<?page no="0"?> Mit Zusatzmaterial The Linguis�c Toolkit for Teachers of English Discovering the Value of Linguis�cs for Foreign Language Teaching Rolf Kreyer <?page no="1"?> Prof. Dr. Rolf Kreyer ist Professor für Sprachwissenschaft des modernen Englisch und zweifacher Lehrpreisträger der Universität Marburg. <?page no="2"?> narr STUDIENBÜCHER <?page no="4"?> Rolf Kreyer The Linguistic Toolkit for Teachers of English Discovering the Value of Linguistics for Foreign Language Teaching <?page no="5"?> DOI: https: / / doi.org/ 10.24053/ 9783823396118 © 2023 · Narr Francke Attempto Verlag GmbH + Co. KG Dischingerweg 5 · D-72070 Tübingen Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikro‐ verfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Alle Informationen in diesem Buch wurden mit großer Sorgfalt erstellt. Fehler können dennoch nicht völlig ausgeschlossen werden. Weder Verlag noch Autor: innen oder Herausgeber: innen übernehmen deshalb eine Gewährleistung für die Korrektheit des Inhaltes und haften nicht für fehlerhafte Angaben und deren Folgen. Diese Publikation enthält gegebenenfalls Links zu externen Inhalten Dritter, auf die weder Verlag noch Autor: innen oder Herausgeber: innen Einfluss haben. Für die Inhalte der verlinkten Seiten sind stets die jeweiligen Anbieter oder Betreibenden der Seiten verantwortlich. Internet: www.narr.de eMail: info@narr.de CPI books GmbH, Leck ISSN 0941-8105 ISBN 978-3-8233-8611-7 (Print) ISBN 978-3-8233-9611-8 (ePDF) ISBN 978-3-8233-0492-0 (ePub) Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http: / / dnb.dnb.de abrufbar. www.fsc.org MIX Papier aus verantwortungsvollen Quellen FSC ® C083411 ® <?page no="6"?> To Marc, for many years of friendship <?page no="8"?> 11 13 1 15 1.1 16 1.2 18 1.3 20 2 25 2.1 27 2.2 28 2.2.1 29 2.2.2 30 2.2.3 32 37 3 39 3.1 41 3.1.1 41 3.1.2 46 3.2 50 3.2.1 50 3.2.2 52 4 59 4.1 59 4.2 63 4.3 67 4.3.1 68 4.3.2 71 4.3.3 72 79 Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “I just want to be a teacher” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Welcome! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . How to use this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A toolkit and what’s in it . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What tools are good for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The art of repairing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Errors and why they are a good thing . . . . . . . . . . . . . . . . . . . . . . . . . . . Interlanguage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What is an error? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “What are you sinking about? ” - sounds in isolation . . . . . . . . Consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The articulatory description of consonants . . . . . . . . . . . . . . . . . . . . . . A comparison of English and German consonants . . . . . . . . . . . . . . . . Vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The articulatory description of vowels . . . . . . . . . . . . . . . . . . . . . . . . . . A comparison of English and German vowels . . . . . . . . . . . . . . . . . . . . “France is bacon” - sounds in speech . . . . . . . . . . . . . . . . . . . . . . . Phonemes, phones and allophones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . When Alastor Moody has a mat eye - allophonic contrasts . . . . . . . . Why France is bacon - connected speech . . . . . . . . . . . . . . . . . . . . . . . . Assimilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Weak forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Linking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Words and beyond . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . <?page no="9"?> 5 81 5.1 81 5.2 82 5.3 87 5.3.1 88 5.3.2 90 5.3.3 91 5.3.4 95 5.4 99 6 105 6.1 108 6.2 113 6.3 118 6.4 121 131 7 133 7.1 134 7.2 136 7.3 140 8 145 8.1 145 8.2 149 8.3 152 8.4 155 8.4.1 155 8.4.2 160 167 9 169 9.1 169 9.2 173 9.2.1 174 9.2.2 180 Nodes and links - the mental lexicon . . . . . . . . . . . . . . . . . . . . . . . . Webs of words, not lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The mental lexicon - some basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stay connected - links in the mental lexicon . . . . . . . . . . . . . . . . . . . . . Defining features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Encyclopaedic relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Mostly) sense relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Collocation and other relations in language use . . . . . . . . . . . . . . . . . . What Taboo can teach us about teaching words . . . . . . . . . . . . . . . . . . Think big! More than words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Why tea can be powerful … and why it shouldn’t . . . . . . . . . . . . . . . . Lexical phrases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patterns and their relevance for the classroom . . . . . . . . . . . . . . . . . . . Teaching patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The corpus in the classroom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The theory of practical corpus analysis . . . . . . . . . . . . . . . . . . . . . The corpus as a collection of authentic language . . . . . . . . . . . . . . . . . Words and their contexts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . More than words - annotation provided in corpora . . . . . . . . . . . . . . A more practical introduction to corpus analysis . . . . . . . . . . . . Wild cards and corpus queries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corpus queries - exploiting lemmatization and tagging . . . . . . . . . . . Corpus queries - collocations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corpora in the classroom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benefits and opportunities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Problems and pitfalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Texts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What makes a text a (good) text . . . . . . . . . . . . . . . . . . . . . . . . . . . . What makes a text a text - standards of textuality . . . . . . . . . . . . . . . . The relevance of text structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Themes and rhemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rhetorical Structure Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Contents <?page no="10"?> 10 189 10.1 191 10.1.1 191 10.1.2 193 10.1.3 196 10.2 198 10.2.1 199 10.2.2 200 10.3 202 11 209 11.1 211 11.2 219 229 12 231 12.1 231 12.2 235 12.2.1 235 12.2.2 242 12.2.3 250 13 253 13.1 254 13.1.1 254 13.1.2 257 13.1.3 259 13.1.4 264 13.1.5 267 13.2 275 279 285 297 301 304 Texts and the relevance of sentence structure . . . . . . . . . . . . . . Principles of text processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The given-before-new principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The principle of end-focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The principle of end-weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Helpful sentence structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-canonical clause patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Secondary clause patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . From text to good text - an example . . . . . . . . . . . . . . . . . . . . . . . . . . . . Texts and the relevance of the situation . . . . . . . . . . . . . . . . . . . . . Spoken and written English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The interpersonal dimension - politeness . . . . . . . . . . . . . . . . . . . . . . . The power of language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The recipient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Humans - the not so rational animals . . . . . . . . . . . . . . . . . . . . . . . . . . . Communication and the unpacking of information . . . . . . . . . . . . . . . Exploiting the Cooperative Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Don’t think of an elephant! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The producer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packaging information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The right word at the right time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Beyond the word level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Of swarms and floods - the power of metaphor . . . . . . . . . . . . . . . . . . Framing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Faulty) logic and argumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Walking uphill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contents 9 <?page no="12"?> Preface This book is an attempt to provide a convincing answer to a fundamental question that many students of English as a foreign language ask or have asked themselves: “Why do I need linguistics? I just want to be a teacher! ” The short answer to this question is “Good linguists are better teachers! ” The long answer is this book. It explores how teachers can benefit from a thorough understanding of basic linguistic concepts and methods and, this way, hopes to show how linguistics can become a very useful toolkit for the English-as-a-foreign-language classroom. To a large extent I have to thank my teaching-degree students for this book. They never stopped challenging the relevance of my lectures and seminars and they always provided me with honest and helpful comments on my teaching. I hope this text will be a valuable resource for them, for future students and for EFL teachers. I would also like to thank Helena Hanneder and Lilly Garlepow for reading earlier versions of this monograph and for providing me with invaluable feedback. Finally, I would like to thank the publisher for considering this text for the narr STUDIENBÜCHER series and, especially, Kathrin Heyng for her advice and encouragement as well as Iris Steinmaier for her patience with my many wishes regarding layout and formatting. Marburg, September 2023 Rolf Kreyer <?page no="14"?> Abstract Contents 1 I just want to be a teacher 2 A toolkit and what’s in it 2.1 What tools are good for 2.2 The art of repairing 2.2.1 Errors and why they are a good thing 2.2.2 Interlanguage 2.2.3 What is an error This part describes my motivation to write this book in the first place. It discusses in which sense linguistics can be regarded as a toolkit for foreign language teachers and, most importantly, it argues that, other things being equal, you will be a better teacher if you are a good linguist. Introduction It hasn’t been made clear at uni, why it should be useful to study linguistics at such depth and how I can incorporate it in my lessons. (A T E A C H E R ) After years of practice, I see linguistics as the centre of lessons, especially when teaching pronunciation, grammar and usage. (A N O T H E R T E A C H E R ) <?page no="16"?> 1 “I just want to be a teacher” I think linguistics is often not regarded as relevant because the connection to the future job is missing in some of the classes. We learn something but are not able to use this knowledge in future teaching. (A S T U D E N T ) [Linguistics] is very important but I realized the importance very late in my studies. The practical orientation of the theory was missing (why do I have to know this as an English teacher? ) (A S T U D E N T T E A C H E R ) My ‘knowledge’ was too academic. Connection to teaching not made clear during university education […] Linguistics was truly horrible at university. (A T E A C H E R ) What this chapter is about You may have experienced linguistics as a somewhat frustrating part of your studies. You have not done any linguistics at school and you know you will not teach it as a teacher. So, it might seem irrelevant for your future job. In your study programme, however, it plays an important role, comparable to literary and cultural studies or language practice. To make things worse, linguistics is not always easy and accessible and can be quite demanding at times. In this chapter you will learn that many students think that way, but you will also learn that linguistics can be very useful for you as a future teacher. <?page no="17"?> linguistics at university … … no linguis‐ tics at school 1.1 Welcome! Welcome to what I hope will turn out to be a successful attempt to convince you, dear reader, that linguistics can contribute to making you a better teacher of English as a foreign language. I wrote this book for several reasons. The first relates to comments like the ones at the beginning of this chapter. The question asked by the student teacher is a very relevant and valid one: “Why do I have to know this as an English teacher? ” Let’s face it: you go to school for 12 to 13 years and will usually have spent at least 10 years learning the English language. In these 10 or more years you focus a lot on learning the language itself, i.e. you learn a lot about words, you are taught about orthography and pronunciation, you have to learn how the simple past and the present perfect are formed and how they are used, your teachers teach you how to form a question or how to negate a sentence and so on. In addition to the practical aspects of speaking and writing English correctly, you do a lot of literary studies. You read short stories, you interpret poems, you analyse the characters of plays and novels etc. Finally, you also explore the cultures of the different countries in which English is spoken. You might have heard about the Royal family in the United Kingdom, you might have learned about the voting system in the USA, you might have heard about the First Nations in Canada or about the Indigenous Peoples of Australia. You have done this for at least 10 years and, obviously, you thought it was so inter‐ esting and fascinating that you decided to become a teacher of English as a foreign language. Well done and congratulations! You are now a student in some EFL-pro‐ gramme at some university (otherwise, you would hardly be reading this book). Within your study programme you recognise a lot of the things that you do from your days at school, even though you do them at a different level now. You improve your English, you keep on reading novels, short stories, poems etc. and analyse the many facets of them, you study cultural phenomena from many different areas of the English-speaking world. So far no surprises. What will be new to you is those parts of the study pro‐ gramme that are concerned with teaching the English language, i.e. didactics. Again, no surprise there, because after all you want to be prepared for your future job as a teacher and learning how to teach the English language will most likely come in handy. However, on top of all of this and as if this wasn’t enough to keep you busy, you find that those who designed your study programme make you spend a significant amount of time with the study of the English language, also referred to as ‘English linguistics’. While you might find English linguistics interesting and fascinating as a subject in its own right (and I’m very glad for you if you do), you might not be so sure about the relevance of English linguistics for your future career. You have spent most of your life in school and you know very well that linguistics was not a part of your English cur‐ riculum. You haven’t learned about phonemes, you have never explored the morphol‐ ogy of the English language and, most likely, you have not discussed the nature of the mental lexicon and you have not learned about collocations or semi-prefabricated 16 1 “I just want to be a teacher” <?page no="18"?> good lin‐ guists make better teachers phrases. What is more, you know that it will not be a part of your school life as a teacher, at least not in the sense that linguistics is something that is taught at school. So, it makes a lot of sense to ask “Why do I have to do all this? I just want to be a teacher! ” Before you read on … … take a minute to try and figure out how you have experienced the linguistic parts of your study so far and to what extent you think linguistics may be relevant for your future career, or not. If you have asked yourself the question “Why do I need to do all this? ”, trust me, you are not alone. In a survey by a colleague and me we found that of the four areas language practice, cultural studies, literature, and linguistics, linguistics is perceived as the least relevant among students, student teachers and teachers alike (Kreyer & Güldenring 2016: 279). Admittedly, we were a little surprised to find that roughly 74 to 80 % of the three groups did find linguistics relevant for future teachers of English, but we also found that our informants regarded the other areas as relevant in 90 % or more of all cases, sometimes even reaching a full 100 %. What is more, we found that the first three areas seem to increase in importance from student to teacher, whereas linguistics seems to become less relevant for teachers. Other researchers come to even more damning conclusions as the following quote shows. In my experience, the waste involved in [Linguistics or Applied Linguistics components of pre-service training courses] can be prodigious: the wasted time and effort of the lecturers, who complain of the apathy or stupidity of their students; the wasted time and effort of the students, who complain of the irrelevance and jargonized complexity of the subject, as well as of the indifference to language teaching of their lecturers. Worst of all is the waste of potential — the continuing alienation of generations of language teachers from an area of knowledge which should be a source of constant support to them. (Edge 1988: 9) Whatever the degree to which you agree with the above, I myself, as a professor of English linguistics, find it very easy to agree with Edge when he laments the waste of potential. Linguistics can be a wonderful and extremely helpful resource for you as a teacher of English as a foreign language. It should and can be “a source of constant support” for you, and it is in this spirit that this book is written. I am actually convinced that, other things being equal, good linguists make better teachers. Obviously, there are many more important things that you need to become a good teacher. For example, the passion that you have for teaching itself but also for what you teach, or your ability to inspire fascination in your students, your ability to create rapport with them and so on. There is no doubt about that and I am not under the delusion that being a good linguist automatically makes you a good teacher. How‐ ever, if you got all of these things under control, if you are good at all of this, you will 1.1 Welcome! 17 <?page no="19"?> linguistics as a skill area be an even better teacher if you are a good linguist. To make linguistics really useful, however, a change of perspective is needed, where “linguistics in the context of TEFL training should be regarded less as content area than as a skill area” (Edge 1988: 11). That is what this book is about. It is supposed to help you become the best teacher you could possibly be by showing the many areas in which your linguistic expertise can help you become more confident and more proficient as a teacher of English as a foreign language. Any of these areas are like compartments in a toolkit and each compartment provides you with tools and material to make you an even more skilful foreign language teacher. The fact that the book was written with linguistics as a skill area in mind means that it does not aim at a comprehensive description of a given content area. Instead if focuses on those aspects of a given content area that will be relevant for future teachers. For example, the book devotes two chapters to phonetic and phonological questions. However, it is not meant to be an introduction to phonetics and phonology. Aspects like acoustic and auditory phonetics, the cardinal vowels or syllable structure, for instance, will be ignored in favour of other aspects that I think are more relevant for the EFL classroom. This does not mean that I regard this publication as some kind of watered-down introduction to linguistics. Quite the contrary: if you want to use your linguistic knowledge as a skillset you need to have a profound and thorough understanding of linguistics. The scenarios that you encounter as a teacher are so varied that no one can prepare you for all the possible situations you might be facing. However, you will be prepared if you have understood the relevant parts of linguistics to such an extent that you can apply your linguistic expertise in any situation that you are confronted with. Linguistics, this way, can become something of a superpower. Or in the words of a teacher of English: As a student, I underestimated the relevance of grammar in EFL-teaching, now, after years of practice, I see linguistics as the centre of lessons, especially when teaching pronunciation, grammar and usage; but now I see what a vital part it is, but it must of course be a vehicle to teach the contents of cultural and literary studies; understanding and producing texts is only possible with a sound knowledge of the English language. (Kreyer & Güldenring 2016: 282) 1.2 An overview We will start off in the next chapter by taking a look at interlanguage and error analysis. In particular, we will find out that the English that your students speak is based on a language system of its own (the interlanguage) and that errors can be very helpful in telling you what your students do not yet know, and, and this may come as a bit of a surprise, often they can tell you what your students do already know. This will then create the background on the basis of which we will explore six different areas of linguistics that I think are especially relevant for you as (future) teachers. The first of 18 1 “I just want to be a teacher” <?page no="20"?> these concerns the sounds of English and German. Chapter 3 will focus on sounds in isolation. After a short introduction into how sounds are produced we will compare the sound inventories of English and German and we will mostly focus on English sounds that can be expected to be difficult for German learners of English, e.g. the th-sounds. Chapter 4 is about how sounds might change when we use them together with other sounds while speaking. A big problem for German learners of English, for example, is the so-called ‘Auslautverhärtung’ or final devoicing, which is a regular process in German but not in English. German learners might, therefore, pronounce bag like back, which, of course, makes a difference. The second large area, words and beyond, looks at words and larger lexical units. We will first explore the mental lexicon. It is best conceived of as a huuuuuge cobweb with tens or even hundreds of thousands of nodes and connections. What these nodes are and how they can be connected is what chapter 5 will be about. Chapter 6 moves beyond the word level and focuses on larger lexical patterns, exploring why it’s okay to speak of a strong tea but not of a powerful tea and explaining why it is good to think big, to go beyond the level of words when you think about vocabulary. A lot of what we explore in the words-and-beyond section of this book is related to a branch of linguistics that studies language by looking at massive amounts of authentic texts, so-called corpora (singular corpus). Corpus linguistics has furthered our understanding of the true nature of language tremendously and recent decades have seen many attempts of putting corpora and corpus-based research to use in the classroom. Accordingly, another major part of this book focuses on the corpus in the classroom. Chapter 7 will first introduce some basics of corpora and corpus-design and explore the theory of practical corpus analysis. That is, it gives you an understanding of how data are stored in a corpus and it helps you understand what you do when you do a corpus query. Chapter 8 provides a more practical introduction to corpus analysis but, similarly to chapter 7, it tries to be as general as possible. So, instead of focusing on one platform or tool only, chapter 8 wants to provide you with a general corpus-tool proficiency which you can, then, apply to any piece of software or interface. This chapter will also explore ways in which corpora can be used in the classroom. In the third part we will move on to texts. We will first discuss, in chapter 9, what it is that makes a text a text and what it is that makes a text a good text. We will also explore ways of analysing and describing the structure of texts. On the basis of this general understanding of text quality we can then zoom in on texts and the relevance of sentence structure (chapter 10). In particular, we will look at principles of text processing and we will explore the options that the English language provides to produce sentences that follow these principles. Chapter 11 will look at the relevance of the situation when it comes to text production. Things to keep in mind, for instance, include whether a text appears as a tweet or as a book, whether it is supposed to be listened to or read, whether you are speaking to a friend or one of your professors, whether you are talking to laypeople or experts etc. Of all of these important situation-related factors, we will be looking at two very important areas, namely the distinction between spoken and written English and questions of 1.2 An overview 19 <?page no="21"?> politeness. The book concludes with two chapters on the power of language. The inclusion of this topic might seem a bit strange. After all, it is not directly related to teaching the English language and it does not usually figure prominently as a topic in the curricula (although it can be an optional topic in some federal states). I have included this topic because I think it becomes more and more relevant in our modern society. Teachers should know about how language can be used to manipulate, so that they can inform their students whenever it is possible. Chapter 12 focuses on the recipient and on those aspects of language processing that make humans vulnerable to manipulation through language. We will find out, for instance, that humans are not as rational as they like to think they are. Chapter 13 sheds light on the many ways in which the producer can shape their message to exploit the vulnerability of the recipient. Consider, for instance, the different effects that a headline like Increasing teaching-related complaints about university teaching from snowflake students! has in comparison to the more neutral Increasing teaching-related complaints about university teaching from students! Framing students as snowflakes, i.e. as persons that think they are special individuals but cannot stand even the slightest bit of pressure or opposition, puts their possibly very reasonable complaints into a certain light. All in all, I hope that the contents I have chosen will be of some use for you, irrespective of whether you are a student, a student teacher or a teacher. 1.3 How to use this book When you go through the text please feel free to make use of it whichever way you think suits you best. Obviously, the six-parts structure of the book suggests to understand each part as some kind of information unit and you will get the most out of each part if you read all of the chapters of that part. Within each of the parts the first chapter prepares some of the grounds for the following chapter(s). It would, therefore, usually not make sense to read the final chapter of a major part without having read the previous one(s). Still, it might make sense for you to skip individual chapters if you are familiar with the content. At the beginning of this chapter you have already seen a section called ‘What this chapter is about’. What this chapter is about The name of this box should explain what it is about. Its primary purpose is to prepare you for the content that is presented in the individual chapters. Even though these sections are usually short, they will still activate relevant knowledge structures in your mind and help you to process the chapter’s content more easily and at a deeper level. In addition, it will support retention, i.e. you will remember more of what goes in. 20 1 “I just want to be a teacher” <?page no="22"?> You will find a box like the one above at the beginning of each chapter and it might help you to decide whether you want to read the chapter or not, although that is not its primary purpose. The respective box for chapter 7, for instance, informs you that chapter 7 provides “a general understanding of what a corpus is and of the kind of information that a corpus may contain in addition to the mere words that you can find in it.” If you are already well-versed in corpus design and basic query methods, there is probably not a lot more that chapter 7 can tell you, so, why not move on to chapter 8? Even within chapters, you might find content that you are familiar with. Feel free to skip it. If you are proficient in corpora and corpus use, for instance, section 8.4., which describes the use of corpora in the classroom, might be the only interesting aspect of the whole part of corpus linguistics and EFL. Within each chapter you will also find boxes of the following kind. By the way: Additional information These boxes provide you with additional information on a topic or with useful hints. You do not have to read these bits to follow the main text. They might be interesting for you but feel free to skip them. You can always come back to them later if you want to. After having talked about what you might not want to read, let’s talk about how to read and, more specifically, about the activities that the book provides. They are titled ‘Before you read on …’. Before you read on… … take a bit of time to consider the relevance of while-reading activities. The book will provide you with a lot of tasks within the actual text. I know from my own experience that as a reader you often do not feel particularly inclined to do these exercises. Please allow me to stress their importance: Of course, you can read the book without doing them, especially since the answers are given in the text but you will really be missing out, because you will process the content of the book a lot less deeply. Or, to put it in the words of one of my favourite memes: Das kannste schon so machen, aber dann isses halt k****! [Thanks to the publishers for leaving that in! ]. At the end of each chapter you will find a summary of the content of the chapter as well as a few select suggestions for further reading. 1.3 How to use this book 21 <?page no="23"?> New tools in your linguistic toolkit The first part of the box at the end of each chapter is a list of the tools that reading this chapter has added to your linguistic toolbox. These ‘tools’ can be research findings, insights into the nature of language or approaches on how linguistics can be put to use in the classroom. Each entry will name the tool and briefly describe what you can get from it. Linguistics is a skill area ▶ This insight suggests to read this text on the background of a particular mindset, i.e. in order to find out how linguistics can help to make you a better teacher. If you want to know more The second part of the box at the end of each chapter provides a short list of suggestions for further reading. The short comments will give you some idea of the content. Obviously, there are other useful texts in addition to the few suggestions I make but I deliberately kept the lists short. I know from my own experience that too many recommendations can sometimes be a little overwhelming. These are the suggestions for this chapter: C A R T E R , Ronald (ed.) (2011): Linguistics and the Teacher. London: Routledge. ▶ This edited volume is a selection of ten papers that demonstrate the relevance of linguistics for teachers. The papers range from theoretical discussions about linguistics and the teaching profession to practical examples, syllabi and course proposals. K R E Y E R , Rolf & Barbara G ÜL D E N R I N G (2016): “’But sometimes this potential is not used at all’. Views on linguistics in EFL teacher training and teaching”. In: R. Kreyer, S. Schaub & B. Güldenring (eds.): Angewandte Linguistik in Schule und Hochschule. Neue Wege für Sprachunterricht und Ausbildung. Frankfurt a.-M., 265-297. ▶ The paper provides insights into how EFL students, student teachers and teachers assess the relevance of linguistics for their (future) profession and reflects on the relevance of linguistics for second language teaching. K R E Y E R , Rolf (2020): “Der sprachwissenschaftliche Blick auf Kommunikation - und was Lehr‐ personen davon haben“. In: N. Meister, U. Hericks, R. Kreyer & R. Laging (eds.): Zur Sache. Die Rolle des Faches in der universitären Lehrerbildung. Das Fach im Diskurs zwischen Fachwissenschaft, Fachdidaktik und Bildungswissenschaften. Wiesbaden, 79-100. ▶ The paper shows why linguistics is still relevant for foreign language teachers even though it is not part of the school curriculum. L I G H B O W N , Patsy M. (2017): “SLA Research and Foreign-Language Teaching”. In: E. Hinkel (ed.): Handbook of Research in Second Language Teaching and Learning. Volume III. London. 103-116. 22 1 “I just want to be a teacher” <?page no="24"?> ▶ The paper explores how findings from SLA research can be useful for foreign language teaching. It argues that an awareness of SLA research can help teachers set more realistic expectations for the classroom. At the end of each chapter you will find a link with additional online material for you. That’s enough for now. Have fun and prepare yourself to discover that linguistics is actually really helpful for you as a future teacher and that it can become your superpower for the EFL classroom! 1.3 How to use this book 23 <?page no="26"?> 2 A toolkit and what’s in it I’m going to use all my tools, my Godgiven ability, and make the best life I can with it. (L E B R O N J A M E S ) You’re going to use all your tools, all of your linguistic expertise, and be the best teacher you can be with it. (R O L F K R E Y E R ) What this chapter is about The present chapter explores how far we can take the linguistics-is-a-toolkit metaphor. From the most general perspective, tools are used to create and repair things. To do this you need to be able to construct, disassemble and analyse. Your linguistic expertise can be used in pretty much the same way. The ‘thing’ you want to construct is the foreign language competence of your students. With people learning a language, of course, this often involves wrong language use, i.e. errors. Errors can be very informative when it comes to language teaching. Obviously, they tell you what your students cannot yet do, but often a wrong construction also tells you what they can already do, what they already know. When it comes to repairing errors, you need to able to analyse what exactly is going on and to disassemble wrong language use. It is only then that you can isolate the exact problem, repair this issue and put it all back together. Linguistics is many things. For me, primarily, it is a fascinating science that helps us to understand that which makes us human, i.e. the ability to use language. Linguistics can also be incredibly educational, for instance, when we look at public and political discourse and when we become aware of the power that language has to manipulate us. Through our understanding of this exploitation of language we can become more resilient towards the many language-based attempts to sway our opinion each and every day. When it comes to teaching a foreign language, linguistics, in my view, is like a toolkit (or skill area). It is a huge box with lots of different things in it. At times it might appear a little chaotic: there are some tools in your toolkit that you are familiar with and you know exactly how to use them, with other tools you can imagine how <?page no="27"?> useless tools useful tools promising tools new uses for old tools they might be put to use, and, lastly, with many of the tools you have no idea what they might be good for and they are or seem useless to you. Starting off with the tools that you do not know what to do with, even I, being a linguist, have to admit that not everything linguistics covers will be useful for your future career as a teacher. For example, the intricacies of X’-theory and genera‐ tive-transformational grammar are not the first things that come to mind when I think of language teaching. Similarly, multivariate analysis of corpus data might not be that helpful in the classroom. Still, a thorough understanding of syntax or concepts of corpus linguistics can be very useful for any teacher of English as a foreign language. More generally, I would contend that you can glean something useful from any linguistic area and I hope that, while reading this book, you will have that experience a lot. On the other end of the spectrum, you have those tools that you feel very confident with, you know exactly what to do with them and you regard them as very useful. From my experience, phonetics and phonology is such an area of linguistics. It doesn’t take a lot to convince my students that knowledge about the articulation of sounds may come in handy if they want to improve their own students’ articulation. Still, it might sometimes be necessary to give some further instruction to enable future teachers to use the tool as skilfully as possible. Between these two kinds of tools in a toolkit, you have a wide range of tools, let’s call them promising tools, of which you do not exactly know what they are good for. You can imagine situations in which they might be helpful, but you are not quite certain how to use them and if you did use them you would not feel confident in using these tools. For example, you might have some knowledge about collocations and you might agree that they are very important for language teaching but you do not really know how to apply your knowledge in the classroom. The present book is about all of the above. It takes a look at those areas of linguistics that are obviously useful for you as a future teacher and wants to help you to get the most out of these areas and maybe even surprise you with some new ways of using a tool that you are familiar with. With regard to those linguistic areas that you do not feel quite confident about, the book will provide you with a thorough understanding of the subject matter and it will illustrate how your expertise can be brought to fruition in the classroom. Finally, even with those linguistic areas that do not really seem to have any connection to the enterprise of language teaching, the book will make suggestions as to how they can be put to use in everyday EFL contexts. Generally, the book is not only about providing you with the necessary skills to deal with the situations that we discuss, but it also, and maybe most importantly, wants to encourage you to think about other ways of making linguistics useful for the foreign language classroom. A good carpenter or a good mechanic can use their tools to address any kind of situation, even those that they have not been explicitly prepared for. Sim‐ ilarly, a good linguist will be well prepared for any kind of language teaching situation that they encounter. I hope that this book will provide you with a useful toolkit that 26 2 A toolkit and what’s in it <?page no="28"?> construct and create disassemble analyse repair you carry in your backpack and that will give you everything you need to confidently deal with any language teaching challenge that you might encounter. 2.1 What tools are good for There are four broad areas where you can put tools to use. You can use them to construct and create things. There is a lot you can do with a hammer, a couple of nails and some pieces of wood, for example. We can also use tools to disassemble things. A hammer, a crowbar or an axe will take you a long way in this respect. The typical toolkit would also contain tools that help you to analyse things. Even such a basic thing as a folding rule, which you only use to measure length or thickness, is a means to analyse an object. A spirit level will tell you whether a table or a wall is slanted or not. Combining all of the three aspects above, tools are used to repair objects. You may notice that something is not quite right. You will then have to analyse the object to see where exactly the problem is, maybe you need to disassemble it to isolate the one part that doesn’t fit and either modify that part or construct a new one. In the end, you can put them back together. To some extent all of this is also true for linguistics as a toolkit. The ‘object’ that you as a teacher construct or create is the competence of the foreign language in your pupils’ minds. It’s a very complicated structure and you need a lot of skill to build it. It helps if you are aware of the large variety of tools that linguistics has to offer to help your students construct this competence. To give you one example. A teacher friend of mine once told me that it was only when he started working as a full-time teacher that he became aware of the usefulness of derivational morphology. This teacher started teaching his pupils the basics of derivational morphology since he thought it might be very useful for them to try and create words for themselves if they needed to express an idea or a concept but didn’t have a word for it. As I see it, this wonderful example shows two things. First, derivational morphology can be fruitfully integrated into a language teaching context even though I doubt that many teachers are aware of it. Secondly, having an awareness of the variety of all the tools that your linguistic toolkit has on offer, increases your creativity and versatility as a teacher. You may have heard the adage “to those that got a hammer, every problem is a nail”, meaning every tool suggests a particular way to look at the world and deal with it. The more tools you have, the better. For example, in my school days, teachers seemed to think that learning vocabulary primarily meant to learn by heart isolated pairs of words. Of course, this is part of expanding your vocabulary, but linguistics has a lot to offer to support and enrich that kind of learning. We will talk about this in more detail later but some discussion of it might be helpful here. 2.1 What tools are good for 27 <?page no="29"?> Before you read on … … try and remember how you were taught vocabulary in school. What did you have to focus on? What kinds of relations between words were made obvious to you (if any)? What kind of tips did your teachers give you when it came to learning new words? Psycholinguistics tells us that the mental lexicon, i.e. the representation of words in your mind, is not at all like the alphabetical lists that you have in dictionaries. The mental lexicon is like a giant cobweb, where each word is linked to a large number of other words on the basis of a fairly huge number of different linking relations. Vocabulary teaching and learning will be more successful if your pupils manage to create such a network with as many links as possible between individual words. One such source of links would be derivational morphology, for example, all the words that end in the suffix -ness. Another kind of link is that between collocations, i.e. words that typically co-occur. For example the two adjectives pretty and handsome mean more or less the same, but pretty would traditionally be used in the context of females whereas handsome would be used with males. What is more, we find that many words typically occur in particular larger structures, so-called prefabricated units. The noun assumption, for instance, would frequently be followed by a that-clause, as in the assumption that linguistics is useless for teachers of English. Another set of links comes from what linguistics calls structuralist semantics, where words are analysed with regard to synonyms (couch - sofa), opposites (hot - cold), wholes and their parts (body - arm), superand subordinates (mammal - cat) and other relations. As a native speaker of your mother tongue you are aware of all of these relations and, consequently, your mental lexicon is a very dense cobweb of words. As a successful language teacher you will help your students build up such a cobweb for the English language. As I said above, competence in any language is a highly complex construction that consists of many different parts, all of which have to be exactly of the right shape, be exactly in the right place and fit together with other parts seamlessly. Obviously, this is not an easy task and in the process of creating this language competence things will inevitably not fit the desired specifications and the construction will have to be repaired. In the language classroom you will know that something is not quite right when your students use the language incorrectly. This is another major area in which your expertise as a linguist will be extremely useful. 2.2 The art of repairing Repairing is a complex process that involves various steps. First of all, you have to notice that something is not the way it should be. Once you’ve done that you need to be able to analyse what is going on and you need to be able to identify the actual problem. 28 2 A toolkit and what’s in it <?page no="30"?> It is only then that you can address this problem and repair the whole structure. This is true for any kind of craft just as much as it is true for language teaching. However, all analogies break down at some point, and it is at this point that our analogy is not quite accurate any more. Whenever you build something as a craftsperson all kinds of errors or mistakes are just that: errors or mistakes that need to be rectified. In a language teaching context and from a linguistic perspective errors are extremely helpful, as we will explore presently. 2.2.1 Errors and why they are a good thing A chapter on errors might be somewhat unexpected in a book of this kind, especially this early in the book. I’ve decided to put it here for two reasons. Firstly, errors will be a big part of your future daily life as a teacher. You need to be able to spot an error as soon as it occurs, you need to know what exactly went wrong and you need to know how to deal with it. All of this needs to happen really fast, in a matter of split seconds. Secondly, errors are usually considered to be something bad, something that is unwanted in the classroom. It seems to show failure both on part of you as the teacher (“I have not done a good job in explaining this to my students! ”) and on the part of the students (“S/ he still hasn’t learned this! ”). However, errors are actually an extremely useful resource for you as a teacher as they provide you with highly relevant information as to the extent to which your student has learned the language, or not. Take a look at the wonderful example of an exchange between the mother and her child below (Corder 1981: 11): (1) M: C: M: C: M: C: M: C: Did Billy have his egg cut up for him at breakfast? Yes, I showeds him. You what? I showed him. You showed him? I seed him. Ah, you saw him. Yes, I saw him. The child’s utterances in the above example point at three problem areas. The first has to do with subject and verb concord and, actually, reveals two difficulties the child has. He or she makes use of the third person present tense singular -s, which is obviously not correct in this case, and the child has not yet learned that the simple past is not marked for concord. A second problem that becomes apparent concerns the lexicon, namely a confusion of the verbs to show and to see. Finally, we see an unawareness of the irregular past tense form of the verb to see. This example also illustrates nicely what can be achieved as a teacher if we deal with errors in the right way. Had the mother merely corrected her child by saying “no, it’s I showed him” only one of the three problem areas would have shown. As Corder (1981: 11) says, “simple provision of the correct form may not always be the only, or indeed 2.2 The art of repairing 29 <?page no="31"?> interlan‐ guage the most effective, form of correction […]. Making a learner try to discover the right form could often be more instructive to both learner and teacher.” Not only does Corder argue for a particular way of treating errors in the classroom, he also advocates a new view on errors. Errors are not some random deviation from a yet to achieve foreign language competence, but they themselves are based on rules, namely on the rules of the language system that the learner himor herself has established so far on the basis of input encountered and teaching received. This language system is usually referred to as the learner’s ‘interlanguage’. 2.2.2 Interlanguage The idea of an interlanguage was introduced into applied linguistics and second lan‐ guage acquisition research by Larry Selinker (1972). It helps us to better understand the process of language learning. A learner starts from his or her first language, either referred to as ‘native language’ (NL) or ‘mother tongue’ and wants to achieve some degree of proficiency in a second language, usually referred to as ‘target language’ (TL). Selinker drew our attention to what happens in between these two languages. He sug‐ gested that errors are not some haphazard amalgamation of native language and target language but are the consequence of a rule system that the learner of the target lan‐ guage has established from the input and the teaching that they have encountered so far. This rule system Selinker refers to as ‘interlanguage’. When it comes to language learning, then, we are not only dealing with two but with three language systems, namely native language, target language and, in between these two, interlanguage (see figure 1). Fig. 2.1: Interlanguage in the learner’s progression from his or her native language to the target language (adapted from James 1998: 3). So, what’s the point, you might ask. What have we gained? One important advantage of looking at the language learning process in this way is a new perspective on what the learner cannot do but also on what the learner can do. Errors are not only a sign of failure, a deviation from a still to be achieved target norm, but they can also tell us a lot about what the learner has already learned. When, in example (1), the child says I seed him, this tells us that the child has already acquired the formation of the regular past tense, this piece of linguistic knowledge is part of the child’s interlanguage. What is not yet part of the interlanguage is that the verb to see has an irregular past 30 2 A toolkit and what’s in it <?page no="32"?> tense form. If something similar happened in your classroom, you could learn from that that you have done a good job in teaching regular past tense formation and that your student has done a good job in learning it. What still remains to be done is to make the learner aware of the exceptions to that rule, as these are not yet included in his or her interlanguage system. All of this is extremely valuable information for you as a teacher. Before you read on … … let’s take a look at the following two examples. The asterisk marks wrong forms, the question mark indicates that the status with regard to correctness is not clear. (a) This is completely *unlogical. (b) ? No fucking way! Try and find out what the use of the form unlogical in (a) tells us about the learner. What has he or she learned already? What about (b)? Is it wrong? Why, or why not? If yes, in which sense? Example (a) is not what a native speaker of English would say and a form that we would like to get rid of at some point in time. Still, the form unlogical shows that the learner has already acquired some understanding of negation in English, namely that it can be done through negation prefixes and that one of these is the prefix un-. This example also illustrates the relevance and value of your linguistic expertise. If you have no idea of derivational morphology, i.e. the study of how we create new words from given words, your only option of dealing with this error is on a word basis resulting in feedback along the lines of “No, this is incorrect, you should say illogical.” Compare this to the much more helpful feedback that you are able to provide if your linguistic superpowers, as it were, kick in: The word unlogical is not quite right. What you did well is that you used a prefix to express that something is not logical. However, you used the wrong prefix. While in German the prefix unis the correct one with the word logisch, in English the adjective logical takes another prefix, namely il-. Maybe we should take a closer look at how we can use prefixes in English to express negation in the near future. Example (b) is a little more complicated, as it is not quite obvious whether we are dealing with an error at all. After all, (b) is an English utterance that, used in the right context and the right circumstances would be an excellent response to, say, a question or a request. In other contexts, such as a classroom context, it would usually be frowned upon and, hence, a piece of unsuccessful language use. Again, the interlanguage system of the learner shows some degree of sophistication because it contains a highly useful and highly idiomatic formula. What the language system lacks, however, are rules about the contexts in which this utterance can be used and in which it cannot be used. 2.2 The art of repairing 31 <?page no="33"?> mistakes are not errors I hope that the above has helped to make clear that it makes sense for you as a future language teacher to keep the concept of interlanguage in mind. It helps you to take a wider perspective on errors. Errors are not merely a sign of failure. Rather, they can provide us with information about the current state of the interlanguage system that the learner has created in their mind. Not only do errors show you where the learner has not yet achieved native-speaker status, but oftentimes they can also show you what the learner has already learned. In this section we have already talked a lot about errors assuming that we all share a similar idea of what an error actually is. However, as often when it comes to the scientific study of something, the situation is a little more complicated than that. This already became obvious in (b) from the task above. Is this an error? If not, why bother? If yes, what exactly is it that is not correct here? We will focus on this question in the next section. 2.2.3 What is an error? From a general perspective we can say that an error is something that happened even though it should not have happened. That is, we have an expectation of how something or someone should behave but the reality is different. In this broad sense, ‘error’ can be equated with ‘something is wrong’ or ‘something has gone wrong’ in comparison with some shared understanding of how things should be. If we apply this broad definition to the language learning context, it is easy to see what our expectation would usually be, namely utterances that are in line with the language system of the target language. If one of your students says something that violates the rules of the English language, as for example the child’s sentence I seed him, we would say this is an error. However, what would you do if your pupils said I seed him, uh saw him? The student in this case has corrected himor herself, he or she obviously knows better. I assume that you can relate to that. At least I can. For instance, I am fully aware of the pronunciation of most words of the English language. What is more, I am even fully aware of the problems that German learners of English have with particular English sounds (as I will prove to you in chapters 3 and 4). Still, sometimes some mispronunciation slips in even though I know exactly what I should be doing. Similarly, if you think about yourself and your use of your native language you will notice that at times you produce a sentence which is not in line with the grammar of your mother tongue. Still, you are a fully competent speaker of your mother tongue. Deviations of this kind are usually referred to as ‘mistakes’. They are phenomena of performance and they can and do happen to everyone from time to time. From a teach‐ er’s perspective, these deviations are, in fact, irrelevant. Recall that your focus as a teacher is on the interlanguage of the learner. You want to know what their interlan‐ guage system currently looks like and which steps you have to take to make this system more similar to that of the target language. The deviations that we are interested in are those deviations that stem from the language system of the interlanguage. They occur 32 2 A toolkit and what’s in it <?page no="34"?> deviation context of use production circumstan‐ ces regularly and systematically in the output of the learner. They are phenomena not of performance but of competence. It is these deviations that we are interested in as lan‐ guage teachers, and these deviations are referred to as ‘errors’. One final problem with our understanding of errors still remains, though. What exactly is a deviation? How would we describe that which we would expect? When is an utterance error-free? When is it not? When we first think about these questions we will most likely think about the correct spelling of words for instance, the correct grammatical form or correct punctuation. However, as example (b) above has made clear, the picture is a little more complicated than that. This is because the competence of a native speaker goes far beyond that which is encoded in a dictionary or a grammar. You as a native speaker of your mother tongue know a lot more about how to use your mother tongue correctly and successfully. For instance, you know when swearing is appropriate and when it is not. You do know how you address friends, colleagues, or your superiors. You do know that some words are more suitable if you want to discuss a topic in an academic setting while you would choose other words if you talked about the same topic with your friends or family. If, then, as above we say that errors are a phenomenon of competence, the challenge is to identify what actually is included in this competence that we want to achieve. You might recall from your introduction to linguistics that Noam Chomsky (1965) defines competence as the knowledge that the ideal speaker-listener has of their language. Paul Lennon, in his definition of errors picks up on this in an elegant way. He defines ‘error’ as a linguistic form or combination of forms which in the same context and under similar conditions of production would, in all likelihood, not be produced by the speakers’ native speakers counterparts. (Lennon, 1991: 182) The beauty of this definition of ‘error’ lies in the fact that it takes a comprehensive perspective, not limiting itself to, say, orthographic or grammatical correctness but understanding appropriateness with regard to all the dimensions that the native speaker of a language is aware of. He focuses on three in particular. The first is the context of use. Example (b), No fucking way! , might be an error, for instance if that was a student’s reaction to the teacher’s request to please open the window, or it might not be an error, e.g. if it was used among friends. It is a question of context and a native speaker would be well aware of the contexts that allow the use of such drastic language and those that do not, just like you know for your mother tongue. The second dimension concerns the production circumstances. Consider the following text messaging ex‐ change. 2.2 The art of repairing 33 <?page no="35"?> native speaker counterpart (2) JP: Why aren’t we all sitting together? RH: Dear Jake, This is the safety protocol. We are less conspicuous traveling as individuals. Sincerely, Raymond Holt JP: You don’t have to sign your name on texts. RH: Dear Jake, suggestion noted. Sincerely, Raymond Holt RH: Dear Jake, There’s a man sitting three rows behind me whom I saw at the library. Sincerely, Raymond Holt JP: What should we do? RH: Dear Jake, At the next stop, grab Kevin and hastily exit the bus. Got it? Sincerely, Raymond Holt You may have guessed that example (2) is taken from the excellent TV series Brooklyn 99. The comic effect of this exchange between Captain Raymond Holt and Detective Jake Peralta resides in Holt’s language use. There is nothing wrong with his texts in the strict sense but the formal greeting and goodbye formulae are not suited for communication via text messaging, even more so since time might be a concern; after all, Peralta, Holt and Holt’s husband are being chased by a criminal who wants to see the husband dead. Lennon introduces a third dimension when he takes into account the “native speaker counterpart” of the learner. This dimension points at person-related sociolinguistic aspects. The age of the speaker, for instance, comes into play here. We all have expe‐ rienced the feeling of embarrassment when people use words and phrases that are not appropriate for their age. For example, while you will find it perfectly normal and acceptable to hear a 25 years old person say “das ist echt cringe” you might feel a hot wave of embarrassment if you heard me, 49 years of age at the time of writing this, use this wording. In Germany, 49 year old men do not use that word. Lennon’s definition of ‘error’ draws our attention to the many different levels and areas in which an utterance can be wrong, ranging from what we would typically assume to be relevant, such as pronunciation, spelling and grammar to the more intricate aspects of language use, such as context, communication channels and aspects of sociolinguistics, to name but a few. All of these aspects are relevant and all of them make a valuable contribution to the linguistic toolkit that will be at your disposal for all of your professional career. Remember: other things being equal, good linguists are better teachers! Enjoy your journey to becoming the best foreign language teacher that you can possibly be! New tools in your toolkit Linguistics as a skill area ▶ You have been introduced to a (maybe new) view on linguistics which claims that your linguistic expertise will help to make you the best teacher you can possibly be. 34 2 A toolkit and what’s in it <?page no="36"?> Errors ▶ You have learned about a new view on errors where errors are not merely something that needs to be eradicated but evidence of an evolving foreign language system. Interlanguage ▶ You have learned that underlying all the utterances of your students is an evolving language system that consists of rules and units influenced by your students’ native language as well as the target language. The interlanguage follows its own rules and it is your job as a teacher to bring the interlanguage rule system closer to the target language one. Lennon’s definition of ‘error’ ▶ This comprehensive definition of ‘error’ widens your understanding of the concept and, accordingly, gives you a much better idea of what correct or successful language use encompasses. If you want to know more C O R D E R , Stephen Pit (ed.) (1981): Error Analysis and Interlanguage. Oxford: Oxford University Press. ▶ A collection of interesting articles by Corder, who argues for a new view on errors. J A M E S , Carl (1998): Errors in Language, Learning and Use. Exploring Error Analysis. London: Routledge. ▶ A very detailed account of error analysis. Demonstrates how a thorough grounding in linguistics makes understanding and dealing with errors a lot easier. 2.2 The art of repairing 35 <?page no="38"?> Abstract Contents 3 “What are you sinking about? ” - sounds in isolation 3.1 Consonants 3.1.1 The articulatory description of consonants 3.1.2 A comparison of English and German consonants 3.2 Vowels 3.2.1 The articulatory description of vowels 3.2.2 A comparison of English and German vowels 4 “France is bacon” - sounds in speech 4.1 Phonemes, phones and allophones 4.2 When Alastor Moody has a mat eye - allophonic contrasts 4.3.1 Why France is bacon - connected speech 4.3.1 Assimilation 4.3.2 Weak forms 4.3.3 Linking The part shows how a thorough understanding of how we produce sounds and how the English sound system differs from the German one can be useful for the EFL classroom. It explores some of the most relevant pronunciation problems and shows how your linguistic expertise can help you to improve the pronunciation of your students. Sounds Senk ju vor träwelling (A B O O K T I T L E ) <?page no="40"?> 3 “What are you sinking about? ” - sounds in isolation - Mayday! Mayday! Hello. Can you hear us? Can you hear us? Over! We are sinking! We are sinking! - Hello. Sis iss se tschöamen Kohstgaht. - We are sinking! We’re sinking! - Wot ah ju … sinking abaut? ( F R O M A L A N G U A G E S C H O O L C O M M E R C IA L O N Y O U T U B E ) What this chapter is about This chapter starts off with an overview of articulatory phonetics, the description of how sounds are produced. On this basis we will, then, compare English and German consonants and vowels and isolate those sounds that are particularly problematic for German learners of English. The chapter also suggests ways of dealing with these problematic sounds in the classroom. The little dialogue above is taken from a Youtube commercial for a language school. It is a young man’s first day on the job, when he receives a distress call from a sinking ship. Obviously, the young coast guard has a very heavy German accent when he speaks. Among many errors we also see the notorious problem with the English th-sound. Unfortunately for the sinking ship, the coast guard seems to assume that this is the correct pronunciation and, consequently, understands we are sinking as we are thinking. While pronunciation problems will hardly (most likely and hopefully: never) lead to such disastrous consequences, it is still a very important area of language teaching, particularly as a strong German accent will usually be noticed, sometimes even ridiculed, particularly if you are a politician, as the case of Annalena Baerbock (the German Foreign Minister at the time of writing) shows. Imagine Baerbock had been one of your pupils and you could have helped her to improve her pronunciation (which is actually not that bad)! But how do we do that, improve someone’s pronunciation? Recently I asked a friend of mine (both of us have studied English as a Foreign Language) who is now a teacher how he helps students with pronunciation problems, particularly, what he would do to help them pronounce the th-sounds correctly. His answer was a mere “I repeat the correct pronunciation again and again and at some point they get it.” <?page no="41"?> three prob‐ lem areas sound in‐ ventory allophonic contrasts final devoicing connected speech While this, I assume, is what happens in many EFL classrooms this approach does not seem to be optimal. Just imagine this would have been how you had been taught to drive a car. Basically, there would have been nothing left for you to learn because you most likely have seen your parents or your older siblings drive a car many, many times before you started learning how to drive. Or, to really drive the point home, imagine you had to undergo some kind of surgery and your surgeon told you “I have never performed this surgery before, no one has taught me how to do it, but I have watched it at least 200 times. So, trust me, you are in good hands! ” Does not sound particularly professional, does it? We’d never accept that kind of approach when it comes to our health. Why should we accept it when it comes to teaching a language? Regarding the challenges of teaching pronunciation in the EFL classroom, I think we professors at university do not always do the best job in pointing out how useful phonetics and phonology can actually be for your future career as a teacher. Instead, we bore you with anatomical aspects of pronunciation (“I didn’t sign up for a course in biology! ”) and we pester you with strange terms like voiceless apicodental fricative or voiced velarized lateral approximant (“You got to be kidding me! Is that really necessary? ”). Actually, all of this (and much more) is relevant to your job and can be a very useful resource for you as I will show in the remainder of this chapter. In particular, we will be looking at what from my experience are the three most relevant problem areas when it comes to pronunciation. The first concerns differences in the sound inventories of English and German. If the sound inventories of two languages differ, i.e. if one language has a sound that the other language does not have, we would usually assume that this causes problems. A good example in this context are the two th-sounds of the English language. They are not part of the German sound inventory and, as we all know, can be quite difficult for some German learners of English. The second problem area is what I like to call ‘allophonic contrasts’. It has to do with differences in which the phonemes of German and English are realised as phones in actual language use. If you are not quite sure what that means, don’t panic, we will take care of this soon. Until then the following example should be helpful. In German, for instance, the two words Rad and Rat are pronounced identically, even though the final phoneme in the first word is actually a / d/ . If you don’t believe me, just create the plural of the two words. The reason why we pronounce Rad like Rat is a phonetic process called final devoicing, or Auslautverhärtung in Ger‐ man. Simply speaking, in German we devoice some voiced consonants at the end of syllables - I’ll provide a more detailed description in the next chapter. For the German language this is a regular process, the English language does not do that, as you can easily see if you compare the two words bad and bat. To pronounce the ‘d’ as a [d] and not as a [t] is really difficult for German learners of English. A last problem area has to do with phenomena that arise in connected speech, i.e. when words are combined to form longer utterances. To give just one example, the English language usually links words in speech whenever possible, while German pronounces each word in isolation. I have found that even very advanced learners of English who show an excellent pro‐ 40 3 “What are you sinking about? ” - sounds in isolation <?page no="42"?> articulatory phonetics consonants and vowels voicing nunciation in all other areas often are not aware of this. For example, the phrase keep out would rather be pronounced as key pout, nude art would be pronounced like new dart (as those of you who have seen How I Met your Mother will know). Linking words in this way (and in others that will be discussed in section 4.3.3) will make you sound a lot more authentic and, in addition, in many cases it can solve the problem of final devoicing. For the remainder of the chapter we will focus on the first problem area, differences in the English and the German sound inventories and probable difficulties arising from those differences. More specifically, a focus will be laid on those sounds that the English language has but the German language does not. The problem with this set of sounds lies in the fact that the pronunciation of them makes use of movements of the tongue and other articulators that German learners are not used to. To understand the difficulties we need to understand how sounds are produced. We will make a first distinction, here, between consonants and vowels. 3.1 Consonants Sounds can be described from a variety of perspectives. For you as future teachers the most relevant one is that of articulation, i.e. the question “How is a sound produced? ”. This question is explored in the field of articulatory phonetics, the scientific study of the production of language sounds. English sounds are produced egressively, i.e. by exhaling air through the mouth and nose. As the air passes through, the articulators, e.g. different parts of the tongue, the lips or the teeth, interfere in different ways with the stream of air, thus, creating the different sounds of human language. A first basic distinction is that between consonants and vowels. With the first class the articulators obstruct the airstream to a larger or lesser extent, whereas with the second class the airstream is only modulated but not obstructed. This leads to two different descriptive systems for the two sound classes. The present section will deal with consonants. 3.1.1 The articulatory description of consonants When we describe how we produce consonants we make use of three features. The first feature is that of voicing. Whether a consonant is voiced or not has to do with the position of the vocal folds in the glottis. If the glottis is fully open the air can stream through this opening and we will perceive the sound as an unvoiced sound. If the glottis, in contrast, is almost closed, the vocal folds will only leave little room for the air to pass through, which results in a vibration of the vocal folds. This vibration is what we perceive as voice. 3.1 Consonants 41 <?page no="43"?> Before you read on … … lightly touch your Adam’s apple with your fingertips. Now alternate the voiced s-sound [z] as in zoo with the unvoiced s-sound [s] and in Sue. What can you feel? Can you think of a way in which that might be useful to explain the regular plural formation to your pupils? (The answer to the second question will be given a little later in this chapter.) When you pronounce the voiced -s you can feel vibration in your fingertips which you cannot feel when you pronounce the unvoiced -s. The vibration that you feel is the vibration of the vocal folds. Another way to experience this vibration is if you alternate between the two sounds but this time close your ears with your index fingers. With the voiced -s you will now hear a loud humming sound in your head which you cannot hear when you articulate the unvoiced -s. By the way: Latin everywhere Many of the terms that you will encounter in the following are based on Latin or contain elements of Latin. For me personally it is easier to remember them if I understand the Latin behind them. For example, the English language has a class of so-called interdental sounds. You all know that inter means ‘between’, as for instance in international. You also know that the dentist is the doctor that takes care of your teeth. An interdental sound, then, is a sound that is created by putting something (in this case the tongue) between your teeth. If you try to understand the meaning of the terms that describe the sounds, these terms become a description of how the sound is pronounced. Below is a list of the Latin roots of (and/ or English words related to) the terms that describe the articulation of sounds. Those words that are merely based on anatomy terms (palate, velum, uvula, glottis) are not included affricate lat. affricare - ‘to rub one thing against another’ apicolat. apex - ‘top’, ‘point’; an apex predator is the predator at the top of the food chain; in the context of the anatomy of the tongue, the word apex refers to the tip of the tongue. approximant lat. approximare - ‘get close to’; two articulators move towards each other, but not so close as to cause friction or a full blockage of the airstream. alveolar lat. alveolus - ‘a small basin, a hollow’; the alveolar ridge is the ridge of the small basins that are the sockets for the teeth. bilat. bi- - ‘two’ dental lat. dens - ‘tooth’ flex lat. flectere - ‘to bend, to curve’; compare engl. to flex fricative lat. fricare - ‘to rub’ 42 3 “What are you sinking about? ” - sounds in isolation <?page no="44"?> manner of articulation interlat. inter - ‘between, among’ labial, labiolat. labium - ‘lip’ lateral lat. latus - ‘side’ lat. lateralis - ‘of or on side of the body’ nasal lat. nasus - ‘nose’ plosive engl. explosion. After you keep the air from moving you let it explode. postlat. post - ‘behind’ retrolat. retro - ‘backwards’, ‘to the rear’ The second feature is the manner of articulation. It describes how the airstream is obstructed. For instance, two articulators may be brought together closely, resulting in a hissing sound. This group of sounds is called ‘fricative’, since the gap between the two articulators is so narrow that it causes friction. Table 3.1 provides an overview of the main manners of articulation together with a description and the relevant sounds of the English, and where relevant German, language. With the example column, voiceless sounds appear on the left, voiced sounds on the right. Manner of articulation Articulation Sound symbols and examples plosive/ stop the speaker closes off their mouth and nose completely and this way builds up pressure, which is then released suddenly resulting in an ‘explosion’ of air. The glottal stop [ʔ] is used in some accents of English instead of [t]. It is also the sound used to indicate gender inclusive language use, for example when you say Student_in. [p] - pen [t] - toad [k] - king [ʔ] - butter [b] - bet [d] - toad [g] - gun fricative the speaker gives the air a little room to escape between two articulators (e.g. upper teeth and lower lip), which will cause audible friction. [f] - fast [θ] - thin [s] - so [ʃ] - ship [h] - help [v] - vast [ð] - this [z] - zone [ʒ] - genre [ʁ] - richtig affricate the speaker combines a plosive and a fricative as one single sound. [tʃ] - chip [dʒ] - Jeff nasal the speaker keeps their mouth closed but lets the air escape through the nose by lowering their velum. [See what happens if you pronounce an [m] and close off your nose with your fingers.] [m] - man [n] - nice [ŋ] - king 3.1 Consonants 43 <?page no="45"?> place of articulation Manner of articulation Articulation Sound symbols and examples flap/ tap the speaker quickly hits the alveolar ridge with the tip of the tongue. You will hear this sound in American English. [ɾ] - better trill the tip of the tongue or the uvula vibrate in the airstream, resulting in something like a ‘rattling’ sound. You can hear it in two versions of the German ‘r’-sound (gerolltes ‘r’), where the trill is either produced with the tip of the tongue either at the teeth, or at or behind the alveolar ridge or with the back of the tongue close to the uvula. The [r] is the r-sound you hear in Scottish English. [r] - richtig [ʀ] - richtig approximant the speaker obstructs the flow of air but not as much as in the case of fricatives. [ɹ] - right 1 [ɻ] - right 2 [j] - young [w] - wait lateral approximant the air has to escape on the left and right of the tongue, where it has enough room so that it does not cause friction. Note that there are two versions of the l-sound in English. We will discuss them in detail in section 4.2. [l] - leaf Table 3.1: Manner of articulation ( 1 RP, 2 GenAm). The third dimension we need to consider is the place of articulation. It describes where the obstruction of the stream of air happens. Usually, we start at the front with our lips and move backwards and downwards until we have reached the glottis. Table 3.2 pro‐ vides an overview together with a description of the articulation process and examples. Note that there is one sound, [w], that has two places of articulation. Place of articulation Articulation Sound symbols and examples bilabial produced by bringing both lips together. [p] - pen [b] - bet [m] - man [w] - wet * labiodental produced by bringing upper teeth and lower lip together. [f] - fast [v] - vast 44 3 “What are you sinking about? ” - sounds in isolation <?page no="46"?> Place of articulation Articulation Sound symbols and examples dental produced by the tongue and the teeth together. With the th-sounds the tongue is between (in‐ ter) the teeth (dental), more specifically, the tip of the tongue (apex). Hence, these sounds are sometimes called ‘interdental’ or ‘apico-dental’. [θ] - thin [ð] - this alveolar the tip of the tongue is placed at or near the alveolar ridge. [t] - toad [s] - so [d] - toad [r] - right 1 [ɾ] - better 2 [z] - zone [n] - nice [l] - leaf [ɹ] - right postalveolar the tip of the tongue is placed at the back of or a little behind the alveolar ridge. [ʃ] - ship [ʒ] - genre retroflex the tongue is curled backwards and the tip of the tongue touches the hard palate, e.g. the American ‘r’-sound. [ɻ] -right 2 palatal the front of the tongue is pulled back towards the hard palate. Notice that the tongue does not curl backwards. [j] - young velar the back of the tongue is raised to the velum (hence: ‘velar’), which is behind the hard palate. [k] - king [g] - gun [ŋ] - king [w] - wet * uvular the back or the root of the tongue is moved to the uvula (hence: ‘uvular’), the bit of tissue hanging down at the back of your mouth [ʁ] - richtig [ʀ] - richtig glottal produced through manipulation of the vocal folds in the glottis. The glottal stop [ʔ] is used in some accents of English instead of [t]. It is also the sound used to indicate inclusive language use, as for example when you say Student_in. [h] - help [ʔ] - butter Table 3.2: Places of articulation. *Note that [w] has two places of articulation and is, hence, described as a voiced labial-velar approximant ( 1 Scot. English, 2 GenAm). 3.1 Consonants 45 <?page no="47"?> the th-sounds You can feel some of the above places of articulation quite well when you change from one sound to another one. I find this most striking with alveolar, postalveolar and retroflex sounds. For instance, when you move through the three sounds [s], [ʃ] and [ɻ], you can feel quite clearly how the tip of the tongue moves in your mouth. 3.1.2 A comparison of English and German consonants Table 3.3 provides an overview of the consonants of German, of British English (or RP, i.e. Received Pronunciation) and American English (or GenAm, i.e. General American). As before, sounds on the left of a cell are voiceless, sounds on the right are voiced. Those sounds that are shared by German, RP and GenAm are in black. You can see that this is the vast majority and we can assume that there will be no difficulties for German learners of English, although there is one exception as we will see. The consonants in blue are the most important ones for us since these are consonants that we find in RP or GenAm but not in the German language. It is here that we would expect most of the problems to occur, and this is actually the case, although not all of them pose severe problems for German learners of English. Place of articulation bilabial labiodental dental alveolar postalveolar retroflex palatal velar uvular glottal Manner of articulation plosive/ stop p b t d k g ʔ fricative f v θ ð s z ʃ ʒ ʁ h affricate tʃ dʒ nasal m n ŋ flap/ tap ɾ approximant w * ɹ ɻ j lateral approx. l Table 3.3: English and German consonants. *Note that [w] is described as a voiced labial-velar approximant, i.e. it has two places of articulation. Let’s start with the most pressing problem first, that of the dental (or interdental/ apicodental) fricatives. As you can see, the German language does not have these but German, just like English, has three pairs of sounds that are similar. Both, the labio‐ 46 3 “What are you sinking about? ” - sounds in isolation <?page no="48"?> the r-sounds dental and the alveolar fricatives share the manner of articulation and have a place of articulation that is adjacent to that of the dental fricatives. The alveolar plosives have a different manner of articulation, but their place of articulation is nearby. Given this similarity in the place and the manner of articulation it does not come as a surprise that German learners of English frequently substitute the dental fricatives with these three pairs of sounds. Your linguistic expertise helps you understand what is actually going on behind the scenes. In addition, your linguistic expertise can also help you teach your students how to pronounce the two sounds correctly. In this case, a lot of information is hidden in the term ‘interdental’ or ‘apicodental fricative’. Let’s start with the fricative part first. This means that we can only leave a little room for the air to pass through, because this is the way we cause the audible friction that gives this class of sounds its name. ‘Interdental’ tells us that we have to put something between (inter) the teeth (dental). The alternative name for this class of sounds, ‘apicodental’, points to the fact that it is only the tip of the tongue (apex) that interacts with the teeth, and not the whole tongue. You see, that these admittedly not quite trivial terms can contain a lot of information that can make you better at doing your job. You just have to invest a little effort to fully understand the names that we give them. Once you’ve done that, it is actually quite easy. Before you read on … … try out the following. It’s an excellent way of teaching the th-sounds: Open your mouth slightly, put your index finger in front of your lips and move your tongue forwards until the tip of your tongue lightly touches your finger. Let your teeth very lightly touch your tongue and then exhale. It might also help to start off from what your pupils already know - and these are the alveolar fricatives, which fortunately come closest to the target sound. If they produce an s-sound, which, being part of the German language, they do not have any problem with, they are basically already doing everything right. Instead of creating the narrow gap through which the air can pass behind their teeth or at the alveolar ridge, they just need to put the tip of their tongue a little further forward and the s-sound will turn into a th-sound. Again, feel free to try that out for yourself. Start with a voiced or voiceless ‘s’ and slowly move your tongue forwards until the tip of the tongue is between the teeth. Another pair of sounds that may frequently cause problems is the voiced alveolar approximants, i.e. the r-sound in RP, [ɹ], and its retroflex counterpart in GenAm, [ɻ]. Especially German learners of English in their early years are likely to replace it by the uvular fricative, [ʁ]. Again, your linguistic expertise can help you teach the correct pronunciation. In particular, you can direct the attention of your students to the place of articulation. The German r-sound is clearly perceptible in the back of our throat when we pronounce words like reiten or richtig. The sound we are looking for in RP is 3.1 Consonants 47 <?page no="49"?> the voiced alveolar tap alveolar, which is a lot further to the front and which is created with the help of the tip of the tongue. Luckily, the German language has a sound that is quite similar, namely the alveolar trill, [r]. You would normally not expect them to use that sound on a regular basis, but they will be able to produce a somewhat ‘exaggerated’ r-sound that sounds a little ‘rattling’. While producing this sound, the tip of the tongue can be felt vibrating clearly at the alveolar ridge. This way they will at least get a feel for the position of the tongue that is involved when the correct r-sound of RP is produced. The final step can then easily be reached when you tell them to lower the tip of the tongue slightly so that the tip does not vibrate anymore. The r-sound of GenAm does usually not cause that much of a problem, as is shown by the fact that it is one of the sounds that Germans who do not speak any English regularly produce it if they want to mimic the sound of American English. Nevertheless, it might be worth your while to make your students aware of the retroflex nature of the sound. This is also important since in many American English accents vowels are rhoticised or r-coloured if followed by an ‘r’. Before you read on… … pronounce the word dart with an American accent. In particular, focus on the change in position of the tip of the tongue as it moves from the voiced alveolar plosive, [d], to the unvoiced alveolar plosive [t]. What can you feel? With dart you can feel the tip of the tongue curling backwards and back as you move from [d] to [t]. This is the retroflection of the tongue, which leads to the r-colouring of the vowel in between. As always, try it out for yourself, and see what you can feel. Fig. 3.1: voiced alveolar plosive [d] and voiced retroflex approximant [ɻ] The voiced alveolar tap or flap, [ɾ], would usually not be expected to be difficult for German learners of English. Basically, it is like a very quick voiced alveolar plosive, [d], where we move the tip of the tongue as soon as it has made contact with the alveolar 48 3 “What are you sinking about? ” - sounds in isolation <?page no="50"?> the voiced labial-velar approxi mant mirror neurons ridge so that there is no time to build up any real air pressure that can be released in an explosive manner. The last sound we will take a closer look at, the voiced labial-velar approximant [w], is problematic, but in a special way. Usually, the students will not have any particularly severe problems with the sound itself, but German learners of English tend to overuse it. More specifically, often they would use this sound if the correct sound is the voiced labiodental fricative [v]. This is really strange, because the German language, as you know by now, also has the voiced labiodental fricative. Anyway, for a long time I was not really aware that I had this problem too. For instance, I would say uniwersity instead of university or wery wague instead of very vague. There might be a good chance that you are not aware of this problem either. It is one problem that I frequently encounter with my university students, often even in those cases where their pronunciation is almost perfect otherwise. It took me a while to get rid of this error (please recall that an error is a matter of competence, i.e. I did not know any better). But even now when I do know better, it is a mistake that might happen to me occasionally. I might, for instance say inwisible instead of invisible. For whatever reason, it seems to be a problem that appears difficult to fully eradicate. What can help you get a grip on these two sounds is, obviously, to practice pronouncing pairs or sequences of words that all start with either [v] or [w], such as very vague, a very westerly wind, will Vance wait while visiting Vanessa, and so on. I am sure you can come up with many more examples of that kind, and if you cannot, the internet will definitely provide many useful exercises for yourself, should you need them, and for your pupils. Before you read on … … look in the mirror (or use your smartphone) as you pronounce the phrase very well - obviously, you need to get the initial consonants right. Take a look at the shape of your lips and teeth. Can you see a difference? With the exercise above, I would like to draw your attention to mirror neurons. Re‐ search has shown that when humans see somebody do something, the perceived action results in the activation of neurons that are involved in performing this action. You can exploit this when working with sequences like the ones above. Simply stand in front of the mirror and look at your mouth when you pronounce the sequences. The shape of your lips and the configuration of lower lip and upper teeth, respectively, are quite distinct. Watching how your articulators move will support the distinction between the two sounds and their respective motor routines. This way, it will be a lot easier to sort out the problem of confusing [v] and [w] once and for all. This concludes our exploration of potentially problematic English consonants for German learners of English. We will now move to the discussion of problematic vowels. 3.1 Consonants 49 <?page no="51"?> 3.2 Vowels Above we have seen that the articulation of vowels, in contrast to consonants, does not entail any obstruction of the airstream. All articulators leave enough room for the air to pass through unhindered. They only modulate the stream of air, leading to the different vowels that we know. Since they are produced in a completely different fashion and since all English vowels are voiced anyway, we need a different system in order to describe them appropriately. This system will be presented in 3.2.1 and will serve as the basis for comparing the English and German vowel system in 3.2.2. 3.2.1 The articulatory description of vowels The articulatory description of vowels is also based on three variables, albeit completely different ones from those necessary for consonants. The first variable is the position of the highest point of the tongue in the oral cavity. If you recall the vowel chart from your introduction, that is what it shows (don’t worry we’ll take a look at that presently). The second is whether a vowel is long or short. This is indicated by a little symbol that looks a bit like a colon, but is actually two small triangles pointing at each other. So, [a] is the short German ‘a’ as in Bann and [aː] the long ‘a’ as in Bahn. Usually, differences in length also involve slight differences in position of the highest point of the tongue. The long ‘u’ and ‘i’, for instance, are a bit higher and a little less central than their respective short counterparts. The third feature is whether the lips are rounded or not. Lip rounding does not play an important role for the English language but it does for the German language, since some vowel pairs show the exact same tongue position and only differ in whether the lips are rounded or not. Before you read on… … try this: Pronounce a long German ‘i’ (e.g. fielen) and, while you keep exhaling, slowly round your lips. What happens? Do the same while pronouncing a long ‘e’ (e.g. Sehne). In order to describe the position of the tongue in the oral cavity linguists make use of what they call the vowel chart, see figure 3.2. This chart is an abstract representation of all the positions that the highest point of the tongue can be in when producing the vowels of a given language. As you can see, the chart is not symmetrical. This is because the upper front teeth are in front of the lower front teeth. As a consequence the tongue has more space to move in the upper front part of the oral cavity than in the lower front part of the oral cavity. The chart has two dimensions, one horizontal, the other vertical. The lines within the chart help us roughly describe the position of the highest point of the tongue with regard to these two dimensions. On the horizontal axis we can describe vowels as ‘front’, ‘central’ and ‘back’ (moving from left to right), 50 3 “What are you sinking about? ” - sounds in isolation <?page no="52"?> and on the vertical axis we describe them as ‘close’, ‘close-mid’, ‘open-mid’ and ‘open’. The category labels for the vertical axis are due to the fact that the high or low position of the tip of the tongue corresponds to different degrees of closedness and openness of the mouth. Fig. 3.2: The vowel chart as a representation of the range of tongue positions in the oral cavity. The vowel chart might still seem a little abstract but things will become clearer quickly when we take a look at the vowels of English in figure 3.3. The position of the highest point of the tongue is shown for the four front vowels and the uppermost four back vowels in figure 3.3. Before you read on … … try the following exercise. Slowly pronounce the words seat, sit, set and sat and see if you can feel your tongue moving down inside your mouth as shown in the pictures on the left-hand side of figure 3.3. Can you feel how your mouth opens quite strongly when you move from set to sat? Do the same with the words zoo, put, saw and pot. Can you feel the tongue moving? What about if you pronounce see and Sue? Can you feel a backward movement of your tongue? Maybe the surrounding consonants are distracting. If that is the case, just pronounce the vowels on their own. 3.2 Vowels 51 <?page no="53"?> Fig. 3.3: The vowels of RP. The blue circles indicate that there is some degree of variation with regard to the the exact tongue position. Note that some scholars and many dictionaries use the symbol [e] to refer to the vowel sound in set. The red circles mark those vowels that are most likely to cause problems for German learners of English. This exercise has hopefully made clear how the vowel chart has to be interpreted. What is more, I hope that you could feel how the tongue moves inside your mouth as you pronounce the individual vowels. If that is the case, you are well prepared for the next section where we discuss differences between the English and German vowel system. If you are still not quite sure, I strongly recommend (even if you have understood all of it) to take a look at this awesome website: https: / / www.seeingspeech.ac.uk/ ip a-charts/ . It provides you with a clickable IPA chart which gives you amazing MRI, animations and ultrasound videos of speakers pronouncing consonants and vowels. It’s really fascinating and I strongly recommend to go and take a look! ! ! 3.2.2 A comparison of English and German vowels Figure 3.4 provides the vowels of German. In this chart you will sometimes see the pipe-character ‘ | ’. In those cases the sound on the left is a sound with unrounded lips, the sound on the right its counterpart with lip rounding. For instance the bold vowels in viel Gefühl. 52 3 “What are you sinking about? ” - sounds in isolation <?page no="54"?> minor differ‐ ences Fig. 3.4: The vowels of German. A comparison of the two vowel charts in figures 3.3 and 3.4 quickly reveals a large degree of overlap. There are some minute differences between the English and the German language which do not show in the two figures. For example, the ‘i’sounds are a little higher in German than they are in English. Similarly, the short ‘o’-sounds are a little different, which is shown in the fact that we even have two different vowel symbols for them. The German [ɔ] is a higher than the English [ɒ]. As a final minor problem, consider the two long ‘a’-sounds of English and German. Both are open, but the English sound is further back. Before you read on … … see if you can feel the difference during pronunciation. Compare: ich bin - the bin das Schott - the shot die Bahn - the barn With the first two pairs you might feel that your tongue moves fractionally down. In the third case, recognise that your tongue moves slightly back and forwards as you shift from the first word to the second and back. This is no surprise. Both vowels are open vowels, so the degree of openness of your mouth doesn’t change, but the English vowel [ɑ: ] is a back vowel whereas the German counterpart [aː] is a central vowel. Minor differences like these might be relevant for you personally as you move from a very good to an excellent pronunciation. However, there are more pressing problems for less advanced learners of English. They have to do with the three vowels circled in red in the English vowel chart. Let’s start with what I think is the major problem, 3.2 Vowels 53 <?page no="55"?> major differ‐ ences: [ɛ] in‐ stead of [æ] major differ‐ ences: [øː] instead of [ɜː] major differ‐ ences: [ɐ] instead of [ǝ] namely the ash-sound [æ]. This is the sound that we have in words like cat, mat or bat. As you can see, it is missing in the German vowel chart. Typically, German learners would use a vowel that is produced at roughly the same position with regard to front and back but a little higher, i.e. instead of an open position of the mouth the learner has his or her mouth only half open. This results in the sound [ɛ], which we have in words like Zen, Kette, Mett or Bett. Accordingly, the English sentence the cat sat on the mat would sound something like this: the kett sett on the mett. The confusion of these two sounds can be quite problematic since it might lead to misunderstanding: I lost the bet is different in meaning from I lost the bat. And while My bad! is a good way to apologise, My bed! might rather be interpreted as some kind of weird invitation. For‐ tunately, the problem is easily solved. What German learners of English should focus on is just to open their mouths a little wider, as this will automatically lead to a lowering of the tongue which already takes us very close to the intended vowel. If then, in addition, they pull back their tongue a little, they have arrived at the perfect pronun‐ ciation of the ash-sound. Another major issue is the long central vowel [ɜː], which you can find right above the schwa, [ǝ], in the vowel chart of the English language. It is the vowel of words like sir, versus or hurt. Again, the German learner of English who is not used to the motor routines necessary to produce the sound will fall back on a similar German sound, in this case the long close-mid front vowel [øː] the vowel sound of words like schön, böse or Höhe. When I was a student I actually had this problem. You might remember from the previous section that I had a problem with the word university, namely that I used [w] instead of [v]. On top of that, I also mispronounced the ‘e’, resulting in uniwöhsity instead of university. With your knowledge of the vowel chart of English and German, you know what to do. The incorrect German sound is a lot closer to the front and a little higher up. In addition, the lips are rounded when you pronounce it. A good way to help your pupils move from [øː] to [ɜː] would be to first tell them to open the mouth a little more while still keeping a generally rounded shape. At the same time they need to pull their tongue back a little which will lead to the desired vowel. As always when it comes to pronunciation exercises, don’t forget the power of the mirror neurons. It might be a good idea to make them watch you as you alternate between the two sounds. If you take a look at figure 3.4 above, you will see that the German language has two central vowels, the schwa [ǝ], which we also find in English and a central sound that is produced with a more open mouth, sometimes referred to as the ‘low schwa’ or the ‘open schwa’, [ɐ]. This sound is often used when we have an ‘r’ in writing, which is why it is sometimes called the ‘vocalised r’. One context is a word-final -er. Compare, for instance, the last sound in the following two German words: (auf die) Schnelle - (sie ist) schneller. The first word ends in the normal schwa, the second one in the open schwa. Another context in which the sound occurs is when a long vowel is followed by an ‘r’ in writing. Take a look at the following examples. 54 3 “What are you sinking about? ” - sounds in isolation <?page no="56"?> diphthongs (1) hört [høːɐt] fährt [fɛːɐt] führt [fyːɐt] friert [fʁiːɐt] Uhr [uːɐ] spart [spaːɐt] Many German learners of English use the open schwa in similar contexts when they speak English. You may recall my orthographic rendition of the German accent in the introductory example about the German coast guard. The word German, in that clip, is pronounced something like ‘Tschöamen’ or, in phonetic transcription, [tʃøːɐmǝn]. You see, the open schwa is a vowel that occurs quite frequently in the German language and, as a result, might cause problems for German learners of English. This becomes particularly relevant if we look at the so-called ‘centring diphthongs’. Diphthongs are vowels that change during articulation. For our purposes it suffices to understand them as vowels that consist of two vowels, the first gliding into the second. The English language has many more diphthongs than the German one, but we will only focus on those that are likely to cause problems and these are the three centring diphthongs that we find in Received Pronunciation. As you can see in figure 3.5 they start off with one of the front or back vowels of English and then move to the schwa, which is the most central vowel in the English vowel chart, hence the name. Fig. 3.5: The three centring diphthongs of RP and their incorrect German counterparts. With some German learners of English the orthographic ‘r’ in words like those in figure 3.5 might trigger the pronunciation of the open schwa which leads to a strong impression of a German accent. The words fear, fair, pure will sound like the German words vier, fair and pur. The problem, in my experience, is rather a problem of awareness, since the German language also contains the schwa. Consequently, working with contrastive pairs like the ones discussed would usually solve that issue. 3.2 Vowels 55 <?page no="57"?> This concludes our discussion of the sound inventories of English and German and we will continue with an analysis of phonetic contrasts in the next chapter. New tools in your linguistic toolkit Contrastive Analysis ▶ An awareness of differences between the English and the German sound systems gives you an idea which sounds are likely to be problematic and why. Articulatory phonetics ▶ You know how sounds are produced and, in particular, are aware of the three variables that are used to describe consonants and the three variables that are used to describe vowels. Because of this you are able to analyse your students’ pronunciation errors much more precisely and you can help them much better. Dental alveolar fricatives, [θ, ð] ▶ You know why these sounds are problematic and you understand why they are substituted by the sounds German learners typically substitute them by. You know how to get your students from the substituted sounds to the target sounds. Voiced alveolar approximant, [ɹ] ▶ Your awareness of the place of articulation of the German voiced alveolar trill tells you that the English sound you are aiming for is created at the same position as the German sound. Your knowledge about the difference between trills and approximants tells you that all it takes to get from the German to the English sound is to lower the tip of the tongue a little bit. Mirror neurons ▶ Your awareness of mirror neurons gives you and your students an excellent way of practicing pronunciation by just looking in the mirror while pronouncing two or more different sounds. Voiced labial-velar approximant, [w] ▶ You know that the sound in itself is not particularly difficult but it often gets confused in pronunciation with the voiced labiodental fricative. One way to solve this problem is to make use of the mirror neurons. Vowel chart ▶ You know how to read the vowel chart and, in particular, are aware that it represents tongue position for the individual vowels. You have made the experience that you can feel your tongue moving according to the vowel chart and you can make your students aware of the differences between German and English vowels by providing them with the same experience. The ash-sound, [æ] ▶ You know why German learners will use the short German ‘e’, [ɛ], instead and you know that by simply opening the mouth more they already will be a lot closer to the ash-sound. The sir-sound, [ɜː] ▶ You know why German learners will use the long German ‘ö’, [øː], instead and you know that opening the mouth and pulling the tongue back a little will usually take your students to the correct sound. Open schwa, [ɐ] ▶ You know that in addition to the schwa-sound [ǝ] of English, German also has an open schwa which is often triggered by the letter ‘r’ in writing. For RP this leads to a problem with those words that end in the letter ‘r’ and result in a centering diphthong in pronunciation, since in those cases German learners would usually substitute the schwa by 56 3 “What are you sinking about? ” - sounds in isolation <?page no="58"?> the open schwa. Your knowledge of the German vowel chart tells you that the simple solution is to close the mouth a little more to produce the schwa. If you want to know more B I E S W A N G E R , Markus & Annette B E C K E R ( 5 2021): Introduction to English Linguistics. Tübingen. ▶ Chapter 3 of this very accessible introduction to English linguistics provides a short introduction to phonetics and phonology. B R O W N , Adam (2014): Pronunciation and Phonetics. A Practical Guide for English Language Teachers. London: Routledge. ▶ Provides a thorough introduction into phonetics and phonology and into teaching pronun‐ ciation. Also provides sample exercises. K ÖN I G , Ekkehard & Volker G A S T ( 4 2018): Understanding English-German Contrasts. Berlin. ▶ A very detailed contrastive analysis of English and German. Chapters 2 and 3 provide a comprehensive comparison of English and German phonology that is much more detailed than the comparison in this book. S C H M I T T , Holger (2016): Teaching English Pronunciation. A Textbook for German-speaking Coun‐ tries. Heidelberg. ▶ A thorough introduction into English pronunciation and how to teach it. In addition to providing methods and tools the book also provides an in-depth discussion of specific pronunciation problems of German learners of English. S K A N D E R A , Paul & Peter B U R L E I G H ( 4 2022): A Manual of English Phonetics and Phonology. Tübingen. ▶ An accessible and concise introduction to phonetics and phonology with university students from German-speaking regions in mind. https: / / www.seeingspeech.ac.uk/ ipa-charts/ . ▶ An amazing website! ! ! It provides you with a clickable IPA chart which gives you MRI, animated and ultrasound videos of speakers pronouncing consonants and vowels. Additional online content. 3.2 Vowels 57 <?page no="60"?> 4 “France is bacon” - sounds in speech When I was young my father said to me: “Knowledge is power, Francis Bacon.” I understood it as “Knowledge is power, France is bacon.” (L A R D _B A R O N ) What this chapter is about While the previous chapter has focused on sounds in isolation, this chapter explores what happens when sounds are used in actual speech. English and German differ in a number of ways. For instance, the l-sound in the German words lief and fiel are pronounced the same, while the corresponding sounds in English leaf and feel are not. The final sounds in German Bad and bat sound the same, the English sounds in bad and bat do not. These and a range of other important differences of sounds in speech are the topic of this chapter. Together with the previous chapter, it provides you with the necessary expertise to improve the pronunciation of your students tremendously. In the previous chapter we have described sounds in isolation. We have looked at the way they are articulated, we have compared the sound systems of German and English and we have identified English sounds that are likely to be problematic for German learners of English since the German language does not have these sounds. However, sounds do not usually occur in isolation but they occur in the context of other sounds in speech. In some cases, English and German may share the same sound, so there is no problem with regard to the aspects discussed in the previous chapter. However, sometimes sounds may be realised in different ways if they are used in actual speech. To understand what is going on we first need to get a clearer idea of what exactly we mean when we use the term ‘sound’. 4.1 Phonemes, phones and allophones In chapter 3, we talked about sounds of a language without really thinking about the concept and what we mean, when we use the term ‘sound’. We didn’t need that much detail then, but a slightly closer look is necessary for what follows. <?page no="61"?> aspiration So, what do we mean when we talk about ‘sound’? What are the sounds of a language? Or, what are the different, the distinguishable sounds of a language? Which sounds go into that set? Presumably, we would be interested in all the different sounds that a language has. From a very strict acoustic perspective one might argue that no two sounds are ever fully identical. So, if you pronounce the letter ‘p’ ten times in a row, none of these would be exactly the same. Still, we would group them together under the sound ‘p’. What about the initial sound in the word pit and the second sound in the word spit? What would you say, are they the same sound or not? Most likely, you will say they are identical, maybe also supported by the fact that we use the same symbol to refer to the two sounds, namely the letter ‘p’. Before you read on … … put your fingers roughly 2 cm in front of your lips. Now pronounce the words pit and spit in alternation a couple of times. What do you recognise? Maybe you recognised that with the word pit the p-sound comes with an extra puff of air, which you do not have when you pronounce the second word. This extra puff of air is referred to as ‘aspiration’ and, accordingly, the first sound is called an ‘aspirated p’, whereas the second is ‘non-aspirated’. So, we are in fact dealing with two different sounds, which also becomes apparent when we look at their phonetic symbols in the more detailed narrow phonetic transcription: [p h ] and [p], the little superscript ‘h’ standing for aspiration. However, somehow this difference does not seem to be as rel‐ evant as the difference between the two sounds [b] and [p]. You may agree that these two consonants are actually two different sounds. However, the difference in that case only consists in the fact that the first sound is voiced whereas the second is voiceless. Still, exchanging one for the other changes the word in which they occur. The word bit is different from the word pit, even though the first sound has only changed a little. In contrast to that, aspiration does not lead to meaning changes. You can see this, if you pronounce the word spit with an aspirated ‘p’, i.e. something like s-pit. You will notice that it sounds a bit strange that way, but you would not think that it is a different word altogether. By the way, there are languages, such as Thai or Korean, where as‐ piration is in fact phonemic, i.e. it distinguishes between words. The question of when sounds are the same or different is not dissimilar to the question of when two letters are different. Take a look at the examples under (1). 1. pit - pit -pit - pit pit---Pit pit - p it - p it - p it pit - bit - sit - fit If you print example (1) and look very closely at the first row you will see that there are very small differences with each of the initial letters. These differences are due to 60 4 “France is bacon” - sounds in speech <?page no="62"?> phonemes minimal pairs phones differences in the structure of the paper and the way the ink or the laser interacted with the paper. So, strictly speaking, none of the four letters would be identical on paper, although they are all identical on a computer screen. The initial letters in the second row are obviously not identical but you will read them as two versions of the letter ‘p’, namely lower case and upper case ‘p’. Compare this to the third row. Like in the second row, the initial letters do not look the same (even on a computer screen) but you will read all of them as some version of the letter ‘p’ and the word is always pit. In the fourth row, in contrast, the initial letters are much more different and they actually do result in four different words. Keep this analogy in mind when, in the following, we discuss the distinction between ‘phoneme’, ‘allophone’, and ‘phone’. What the different letters are for written language (for example in the fourth row of example (1)) the phonemes are for spoken language. Phonemes are abstract ideas of sounds, just like letters are abstract concepts for the ‘things’ that make up the alphabet. We use forward slashes to make clear that we are talking about phonemes, e.g. the phoneme / p/ . Phonemes are identified with the help of a procedure that is based on whether an exchange of one sound for another makes a difference in meaning or not. A phoneme is, therefore, usually defined as ‘the smallest unit that distinguishes mean‐ ing’, or ‘the smallest distinctive unit’. What does that mean? Well, if you take the string of sounds / bɪt/ and replace the initial consonant with the sound / p/ , you get a different word with a different meaning, namely / pɪt/ (Note that we use slashes instead of an‐ gular brackets to refer to phonemes; see below). A pair of words like bit and pit is what we call a ‘minimal pair’, i.e. two words that are identical apart from one sound only. We use these minimal pairs to identify the phonemes of a language. Phonemes are abstract units of linguistic description, that is, no one has ever heard a phoneme. What we actually hear are the phones of a language (marked by angular brackets, [ ]). We say phonemes are instantiated as phones. So, if you pronounce the letter ‘p’ ten times in a row, you will have instantiated the phoneme / p/ ten times as a phone [p]. Just like the different letters ‘p’ in the first row of example (1), neither of these are, strictly speaking, fully identical. However, these differences are completely irrelevant. Before you read on … … answer the following question: When we identify the phonemes of a language we are interested in the function that these phonemes have in the language system. An analogy that I find to be useful is that of furniture in a room and the different functions that pieces of furniture have. Let’s say in a room you find a table with six chairs. Applying what we have said above, what would be the equivalent of phonemes in that room and what would be the equivalent of phones? When we look at a room with a table and six chairs we can easily distinguish two functionally different kinds of furniture, namely the table and the chairs. We use tables 4.1 Phonemes, phones and allophones 61 <?page no="63"?> allophones complemen‐ tary distri‐ bution free variation to put things (e.g. food, plates and cutlery) on, and we use chairs to sit on. These would be the equivalents to phonemes, or to be more precise, the abstract ideas ‘table’ and ‘chair’ (together with their respective functions) are the equivalents to phonemes. The abstract idea ‘table’ is realised in the one table in that room. This table, the table we can look at and the table we can touch is the equivalent of the ‘phone’ to the ‘table phoneme’. The six chairs are six individual realisations of the abstract concept ‘chair’. They would be the ‘phones’ to the ‘chair phoneme’. Coming back to example (1). What about the second row, lowercase and uppercase ‘p’, and the third row, the different fonts p, p , p , and p ? Do we have something anal‐ ogous in spoken language, too? Yes, we have. These different versions of the letter ‘p’ are similar to what linguists call ‘allophones’. With some phonemes we can see par‐ ticular regularities when it comes to their realisation. You have seen in the example above the distinction between the p-sound in the words pit and spit. In the first case, the sound is aspirated, in the second it isn’t. As linguists we can predict when exactly this aspiration occurs and when it doesn’t. If this is the case, we speak of allophones. The prefix ‘allo-’ is of Greek origin and means ‘other’ or ‘different’. So, a phoneme is realised not only as one phone, but also as one or more other phones. A good German example is the phoneme / d/ , which is usually realised as the phone [d] but, for example at the end of a syllable, would usually be realised as [t] as a result of final devoicing in German. The two examples discussed above are instances of allophones in comple‐ mentary distribution. This means that one allophone occurs in one set of contexts (the position of the allophone and its neighbouring sounds) whereas the other allophone occurs in another set of contexts. The phoneme / d/ , for instance, is realised as the allophone [t] at the end of a syllable and as the allophone [d] in all other cases. This way the two allophones complement each other. Allophones in complementary distri‐ bution would be similar to uppercase and lowercase ‘p’, because we can cleary state when the two appear. A simple rule for the English language would be “Use ‘P’ at the beginning of a sentence or at the beginning of a proper name. In all other contexts use ‘p’.” We could also argue that bold or italics are examples of complementary distribu‐ tion. For instance, I use bold font to emphasise a stretch of text and normal font if I do not want to put particular emphasis on something. Italics is used to mark what is called ‘object language’, i.e. a word etc. that I talk about. In addition, we also have allophones in free variation. This means that you can choose either allophone and it doesn’t make any difference. An example of this is the phoneme / t/ , which has as its two realisations the allophones [t] and [ʔ]. The second is called the ‘glottal stop’ and it is a sound that can replace the first allophone in some accents of English. In the London area, for instance, you could pronounce the word butter with either sound, nobody would have problems understanding the word. This would be similar to the different fonts in line 3 of example (1). Admittedly, this was a somewhat lengthy discussion of some technical terms but it is important because we will be looking at allophonic processes in the following. These are important since there may be sounds in the sense of ‘phoneme’ that both the English and German language 62 4 “France is bacon” - sounds in speech <?page no="64"?> final devoicing have, however, one language may show a wider variety of the possible realisations of that phoneme than another language. In the following we will discuss three of these differences. 4.2 When Alastor Moody has a mat eye - allophonic contrasts When it comes to allophonic contrast, one of the major problems for German learners of English is a phenomenon called ‘final devoicing’. The term refers to the fact that in the German language voiced obstruents, i.e. sounds that obstruct the flow or air to some extent during production (/ b, d, g, v, z, ʒ, dʒ/ ), are devoiced when they occur in syllable-final position, as exemplified in the German part of table 4.1. The first two columns make clear that the sound in question is actually a voiced obstruent, which gets devoiced in syllable-final position (columns 3 and 4). German English Räder Stäbe Flüge Lose Brave Orange Manager [ʁεːdɐ] [ʃtεːbə] [flyːgə] [loːzə] [bʁaːvə] [oraŋʒə] [mɛnɛdʒɐ] Rad Stab Flug Lose brav orange manage [ʁaːt] [ʃtaːp] [fluːk] [loːs] [bʁaːf] [oraŋʃ] [mɛnɛtʃ] mad stab hug buzz brave rouge badge [mæd] [stæb] [hʌg] [bʌz] [bɹeɪv] [ɹuːʒ] [bædʒ] Table 4.1: Final devoicing in German and its absence in English. If we phrase this situation in terms of phonemes and allophones we can say, for instance, that the phoneme / d/ has two allophones, namely [t] in syllable-final position, and [d] in all other contexts. The phoneme / b/ has two allophones, namely [p] in syllable-final position, and [b] in all other contexts. And so on. So, final devoicing is a regular process in German and it is final devoicing that makes the joke “Feldsalat? - Wenn Du ihn runterschmeißt! ” work, but it does not happen in English. The same set of sounds remains voiced in syllable-final position, as the righthand side of table 4.1 makes clear. Keeping final devoicing under control is easier said than done for German learners of English, since it is a regular phonetic process in the German language. It might help to make clear to your students that substituting voiceless sounds for their voiced counterparts in syllable-final position might often result in a change of meaning. For example being mad is different from being a mat, a lab is not a lap, a bag is different from a back, a buzz is not a bus, to save someone is different from being safe, and a badge is not the same as a batch. Unfortunately, I cannot find a minimal pair for [ʒ] and [ʃ], but the point I want to make here should be clear nonetheless. 4.2 When Alastor Moody has a mat eye - allophonic contrasts 63 <?page no="65"?> [ŋ] or [ŋg] Before you read on … … try this other way of tackling the problem of final devoicing. The Harry-Potter character of Alastor “Mad Eye” Moody runs the risk of becoming “Mat Eye” if Germans pronounce the name. This risk disappears if you make the last sound of the first word the first sound of the second word. “Mad Eye”, thus, is not pronounced [mæd aɪ] but [mæ daɪ]. Try it! This is called ‘resyllabification’ and will be discussed in detail in section 4.3.3. The second problem we would like to take a closer look at concerns the phoneme / ŋ/ as in sing or song. Let us take a look at the dataset provided in table 4.2. German English lang Ding Finger Hunger Sänger Dünger länger jünger [laŋ] [dɪŋ] [fɪŋɐ] [hʊŋɐ] [sɛŋɐ] [dyŋɐ] [lɛŋɐ] [jyŋɐ] long thing finger hunger singer banger longer younger [lɒŋ] [θɪŋ] [fɪŋgə] [hʌŋgə] [sɪŋə] [bæŋə] [lɒŋgə] [jʌŋgə] Table 4.2: [ŋ] in German and [ŋ] or [ŋg] in English. In the German language the letters ng always seem to be pronounced [ŋ] while in English there appears to be some random alternation between [ŋ] and [ŋg] (apart, of course, from words like Orange in German or orange, challenger and exchange in English, where the 'g' is pronounced as [ʒ] and [dʒ], respectively). What is going on here? Looking at the first four English words we might hypothesise that if ng occurs at the end of the word it is pronounced [ŋ] (just like in German) and that it is pronounced [ŋg] within a word. However, this hypothesis has to be refuted if we take a look at the words singer and banger. Here, again, the letters ng are pronounced [ŋ]. What is it that distinguishes, say, finger from singer? Apart from the first letter they are identical and the letter sequence in question occurs within the word in both cases. Before you read on … … see how far you will get with this (a little tougher) exercise. If you look at the list of English words in table 4.2, can you come up with a hypothesis about when ng is pronounced [ŋ] and when it is pronounced [ŋg]? You need to take morphemes (the smallest units of meaning) into consideration. For instance, the word singer consists of two morphemes, sing and -er. 64 4 “France is bacon” - sounds in speech <?page no="66"?> allophones of / l/ in English The words finger and singer differ if we move from the level of the word to the level of morpheme, i.e. the smallest unit of meaning. Finger, just like hunger, consists of one morpheme only, we cannot break it down into smaller meaningful parts. In contrast to that, singer and banger both consist of two morphemes namely the verb sing and bang, respectively, and the morpheme -er, which expresses the meaning ‘someone who does X’. A singer is someone who sings etc. So, we adapt our hypothesis: the letter combination ng is pronounced [ŋ] at the end of morphemes (sing, singer) and [ŋg] within morphemes (finger) (with the exception, of course, of words like orange or challenge, where the 'g' is pronounced as [ʒ]). However, the words longer and younger call this hypothesis into question. In both cases the letter combination ng is at the end of a morpheme and, still, it is pronounced [ŋg]. The same, by the way, holds true for the respective superlative forms longest and youngest. The comparative and superlative forms are actually exceptions to our hypothesis. Fortunately, they are the only ones. So, we can describe the rules for whether to use [ŋ] or [ŋg] as follows: ● within a morpheme ng is always pronounced [ŋg], e.g. finger, hunger ● at the end of a morpheme ng is pronounced [ŋ], e.g. sing, song, singer, singing, unless ● it occurs before a comparative or superlative morpheme, in which case ng is pronounced [ŋg], e.g. longer, stronger While this is the linguistically correct description of the situation, you would maybe not teach it that way to your students, as they would usually not have any idea what a morpheme is. Nevertheless, the rule above is important when it comes to pronunciation, as it concerns many words. Fortunately, the problem of explanation is reduced quite substantially if we start off from the situation in the German language (the letter combination ng is always pronounced [ŋ]) and describe exceptions from that rule. A possible description could be (after having ruled out those words where ng is realised as [ʒ]): Just like in German the letter combination ng is pronounced [ŋ]. However, if a word ends in -nger or -ngest the letters ng are pronounced as [ŋg]. There is one tiny exception, though: if the word refers to someone who does something, then it is pronounced [ŋ]. The last allophonic contrast concerns the phoneme / l/ . Just like above, in this case, too, the German language only has one realisation of this phoneme, namely the voiced alveolar lateral approximant [l]. This sound is often called the ‘clear l’. In the English language it is contrasted with the so-called ‘dark l’, which phonetically is described as the voiced velarised lateral approximant. It is called ‘velarised’ since the back of the tongue is moved upwards towards the velum, as you can see in figure 4.1. This gives the l-sound a somewhat darker quality. 4.2 When Alastor Moody has a mat eye - allophonic contrasts 65 <?page no="67"?> Fig 4.1: The English allophones of / l/ : clear l (left) and dark l (right). The velarized ‘l’ is quite a frequent sound in the English language, as it occurs at the end of a syllable or word and in front of consonants. For you as teachers of English it might be useful, first, to make your students aware that there actually is a difference between German and English. Word pairs like viel and feel or Feld and felt should make the contrast quite obvious. Once your students have recognised that there is a difference maybe some assistance is needed in articulating the dark ‘l’. In my experience, a useful approach is to start with the clear ‘l’. While your students are pronouncing that sound they should try to produce a vowel that is similar to the ‘o’ in Gott only long, while keeping the tip of the tongue at the alveolar ridge. This way, they will come very close to the dark ‘l’. By the way: Broad and narrow transcription Now that we have discussed some of the major allophonic contrasts, we can take a closer look at the transcription of sounds. In dictionaries (printed or online) you would usually find what is called ‘phonemic transcription’, i.e. a transcription that focuses on the smallest distinctive units and dispenses with the smaller phonetic details. Hence, it is also referred to as ‘broad transcription’. For instance, above we have seen that the phoneme / l/ is realised in two allophones, clear and dark ‘l’. The distinction between the two does not show in broad transcription. It only becomes apparent in so-called ‘phonetic’ or ‘narrow transcription’, where the velarisation of the [l] is represented by a small extra symbol, ‘ ̴ ̴ ‘. So, the phoneme / l/ is realised as clear ‘l’, [l], and dark or velarised ‘l’, [ɫ]. Another example where broad and narrow transcription differ is the case of the phoneme / r/ . In dictionaries the word try would be transcribed as / traɪ/ both, for British and American English, even though, as we have seen in section 3.2.1, in RP the / r/ is realised as an alveolar approximant, [ɹ], while in GenAm it is realised as a retroflex approximant, [ɻ]. 66 4 “France is bacon” - sounds in speech <?page no="68"?> 4.3 Why France is bacon - connected speech If we take a look at connected speech, we quickly realise that there is a lot going on during articulation: adapting the state of the glottis to produce voiced and unvoiced sounds, the exact positioning of the tongue in your oral cavity and putting the other articulators in the right position or the right shape. You can imagine that all of this takes quite some effort and, consequently, it does not come as a surprise that speakers are, in a sense, looking for ways to reduce that effort. The heading of this chapter points at one way to reduce effort. It is motivated by the following story that can be found all over the internet (e.g. knowyourmeme.com/ memes/ france-is-bacon). When I was young my father said to me: “Knowledge is power, Francis Bacon.” I understood it as “Knowledge is power, France is bacon.” For more than a decade I wondered over the meaning of the second part and what was the surreal linkage between the two. If I said the quote to someone, “Knowledge is power, France is Bacon,” they nodded knowingly. Or someone might say, “Knowledge is power” and I’d finish the quote “France is bacon,” and they would look at me like I’d said something very odd, but thoughtfully agreed. I did ask a teacher what did “Knowledge is power, France is bacon” mean and got a full 10-minute explanation of the “knowledge is power” bit but nothing on “France is bacon.” When I prompted further explanation by saying “France is bacon? ” in a questioning tone, I just got a “yes.” At 12 I didn’t have the confidence to press it further. I just accepted it as something I’d never understand. It wasn’t until years later I saw it written down that the penny dropped. (Answer by Lard_Baron to a question on Reddit in 2011: “What word or phrase did you totally misunderstand as a child? ”) The name Francis Bacon can only be misunderstood as France is bacon, since the English language tries to link words as much as possible, which makes France is sound like Francis. Linking comes in many different forms, all of which will be addressed further below, and it is important for German learners of English because the German language is a language that does not link its words in spoken language. Ein Extrem ist, for example, would not be confused with Ein Extremist. On the contrary, the German language usually makes sure that new words beginning in a vowel are actually separated from the previous word by inserting a glottal stop. For example, if you read out example (2a) below, you will feel that you actually put effort into separating the individual words. The English way of pronouncing the same sentence would be much closer to what you can see under (2b). (2a) Kauf ein Ei und ein Eis im Edeka. (2b) Kau fei nEi un dei nEi si mEdeka Admittedly, it looks a bit strange if you see it written down like that, but, if you listen closely to the speech of native speakers of English, you will hear that kind of linking. 4.3 Why France is bacon - connected speech 67 <?page no="69"?> regressive assimilation As we will see below, for example, for English-speaking people there is not really a difference between nude art and new dart. The English language shows another feature of connected speech that seems to be motivated by questions of efficiency, namely the extensive use of weak forms as we see, for instance, in the term rock ‘n’ roll. The existence of such reduced forms should not come as a surprise to German learners of English as we also have weak forms in the German language. However, the use of weak forms in German seems to be somewhat frowned upon in certain contexts and seems to be more acceptable when talking among friends or in other informal situations. For the English language, in contrast, weak forms are the norm in any kind of context. To put it differently, if you always used the full form of every word, i.e. the form you would use if you spoke the word in isolation, you would sound quite odd. More of this later. Sometimes the pronunciation of one word influences the pronunciation of a neigh‐ bouring word. This process is called ‘assimilation’. Before you read on … … you may try this for yourself: take the phrase that man and repeat it quickly ten or more times. What happens? Usually, with your last repetitions the phrase will start to sound like thap man rather than that man. How can we explain that? Having worked through the previous section you’re in the possession of every bit of information that you need to understand why this would happen. So, give it a try and see if you can find an explanation (otherwise wait for the explanation in the next section). This chapter will take a look at all of the three features of connected speech, starting off with different kinds of assimilation. 4.3.1 Assimilation Let’s get back to thap man and that man. What is happening here? Originally, we have the two sounds [t] and [m]. The first, as you know by now, is a voiceless alveolar plosive. The second is a voiced bilabial nasal. If you repeat the phrase that man over and over again, the [t] will change to [p]. This sound is a voiceless bilabial plosive. It is identical to the sound [t] apart from the place of articulation. This has shifted from alveolar to bilabial. Why would the speaker do that? The most reasonable hypothesis is that he or she does that in anticipation of the bilabial sound [m] that follows in the next word. This whole process is called ‘assimilation’ which generally means that two things be‐ come more similar. In this case it is two sounds that become more similar, more spe‐ cifically, one sound, [m], influences the pronunciation of the preceding sound [t]. The direction of influence is backwards, which is why we talk about ‘regressive assimila‐ tion’. Another instance of regressive assimilation can be found in the phrase have to. 68 4 “France is bacon” - sounds in speech <?page no="70"?> elision progressive assimilation assimilation and nega‐ tion prefixes If you speak the word have on its own, it is pronounced as [hæv], i.e. the final consonant is voiced (even though German learners of English might not be aware of it because of final devoicing). Again, in anticipation of the following sound, which is a voiceless alveolar plosive, the [v] will get voiceless as well, resulting in [f] and the pronunciation of the whole phrase will move from [hævtuː] to [hæftuː] (we’ll ignore the weak form of to for the moment). As a final example consider the word handbag. Two things can happen here. The first is a process that we call elision, i.e. the omission of a sound. So, handbag would often be pronounced without the [d], i.e. [hænbæg]. As before, while pronouncing the last sound of the first syllable, the voiced alveolar nasal [n], the speaker anticipates the place of articulation of the first sound of the next syllable. The sound [b] is bilabial and, hence, handbag might be realised as hambag [hæmbæg], be‐ cause [m] is bilabial, too. In addition to regressive assimilation we also find ‘progressive assimilation’ in Eng‐ lish, even though it is more rare. One possible context is the one instantiated by phrases like read this or find this. The final [d] of the first word might influence the pronunci‐ ation of the voiced apicodental fricative in this, resulting in [dɪs] instead of [ðɪs]. An‐ other context is words ending in the postalveolar fricatives [ʃ] and the next word be‐ ginning in the respective alveolar fricative [s]: Church Street, for instance, might be pronounced as [tʃɜːtʃ ʃtriːt] rather than [tʃɜːtʃ striːt]. In the case of would you the final [d] would change the pronunciation of the initial [j] to [ʒ], resulting in [wʊd ʒuː] instead of [wʊd juː] (again, ignoring the weak form of you for the moment). When it comes to language teaching, assimilation, in my view, is nothing that needs to be taught explicitly since it’s a natural process that will happen anyway - if we let it happen, that is. Maybe this is a point worth mentioning in the classroom. Not only is it okay to not pronounce words in connected speech as ‘carefully’ as if we pronounced them in isolation, it is what English native speakers would expect you to do. Accordingly, you and your pupils will sound more authentic if you and them allow this natural process to happen. An awareness of assimilation can also help to explain phenomena of the English language that might be puzzling for your students. I would like to give you two exam‐ ples. In the context of errors in chapter 2 we have talked about negation and a variety of negation prefixes in the English language. Before you read on… … take a look at the pronunciation of the first sound after the negation prefixes beginning in ‘i’ below. If you assume that inis the basic negation prefix, can you explain what is going on with the help of regressive assimilation? insane immoral illogical irresistible inactive impossible illegal irregular incredible imbalance illiberal irrelevant 4.3 Why France is bacon - connected speech 69 <?page no="71"?> If we assume that the prefix in the first column is the basic negation prefix, the three other versions can all be explained through regressive assimilation. The prefix inbecomes imin front of bilabial consonants (second column), due to the anticipation of the place of articulation: the last sound of the prefix changes from voiced alveolar nasal to voiced bilabial nasal. The third column can be explained by an anticipation of the manner of articulation. Both, [n] and [l] are voiced alveolar sounds, the only difference being that the first is a nasal, the second is a lateral approximant. We find a similar change of the place of articulation in the words of the fourth column with the [ɹ] of irbeing an approximant. Obviously, you’re not likely to teach assimilation as a topic in school. If that was what gave linguistics its legitimacy as part of your education as a future teacher, there would be no point in telling you about this phenomenon, or about most of the other linguistic phenomena for that matter. However, you do see in the example above how seemingly random restrictions in the English language can be explained with the help of linguistics. It stands to reason that providing such an explanation is a much better answer to the question “but why is this so complicated? ” than a mere “That’s just the way it is. You just have to learn it! ” Progressive assimilation can be helpful in a similar way. Some years ago one of my nieces was confused about the formation of the regular plural in English: I don’t understand this plural thing in English. I know that you use [ɪz] if the noun ends in some kind of ‘s’ or ‘sch’. But I never know when to use the soft ‘s’ or the ‘sharp’ ‘s’. Being a linguist you know that the regular plural comes in three forms, as shown under (3) (recall from your introduction that the sibilants of English are [s, z, ʃ, ʒ, tʃ, dʒ]). (3) [-ɪz] when following a sibilant, e.g. houses, kisses, churches [-s] when following a voiceless sound other than a sibilant, e.g. hats, packs, hips [-z] when following a voiced sound other than a sibilant, e.g. hugs, dads, tubes (note that the ‘e’ is silent). Admittedly, your linguistic expertise will not be particularly helpful on its own, since your students will not necessarily know what voiced or voiceless sounds, let alone sibilants are. However, you can boost your linguistic expertise with a bit of creativity and explain the occurrence of voiced and voiceless ‘s’ in a way that even pupils from primary school will understand. You can exploit the fact that voicing is easily perceptible at the Adam’s apple when you touch your throat with your fingers. Even children at a very early age can feel this difference. So, first I made sure that my niece could feel the difference between the voiced and the voiceless alveolar fricative when she touched her Adam’s apple. Once this was established it was easy: Just say the word in the singular form. If you can feel your Adam’s apple vibrating with the last sound of this word use the soft ‘s’. If you don’t feel any vibration, use the ‘sharp’ ‘s’. This was a real eye-opener for my niece. 70 4 “France is bacon” - sounds in speech <?page no="72"?> citation form weak form reduction of full vowels elision weak forms and effi‐ ciency 4.3.2 Weak forms All English words have a so-called ‘citation form’. This is the form that is used if you pronounce a word in isolation. For example if you said the sentence the words and and of are function words, the first and would be pronounced in its full form, which is [ænd], whereas the second and would usually be reduced to some extent. This could be [ǝnd] or [ǝn] or just [n], all of which are weak forms of the word and. Sometimes we can even see that in writing as in the name of the popular music style rock ‘n’ roll. The example of and already shows the two processes that are at work when it comes to creating a weak form from a strong form. This can be done through a reduction of the vowel. One way is to change the vowel to schwa, as shown in the case of and where we change the [æ] to [ǝ]. Another kind of reduction is to shorten long vowels into their short counterpart. For example the long vowel [iː] in she might be changed to [i] or [ɪ]. In all cases, the vowels have become weaker in the sense that it takes less effort to articulate them. Finally, some sounds can be left out as we see in the example of [ænd], which can be reduced to [ǝn] or even just [n]. The process of leaving out sounds is called ‘elision’ and we say that a sound has been elided. Similar to assimilation, weak forms can be understood as being largely driven by questions of efficiency. Long vowels take more effort in production than short vowels. This is not just because we exhale longer with long vowels, but as you can see in figure 3.3 the long vowels that may be shortened ([iː] and [uː]) are produced either far to the front or far to the back of the oral cavity and with the highest point of the tongue close to the roof of the mouth. This is a lot of effort for the tongue. The same explanation holds for the replacement of a full vowel by schwa. The schwa-sound is the most relaxed vowel a language can have, as the position of the tongue is maximally relaxed with this sound. This degree of relaxation means a huge reduction of effort. Table 4.3 provides a list of some of the most important weak forms of the English language. Strong Weak Strong Weak a am an and are be been but can could do does for from [eɪ] [æm] [æn] [ænd] [aː(ɻ)] [biː] [biːn] [bʌt] [kæn] [kʊd] [duː] [dʌz] [fɔː(ɻ)] [fɹɒm] [ǝ] [(ǝ)m] [(ǝ)n] [(ǝ)n(d)] [ǝɹ, ǝɻ ǝ, ɹ, ɻ] [bɪ] [bɪn, bin] [bǝt] [kǝn] [kǝd] [dǝ, dʊ, du, d] [dǝz] [fǝ] [fɹǝm] he her him I me must of she should the they them to us [hiː] [hɜː(ɻ)] [hɪm] [aɪ] [miː] [mʌst] [ɒv] [ʃiː] [ʃʊd] [ðiː] [ðeɪ] [ðɛm] [tuː] [ʌs] [hɪ, ɪ, hi, iː] [hǝ, ɜː, ǝ] [ɪm] [ʌ] [mɪ] [mǝst, mǝs] [ǝv, ǝ, v] [ʃɪ, ʃi] [ʃǝd] [ðǝ, ðɪ, ði] [ðǝ] [ðǝm, ǝm] [tǝ] [ǝs, s] 4.3 Why France is bacon - connected speech 71 <?page no="73"?> Strong Weak Strong Weak had has have [hæd] [hæz] [hæv] [hǝd, ǝd, d] [hǝz, ǝz, z, s] [hǝv, ǝv, v] were will you [wɜː(ɻ)] [wɪl] [juː] [wǝ] [wǝl, ǝl, l] [jǝ, jʊ, ju] Table 4.3: Some weak forms of English. Making use of weak forms might take some getting used to. However, it is a very natural thing to do when speaking English, so much so that if we have a sentence that does not have any reduced vowels, this sounds really strange. Before you read on … … try the two sentences below (Bolinger 1968: 26)! (a) Minnie Abbott’s kitten Missie’s fond of liverwurst. (b) Irene Carstair’s pet chimpanzee Nimrod dotes on fresh horehound drops. Reading out aloud the first example should feel quite natural to you, whereas the second example might feel strange. If we compare the two, we see that the first example shows a regular alternation of what is called full vowels and reduced vowels. The second example, in contrast, a perfectly correct English sentence, sounds strange to us, because all of the vowels are full vowels. 4.3.3 Linking We have already seen a beautiful example of linking in Francis Bacon and France is bacon. The following example is taken from the excellent TV series How I Met Your Mother. Barney has discovered one of Lily’s paintings, depicting her naked boyfriend Marshall, behind the piano. Below is how Barney, Robin and Ted tell Marshall about it. Because English links words as much as possible, the string nude art would sound exactly like new dart. (4) B: Hey guys, guess what I got! A new dart! R: Oh wow! A new dart! T: Hey, that new dart is great! R: I did not know you were such a fan of new dart, Barney . B: Oh yes, Robin. I just love new dart - nude … art. Linking of words is usually not found in German, but a very important feature of spoken English, which even highly advanced learners of English (and teachers) do not seem to be aware of. This is why you as a future teacher should put some effort into helping your students master this feature of English. Linking can occur between two words 72 4 “France is bacon” - sounds in speech <?page no="74"?> linking and efficiency resyllabifi‐ cation resyllabifi‐ cation and final devoic‐ ing whenever the second word begins in a vowel. Again, linking seems to be largely driven by questions of efficiency, because if you want to have a clear boundary in connected speech between two words you will always have to insert the glottal stop which takes some effort (and which the German language actually does). Before you read on … … repeat each of the following three phrases in quick succession ten times or more. bin ich sie auch tu ich Now, if you don’t force yourself to isolate each word and if you speak quickly you may find that at some point you end up saying bi nich, sie jauch and tu [w]ich (the [w] indicating the sound in English what) and you may experience that this takes less effort. That is what linking is about. The first type of linking we will discuss is that which occurs when the first word ends in a consonant and the second begins in a vowel. This type is usually referred to as ‘resyllabification’. Let’s consider the example provided at the beginning of this sec‐ tion. Nude art would usually be pronounced as new dart. If you look at the phonetic transcription of the two words, you can easily see why this technical term ‘re-syllabi‐ fication’ makes sense. (5) nude art new dart [njuːd ɑːt] [njuː dɑːt] It becomes obvious if we look at example (5) that the pronunciation of the two phrases is identical as far as the sounds are concerned. What has changed, however, is how we combine the sounds to form syllables. The term ‘syllable’ does not need to concern us here, as your general understanding of what a syllable is is enough at this point. When we pronounce nude art as new dart (as we would usually do) the last sound of the first syllable becomes the first sound of the second syllable. So, in this type of linking we change the original syllables, hence the term ‘resyllabification’. This is a normal process in the English language and you should use it as much as possible and you should teach your students to do it as well. Not only is this type of linking typical of the English language, it can also be really useful for us as German learners of English since we are likely to apply final devoicing when we speak English. For instance, if we pronounced the phrase nude art without resyllabification we would have to put a lot of effort into keeping the final consonant of the first word voiced, i.e. not say [njuːt] but [njuːd]. If we use resyllabification, this is no longer a problem as the voiced consonant will have moved to the beginning of the next syllable. In this position, German learners of English do not have any problems keeping the consonant voiced. 4.3 Why France is bacon - connected speech 73 <?page no="75"?> liaison linking with approxi mants linking r In addition, resyllabification can be helpful if we combine it with elision in weak forms of pronouns. For example, the phrases hug her and hug him are quite difficult to pronounce if we use the citation forms of each word, i.e. [hʌg hɜː] and [hʌg hɪm], respectively. Your life gets a lot easier if you elide the [h] of the pronouns, which then gives you the option to use resyllabification, [hʌ gɜː] and [hʌ gɪm]. Try it out for yourself: pronounce the phrases with their citation forms a couple of times and then repeat the same with the resyllabified versions - it should be a huge difference. The second type of linking includes the insertion of a sound between two words. This type of linking is referred to as ‘liaison’. Liaison can happen when the first word ends in a vowel and the second word begins in a vowel. To keep the words separate in this case would usually mean to insert the glottal stop, as we do in German. We have already seen that this takes a lot of effort. What we often find in English is that speakers insert a special kind of consonant in this position, namely an approximant. Please recall from above, that an approximant is a consonant which leaves the air a lot of room to pass through. In a sense, then, it is somehow in between a ‘real’ consonant with a more substantive obstruction of the airstream and a vowel where we have no obstruction in the airstream. Two of these approximants, [w] and [j], are actually also referred to as ‘semi-vowels’. Let us take a look at [w] first. The insertion of this sound is an option if the first word ends in the long or the short ‘u’, [uː] or [ʊ], and the next begins in a vowel. Do it would, accordingly, be pronounced something like [duːwɪt] or, using a weak form of do, [duwɪt] rather than [duː ɪt]. Show us is not pronounced [ʃǝʊ ʌs] but [ʃǝʊwʌs]. Who is it, if we also apply resyllabification, would be pronounced as [huːwɪzɪt]. To distinguish linking sounds from ‘regular’ ones, some transcription practices would use superscript symbols, e.g. [du w ɪt] or [ʃǝʊ w ʌs]. Again, [w] after ‘u’-sounds instead of the other approximants [j] and [ɹ]/ [ɻ] makes sense if we take into consideration that the two ‘u’-sounds and the bilabial approximant [w] show lip-rounding, whereas the other approximants do not. The approximant [j], to some extent can be seen as the unrounded counterpart to [w]. So, while the rounded close back vowels ([uː] or [ʊ]) can lead to an insertion of [w] before another vowel, the unrounded close front vowels ([iː] or [ɪ]) may lead to an additional [j]. The phrase see it may, thus, be pronounced [siːjɪt]/ [siː j ɪt]rather than [siː ɪt] and my arm would be pronounced [maɪjaːm]/ [maɪ j aːm] rather than [maɪ aːm], etc. While the insertion of the two semi-vowels are rather subtle and, accordingly, teach‐ ers are often not aware of it, the insertion of the voiced alveolar or retroflex approx‐ imants, i.e. [ɹ] or [ɻ] are more conspicuous and come under the term ‘linking r’ or ‘intrusive r’. The first can only be found in so-called ‘non-rhotic’ accents of the English language. The term comes from the Greek word for ‘r’ and refers to those accents that do not pronounce a letter ‘r’ in writing if it immediately follows a vowel. RP is such an accent, as you can see in words like for, more or further, where none of the ‘r’-s that you see in writing are pronounced. Compare this to GenAm, a rhotic accent of English, which does pronounce these ‘r’-s. If a word ends in such a post-vocalic ‘r’ and the next 74 4 “France is bacon” - sounds in speech <?page no="76"?> intrusive r begins in a vowel, the ‘r’ that we see in writing will also be pronounced in non-rhotic accents and become the initial sound of the next syllable. For example, for example, will be pronounced [fɔː ɹɪgzɑːmpl] and not [fɔː ɪgzɑːmpl]. More of it will be pronounced [mɔː ɹɒ vɪt], and further up will be [fɜːðə ɹʌp]. The second kind of ‘r’-insertion is called ‘intrusive r’, and describes the insertion of an r-sound in those cases where we do not have an ‘r’ in writing and insert it between two vowels at the word boundary, nonetheless. The r-sound, in a way, intrudes, hence, the term ‘intrusive’. Note that this only happens if the last sound in the first word is not one of the close vowels, i.e. the iand the u-sounds (if that was the case we would use one of the other approximants as we have seen above). An example frequently used to illustrate intrusive ‘r’ is the phrase law and order, which, using weak forms as well as resyllabification, would be pronounced [lɔː ɹǝn dɔːdǝ]. An example that shows both, linking and intrusive ‘r’ is the phrase India or Asia, which is pronounced [ɪndiǝ ɹǝ ɹeɪʒə]. To conclude this chapter, table 4.4 provides a couple of examples in orthographic and citation form and what they might sound like in connected speech, including assimi‐ lation, weak forms and linking. I have included in the last column a version without any spaces. This is to make clear that the examples should be read as one word, with the main stress on the syllable following the apostrophe - otherwise it will not sound authentic at all. Please note, that different versions might be possible. Orthographic from Citation form Connected speech Pronounced like Does he have to? Do I know it? She saw an ox. I used to love her. That man knows them. Could you watch him? She is for it. They are angry. [dʌz hiː hæv tuː] [duː aɪ nǝʊ ɪt ] [ʃiː sɔː æn ɒks] [aɪ juːzd tuː lʌv hɜː] [ðæt mæn nǝʊz ðɛm] [kʊd juː wɒtʃ hɪm] [ʃiː ɪz fɔː ɪt] [ðeɪ aː æŋgɹi] [dǝ zi hæf tǝ] [dʊ w aɪ nǝʊ w ɪt] [ʃi sɔː ɹǝ nɒks] [aɪ juːs tǝ lʌ vǝ] [ðæp mæn nǝʊ zǝm] [kǝ dʒʊ wɒ tʃɪm] [ʃi j ɪz fɔː ɹɪt] [ðeɪ j ǝ ɹæŋgɹi] [dǝzi’hæftǝ] [dʊ w aɪ’nǝʊ w ɪt] [ʃisɔːɹǝ’nɒks] [aɪjuːstǝ’lʌvǝ] [ðæpmæn’nǝʊzǝm] [kǝdʒʊ’wɒtʃɪm] [ʃi j ɪz’fɔːɹɪt] [ðeɪ j ǝ’ɹæŋgɹi] Table 4.4: Examples of connected speech, including assimiliation, weak forms and linking. New tools in your toolkit Final devoicing ▶ Your awareness of this regular allophonic process of German draws your attention to a highly significant problem area for German learners of English. Assimilation ▶ Your knowledge of this important process of connected speech makes you aware that it is not necessary, in fact, not wanted, to always pronounce every word in its citation form. 4.3 Why France is bacon - connected speech 75 <?page no="77"?> Weak forms ▶ Your knowledge of this important feature of connected speech in English makes you aware that it is not necessary, in fact, not wanted, to always pronounce every word very carefully. Glottal stop in German ▶ Your awareness of this central feature of German connected, or rather disconnected, speech helps you understand why it often sounds strangely choppy when Germans speak English. Linking ▶ You are aware how important linking is when speaking English. In particular, you know that, in contrast to German, there is usually no glottal stop before a word that starts in a vowel. Resyllabification ▶ You know that in those cases where the first word ends in a consonant and the second begins in a vowel, you can pronounce the final consonant as the first sound of the second word to make pronunciation easier. This can also help to deal with many cases of final devoicing. Liaison ▶ You are aware that you may insert certain sounds if one word ends in a vowel and the next begins in a vowel. This makes pronunciation easier. The approximant [w] ▶ This sound can be inserted if the first word ends in long or short ‘u’ and the second begins in a vowel; the counterpart to [j]. The approximant [j] ▶ This sound can be inserted if the first word ends in long or short ‘i’ and the second begins in a vowel; the counterpart to [w]. Intrusive ‘r’ ▶ You can insert an ‘r’ if one word ends in a vowel (other than long and short ‘u’ or ‘i’) and the other word begins in a vowel. Linking ‘r’ ▶ You know that a written post-vocalic ‘r’ at the end of a word is pronounced if the next word begins in a vowel. If you want to know more (identical to chapter 3) B I E S WA N G E R , Markus & Annette B E C K E R ( 5 2021): Introduction to English Linguistics. Tübingen. ▶ Chapter 3 of this very accessible introduction to English linguistics provides a short introduction to phonetics and phonology. B R O W N , Adam (2014): Pronunciation and Phonetics. A Practical Guide for English Language Teachers. London: Routledge. ▶ Provides a thorough introduction into phonetics and phonology and into teaching pronun‐ ciation. Also provides sample exercises. K ÖN I G , Ekkehard & Volker G A S T ( 4 2018): Understanding English-German Contrasts. Berlin. ▶ A very detailed contrastive analysis of English and German. Chapters 2 and 3 provide a comprehensive comparison of English and German phonology that is much more detailed than the comparison in this book. S C H M I T T , Holger (2016): Teaching English Pronunciation. A Textbook for German-speaking Coun‐ tries. Heidelberg. 76 4 “France is bacon” - sounds in speech <?page no="78"?> ▶ A thorough introduction into English pronunciation and how to teach it. In addition to providing methods and tools the book also provides an in-depth discussion of specific pronunciation problems of German learners of English. S K A N D E R A , Paul & Peter B U R L E I G H ( 4 2022): A Manual of English Phonetics and Phonology. Tübingen. ▶ An accessible and concise introduction to phonetics and phonology with university students from German-speaking regions in mind. https: / / www.seeingspeech.ac.uk/ ipa-charts/ . ▶ An amazing website! ! ! It provides you with a clickable IPA chart which gives you MRI, animated and ultrasound videos of speakers pronouncing consonants and vowels. Additional online content. 4.3 Why France is bacon - connected speech 77 <?page no="80"?> Abstract Content 5 Nodes and links - the mental lexicon 5.1 Webs of words, not lists 5.2 The mental lexicon - some basics 5.3 Stay connected - links in the mental lexicon 5.3.1 Defining features 5.3.2 Encyclopaedic relations 5.3.3 (Mostly) sense relations 5.3.4 Collocation and other relations in language use 5.4 What Taboo can teach us about teaching words 6 Think big! More than words 6.1 Why tea can be powerful … and why it shouldn’t 6.2 Lexical phrases 6.3 Patterns and their relevance for the classroom 6.4 Teaching patterns As an EFL teacher teaching vocabulary is one of the major parts of your job. You are better at this, the more you know about the mental lexicon and the more you are aware of the vast range of lexical units (words and beyond) that you can find in it. This is what the present part focuses on. As regards larger patterns, we see that these are a still somewhat neglected area of language teaching. Accordingly, one aspect is to provide you with examples of how we can raise awareness of such patterns and how we can teach them. Words and beyond More than words is all you have to do to make it real (E X T R E M E ) More than words is all you have to do to sound more idiomatic (R O L F K R E Y E R ) <?page no="82"?> 5 Nodes and links - the mental lexicon I am playing Taboo with a couple of friends. In the other team there are two sisters, A and B. A has to explain the word penguin to her team. A: The funny animal in the zoo last summer. B: Penguin! ! ! A: Yes! ! ! I: ? ? ? ? ? ? What this chapter is about Teaching vocabulary is probably one of most important tasks in the EFL classroom, as Wilkins already made clear more than 50 years ago: “[…] while without grammar very little can be conveyed, without vocabulary nothing can be conveyed” (1972: 111-112). Similarly, Michael Lewis describes “lexis [… as] the core or heart of language” (1993: 89). It stands to reason that teachers should be at their best when teaching words. One way to increase the degree of vocabulary retention (i.e. how many of the words we teach are actually remembered) is to integrate words (and other lexical units) into a dense lexical network. This chapter provides a description of the possible relations in such a network. I remember that when I was a child and I had to learn my vocabulary I would do that on the basis of alphabetical lists in the back of my schoolbook. Later, I would write my own word cards and take them with me to look at them whenever I had five minutes. Don’t get me wrong, word cards (and word card apps) are very helpful but in my case one side had the English term, the other the German term in isolation. This is not particularly helpful. To make things worse, the cards would be ordered alphabetically. Looking back, this was actually one of the least efficient and effective ways to increase my vocabulary. The following sections will show why - and suggest better ways. 5.1 Webs of words, not lists When we think about a lexicon or, more commonly, a dictionary, we think about a long list of words in alphabetical order. The mental lexicon is very different from this, which <?page no="83"?> becomes apparent quickly when we look at the game Taboo. Taboo, for those who do not know the game, is about explaining a given word within a given time limit to the members of your team. For each word there is a set of five taboo words that you are not allowed to use. As you can see in figure 5.1, for the term penguin these taboo words are bird, South Pole, swim, fly and biscuit. None of these are even somewhere in the alphabetical vicinity of the word penguin. Instead, we can see from this list that words are organised in a way that relates to our knowledge of the world and the things within it, in this case our knowledge about penguins. Note that four of the five taboo words would also be used in a description of a penguin, e.g. a bird that lives on the South Pole and cannot fly but can swim well. The fifth taboo word is very interesting, since it betrays a link of a completely different kind, namely a link that relates to aspects of cultural knowledge. This becomes particularly apparent if you do not understand why the word biscuit should be taboo. The answer lies in the fact that in the UK there is a chocolate bar called ‘Penguin’, which we do not have in Germany. Yet another fact about the way our mind stores words is illustrated at the beginning of this chapter. Person A does not even try to describe the word penguin by recourse to her knowledge of the animal or her cultural knowledge but she addresses her sister directly trying to activate some part of their shared lives. Obviously, words in the mental lexicon also contain links to what psychologists call our episodic memory, which, of course, differs from person to person, but may contain some overlap, as the scene above shows. Fig. 5.1: Penguin. All in all, there are a number of different ways in which one word can be linked to other words in our mental lexicon. We will explore these links in this chapter and we will see how they can be made useful for the teaching of vocabulary. Before we do this, however, a quick glance at the general makeup of our mental lexicon is in order. 5.2 The mental lexicon - some basics Current models conceive of the mental lexicon as a structure which consists of nodes and links between these nodes. Together these two basic ingredients form 82 5 Nodes and links - the mental lexicon <?page no="84"?> relevance of data vast networks which represent our knowledge of words and how to use them. This idea is not new and goes back at least to the late sixties and early seventies, when researchers tried to account for experimental results on how informants process words and sentences. The network model of Collins and Quillian (1969: 241), for instance, describes hierarchies in so-called taxonomic networks, consider figure 5.2 below. Fig. 5.2: An example of a taxonomic network model (after Collins and Quillian 1969: 241). A taxonomy is a classification system that (usually) organises elements into classes, which then again are organised into other classes and so on. In the example above, for instance, the superordinate class of ‘animal’ falls into the two subclasses ‘bird’ and ‘fish’, the former of which distinguishes canaries from ostriches while the latter is divided into sharks and salmon. On each level of the taxonomy we have features that are attributed to the relevant class. Animals have skin, for instance, while only birds have wings and only fish have fins. Lower classes will inherit attributes from their superordinate class(es). Since a canary is a bird it, too, has wings, can fly and has feathers. Being a bird, it is also an animal, which means it has skin, eats, breathes and can move around. Anybody can come up with a model of anything. Models only make sense if they manage to explain something, for example, experimental data. Collins and Quillian asked their informants to verify sentences like a canary can sing or a canary has skin. Both sentences are obviously correct, however, it took informants longer to verify the second sentence. According to the authors this finding validates their model, since the feature ‘can sing’ is immediately attached to the canary node, whereas the feature ‘has skin’ is further removed from the node ‘canary’ (we will discuss another explanation further down below). Figure 5.3 shows another example of a network model of the mental lexicon. As you can see, the model does not involve any hierarchies. Rather, it is a model that captures loosely organised associations. For example, vehicles can be found on the street, which 5.2 The mental lexicon - some basics 83 <?page no="85"?> association strength leads to a connection between the two nodes ‘vehicle’ and ‘street’. Different kinds of vehicles are connected to the ‘vehicle’ node since they are members of that category, each of these vehicles, of course, can also be found on the street. In this network model a distinction is made between different degrees of relatedness of concepts. The nodes we have discussed so far are strongly related since they share a number of nodes. In the model by Collins and Loftus this is expressed through proximity of nodes. In contrast to that the nodes ‘fire engine’ and ‘cherries’ are far apart since they only share one feature, namely the fact that both are red. Fig. 5.3: An example of an associative network model (after Collins and Loftus 1975: 412). In both of the models above the concept of relatedness or strength of association is important. More recent models would represent differences in that respect by positing links that are stronger or weaker. For example, having skin is a feature of the canary just like the fact that a canary sings is one of its features. So, in more modern models both, the node ‘skin’ and the node ‘sing’ would be immediately connected to the node ‘canary’. However, the link from ‘sing’ to ‘canary’ would be stronger than that from ‘skin’ to ‘canary’. This is shown in figure 5.4, where the link at the top right is thicker than the link at the bottom right. 84 5 Nodes and links - the mental lexicon <?page no="86"?> activation spreading Fig. 5.4: A lexical network around the node ‘canary’. The fact that something has skin holds true for many different animals, and, therefore, is not particularly indicative of a canary. In contrast to that, there are not that many objects in our world that sing. This as well as the colour yellow is more typical of canaries. When it comes to the game Taboo, words that refer to such typical features would usually be taboo words. A pivotal concept in the network models is that of (degree of) activation and that of activation spreading. Without any input the nodes in our lexical network will be in‐ active. Nodes can be activated as a consequence of visual or auditory input, i.e. when you read or hear a word. Once a node is active it will spread its activation to all the neighbouring nodes. Let’s again take the game Taboo to illustrate this. If your teammate gave as a hint ‘something that sings’, this would result in the activation of the node ‘sing’ and the node would spread its activation through the network to all the nodes it is connected with, one of these being the node ‘canary’. However, even if your team‐ mate repeatedly told you that they are looking for something that sings you would usually not get that they are looking for the word ‘canary’. This is because every node in a network has an activation threshold. This threshold has to be met to activate a node. Your chances of getting the word would be higher if they said ‘a bird that sings’ or even ‘a yellow bird that sings’. In the latter case the node ‘canary’ receives activation from three different nodes which will eventually result in its full activation. A model of this kind is intuitively clear and you can see it at work whenever you play the game Taboo. However, there is also experimental evidence that makes a strong case for the validity of this kind of network model. One piece of evidence are the findings by Collins and Quillian (1969) discussed above. The fact that informants take longer to verify the sentence a canary has skin than the sentence a canary can sing can be explained by the different strengths with which the respective features are associated to our concept of a canary. A network model represents this by the thickness of the individual connecting lines. 5.2 The mental lexicon - some basics 85 <?page no="87"?> Before you read on… … think about the following little game. Take a friend and tell them that you will ask them a series of questions which they have to answer loudly and as quickly as they can. Then you show them a number of (maybe ten or so) things that are white and ask them each time “What is the colour of this? ”, making them say the word white many times. After that you ask them “What does the cow drink? ” What do you think will happen? What would usually happen (if your friends do not know the game) is that they will answer “Milk! ”, even though, of course, they know that cows drink water. However, they also know that cows produce milk which leads to an association of the nodes ‘milk’ and ‘cow’. This association is a lot stronger since, at least in our cultural context, when we think of milk we usually think of cow’s milk, even though we know that all mammals give milk. As you play this game with your friends every time they utter the word white the relevant node in the mental lexicon will get activated and this node will spread its activation to all the nodes that represent things that are white. This will not result in full activation of the node ‘milk’ but it will activate the node to some extent, thereby warming it up, if you will. As soon as you then mention the word cow the additional activation from the node ‘cow’ will lead to a fully activated node ‘milk’ before the node ‘water’ can be activated, even though the node ‘drink’ might have a somewhat stronger connection to ‘water’ than ‘milk’. The relevant portion of a possible network model is shown in figure 5.5. Fig. 5.5: A lexical network around the node ‘milk’. A final piece of evidence I want to provide comes from an experiment by Garnes and Bond (1976; cf. Aitchison 3 2003: 228). Informants were played sentences like the ones shown under (1). (1) Paint the fence and the [? eɪt]. Check the calendar and the [? eɪt]. Here is the fishing gear and the [? eɪt]. 86 5 Nodes and links - the mental lexicon <?page no="88"?> networks and teaching vocabulary The string in phonetic transcription represents the final sounds of words like gate, date and bait. The researchers manipulated the first sounds such that the initial consonants were ambiguous, falling between [g], [d] and [b]. However, the informants reported hearing gate, date and bait, respectively. Again, this can easily be explained by a network model of the mental lexicon, since the three words would be associated with the words fence, calendar and fishing gear. As soon as one of these, say, fence is mentioned, the respective node ‘fence’ in the mental lexicon would be activated, spreading activation to the associated node ‘gate’ which, as a consequence, is activated to some extent. When the informants hear a sound sequence that is very similar to the word gate, this node gets fully activated and the informants understand gate. As you can see, network models are very powerful when it comes to explaining observations derived from experiments (and other observational data, too) about how words are associated with one another. It therefore makes sense to assume that words are, in some way or other, represented as networks in our minds. Teaching vocabulary, as a consequence, involves helping your students create such mental networks. If you recall your own school days you will find that vocabulary work was often driven by these considerations. Usually you would not have been given a list of words arranged in alphabetical order, for example the ten most important adjectives beginning with ‘a’. Instead vocabulary would have been grouped into word fields, such as the word field of ‘food’. This way, adjectives would already be grouped in a meaningful way, such as adjectives that describe how food can taste, nouns that describe possible in‐ gredients for recipes or verbs that describe typical actions while cooking. When it comes to learning new words it is helpful to integrate them as deeply as possible into the already existing structures of the mental lexicon. This means the larger the number of links that the word has to other words, the more likely your students will remember that word (although, of course, there are other factors that come into play as well). This is why, as a teacher, it helps to have an awareness of the wide variety of links that can exist between words. 5.3 Stay connected - links in the mental lexicon Before you read on … … take a few moments and think about Taboo cards for the four words present, tense, bush and butterfly. For each of these four words write down five taboo words that people would not be allowed to use if they wanted to explain one of these four words. We will work with these later, so make sure you take a minute to write the four times five words down. 5.3 Stay connected - links in the mental lexicon 87 <?page no="89"?> The links in our mental lexicon are of different kinds. The first kind is typical or ‘defining’ features. For example, it is one of the typical features of dogs that they bark and, consequently, it makes sense to assume that there is a connection between our mental representation of the word dog and the word bark. The next kind of relation can best be understood by contrast to the first relation. Recall our discussion of the taboo card for the word penguin. The first four taboo words were bird, South Pole, swim and fly. All of these are related to typical features, which becomes apparent in a possible definition of a penguin as a bird that lives in the South Pole and can swim but cannot fly. This is a definition that could be found in a dictionary. In contrast to dictionaries encyclopedias provide a lot more information on a certain topic. In the case of the penguin this might include additional information of a cultural kind like, for example, the fact that a chocolate bar or a biscuit is named Penguin in the UK or maybe that there is a publishing house of the same name. This kind of relation between words, therefore, can be described as ‘encyclopaedic’ relations. In addition there are links that are of a more linguistic nature. They are usually referred to as ‘sense relations’ and have been explored in the field of structuralist semantics. Examples are relations such as that of synonymy, i.e. sameness of meaning, for instance couch and sofa, or hyponymy, i.e. the relation of subordination among categories, the category dog, for example is subordinate to the category mammal. While this last kind of relation is rather general or abstract and independent of language use and experience, the final kind of relation, ‘collocation’, has to do with frequencies and expectancies. Every competent native speaker knows that some words go well together whereas other words do not. Most words have a particular set of co-occurring words. The word handsome for instance is more likely to occur together with words relating to men than words relating to women, whereas the situation is reversed for the word pretty. These kinds of expectancies are also represented as links between nodes in the mental lexicon. In the following we will discuss each of the four relations in more detail. A table at the end of section 5.3 will provide an overview of all the links that are relevant for the mental lexicon. 5.3.1 Defining features The linguistic subdiscipline of semantics, i.e. the scientific study of decontextualised meaning communicated through language, among other things, aims to describe the 88 5 Nodes and links - the mental lexicon <?page no="90"?> extension and intension cue validity meaning of words. In principle, there are two options, ‘extension’ and ‘intension’. The first term describes the meaning of a word by listing or pointing at all of the objects, states, feelings et cetera that are captured by a given term. For example, a possible way to define the meaning of the word dog would be to list all the animals in the world that fall into that category. To a large extent this is how infants learn the meaning of words, when their caretakers point to an animal and say “Look! A dog! ” The description of meaning by intension (also called ‘sense’) does not make any reference to the non-lin‐ guistic world but captures the meaning of a word with the help of other words, i.e. words that help to define the term in question. The distinction of extension and inten‐ sion goes back to Friedrich Ludwig Gottlob Frege (1848 - 1925), a logician, mathema‐ tician and philosopher. Frege (1892) draws our attention to the fact that two words may have the same extension but different intensions, for instance the words morning star and evening star. Both of them refer to the planet Venus, i.e. they have the same ex‐ tension, but they mean two different things with regard to the intension of the two terms. The morning star is the last ‘star’ that you can see in the morning sky, whereas the evening star is the first ‘star’ in the evening sky. Note that both kinds of relations are represented in our mind. For example, part of the meaning of the word cat for me is that the word makes me think of all the different cats that we had in our family. This kind of knowledge or experience comes close to encyclopaedic relations, which we will discuss further below. The concept of intension makes reference to the features that define a concept. Ob‐ viously, all of these are important when we want to describe the mental lexicon but some are more important than others in the sense that they are more typical than others. When it comes to teaching words and establishing lexical networks in the minds of your students it makes sense to focus on words relating to the more typical features. The idea of typicality is captured in the concept of ‘cue validity’ (Rosch & Mervis 1975). The term describes the degree to which a given feature can be seen as a cue for a particular category. The feature ‘barks’, for example, is highly indicative of dogs and we would assume that the two nodes ‘dog’ and ‘bark’ will be strongly linked in the mental lexicon similar to the strong link that exists between the node ‘cat’ and the node ‘purr’ or that between ‘feathers’ and ‘bird’. Note, however, that a given feature does not have to be restricted to one category only to be typical. I am not a biologist but as far as I know there are three kinds of animals that have beaks, namely the platypus, the octopus and birds. Still, it would make sense to establish a link between the node ‘beak’ and the node ‘bird’ since we strongly associate that feature with birds. We also find features that, though not true for every member of a category, might still be typical of the category as a whole. A case in point is the feature ‘can fly’: even though ostriches and penguins cannot fly and even though there are a lot of non-bird animals that can fly (e.g. bats or insects), the ability to fly is closely connected to birds in our represen‐ tation of the outside world. Hence, it makes sense to assume that our mental lexicon has a strong connection between the node ‘bird’ and ‘to fly’. 5.3 Stay connected - links in the mental lexicon 89 <?page no="91"?> 5.3.2 Encyclopaedic relations The term ‘encyclopaedic relations’ is a broad cover term that makes reference to any kind of background knowledge that we have of an entity in the non-linguistic world and that goes beyond the information captured by the definition. It is not defining for the concept ‘penguin’ that there is a biscuit or a publishing house of the same name, but it is a piece of knowledge that some people might associate with the word. For those people the respective node in their mental lexicon would be connected to the node for ‘biscuit’ or ‘publishing house’. Again, we see that the authors of the game Taboo are very well aware of the fact that bits of encyclopaedic knowledge can betray the word someone is looking for. The taboo card for Hardy, for instance, lists Thomas and author as taboo words, since (as you may know) there is a nineteenth century British author called Thomas Hardy. The other three taboo words are tough, strong and sturdy, all of which mean more or less the same as the word hardy. A comparison of the two groups of taboo words makes it clear immediately that encyclopaedic relations are very dependent on the time one lives in and on the culture one grew up in. A more efficient taboo word for the current times might be actor instead of author, since it is likely that more people nowadays know Thomas Hardy the actor than Thomas Hardy the author. Another good taboo word would be Venom, since this might be one of the best-known roles of the actor. It would not make much sense to change the first three taboo words: Language changes a lot more slowly than popular culture. Fig. 5.6: Hardy. As pointed out above encyclopaedic relations are concerned with background knowledge. This kind of knowledge goes beyond information you would find in an encyclo‐ paedia and also involves facts that are stored in your own episodic memory, i.e. that part of your memory that contains recollections of things that have happened in your life. That this kind of information is also relevant for our mental lexicon is shown by the fact that sister A could make reference to a visit to the zoo to make sister B aware that the word she was looking for is penguin. 90 5 Nodes and links - the mental lexicon <?page no="92"?> When it comes to teaching words the power of encyclopaedic relations can be exploited to help your students remember particular words better, since a larger number of relations would usually result in better storage and better retrieveability of a given word. Textbooks make use of this and would usually introduce new words by providing a rich background either of encyclopaedic information or of information contained in a story. This way, it is more likely that the vocabulary that needs to be learned becomes part of a dense network in the learner’s mind. 5.3.3 (Mostly) sense relations The concept ‘sense relations’ (see, among others, Cruse 1986, Lipka 2002, or Lyons 1977) relates to the description of the meaning of a word by pointing out how this word relates to other words in a language. Sense relations are used a lot in dictionaries. For example, the online version of the Collins English Dictionary (www.collinsdictionary .com) defines ‘hand’ as “[…] the part of your body at the end of your arm. Each hand has four fingers and a thumb” and, this way, makes use of the part-whole relation: the hand is a part of the arm and is a whole to the fingers and the thumb. Another relation that is used frequently is that of (near-)synonymy, as in “If you are unhappy, you are sad and depressed”. When the dictionary defines ‘eagle’ as “a large bird that lives by eating small animals” the kind-of-relation, also called ‘hyponymy’ (an eagle is a kind of bird) plays an important role. An awareness of the different kinds of sense relations is important for language teaching as each type provides a new way of explaining a term and of anchoring a term in the mental lexicon of your students. By the way: Ancient Greek everywhere Many of the terms that you will learn about in the following section involve elements coming from Ancient Greek. While it is not necessary for understanding the linguistic concept behind the individual terms I still think it makes it easier for you to connect the linguistic terminology to the concept behind it. A very important Greek element is -onym or -onymy. The root -onym is a form of onoma, meaning ‘name’. We will find it a lot together with a range of prefixes. One of these is the prefix syn-. It, too, is of Greek origin and means ‘similar’ or ‘alike’, as in synonym, i.e. a word that means (more or less) the same as another word, or synonymy, i.e. the relation between synonymous words. Its opposite antstems from Ancient Greek antiwhich means ‘opposite’ or ‘against’. Another pair of opposites is that of hyperand hypo-, coming from Ancient Greek huper, meaning ‘over’ and hupo, meaning ‘under’, respectively. It is the same prefix we find in words like hypertension or hypoallergenic. Yet another pair of Greek-based opposites is that of holos and meros, meaning ‘whole’ and ‘part’, respectively. The first is used in a word like hologram, where the prefix can be understood to indicate that a whole, i.e. a three-dimensional, object is depicted as opposed to a drawing 5.3 Stay connected - links in the mental lexicon 91 <?page no="93"?> synonymy antonymy with only two dimensions. It is difficult to find non-jargon examples of words starting with the prefix mero-, which is why it makes just as much sense to take the linguistic term meronym as an example. The meronym is the name for a part. The well-known prefix polystems from Greek polys and means ‘many’. Someone who is polyglot speaks many languages. We also find the prefix in polysemy or polysemous where it combines with the Greek root sema, meaning ‘sign’ (the same as in semantics). A polysemous word, then, is a word that is many signs, i.e. has many meanings. Another well-known prefix is homomeaning ‘same’, as in homosexuality or homogenous. The prefix combines with stems based on Ancient Greek phone, meaning ‘voice’ and graphe, meaning ‘writing’, as in homophony and homography. An extremely useful way to describe the meaning of a word, of course, is through synonymy, i.e. the sameness of meaning. True synonymy is extremely rare and there are only very few examples, such as the word pairs couch - sofa or car - automobile. More frequently we find partial synonymy, where the meanings of two words are very similar but do not overlap completely. Often this is the case in highly sensitive areas of human life, such as death, bodily functions and sex. Depending on the context, for example, you might say that someone died or that they passed away or that they kicked the bucket. All of these expressions mean the same but they cannot be used fully in‐ terchangeably. Similarly, you might tell someone that you need to go to the restroom, that you need to find the loo, that you are looking for the toilet, that you need to do number two or that you need to take a s*** (among other alternatives). The example of the taboo card Hardy illustrates the relevance of synonymy since three of the five taboo words were (half-) synonyms. Another basic sense relation is that of antonymy. This is the opposite of synonymy, i.e. it refers to oppositeness of meaning (note how I used this sense relation to describe the meaning of antonymy in relation to the meaning of synonymy). There are different kinds of antonymy, the most basic one being that of ‘complementary antonymy’. The term refers to the fact that a word and its complementary antonym complement each other. That is both words together represent all the possible alternatives there are. It can be an either-or relationship, sometimes also referred to as ‘binary antonymy’: if you are not the one, then you have to be the other and vice versa. An example would be the pair dead and alive: there is nothing in between or, if you want to sound especially intelligent and impress with your knowledge of Latin, tertium non datur - a third (op‐ tion) is not given. In contrast to binary antonymy ‘gradable antonymy’ does allow for in-between states. Something can be hot or it can be cold or it can be many things in between like warm, tepid or cool. Because in-between states are possible it follows that if something is not one thing you cannot say that it, then, has to be the other. If some‐ thing is not hot, it does not mean that it is cold, it could just be warm or cool and so on. ‘Directional antonymy’ describes opposite directions of movement, for example a 92 5 Nodes and links - the mental lexicon <?page no="94"?> hyponymy and hyper‐ (o)nymy holonymy and mero‐ nymy polysemy person at point A will go to point B from point A but they will come to point B from point A. The more general sense relation is that of ‘relational antonymy’ where two terms describe the same situation from opposing perspectives. If a person A transfers goods to a person B in exchange for money we would say that A sells something to B and that B buys something from A. The two words sell and buy are relational opposites. Other examples would be give and take or teach and learn. The next pair of sense relations is indeed an example of relational oppositeness. We speak of ‘hyponymy’ if one word describes a category that is subordinate to another category, e.g. a mammal is a particular kind of animal, a dog is a particular kind of mammal, a golden retriever is a particular kind of dog. In all of these cases, the first word is a hyponym (a subordinate term) to the second word. Looking at the same situation from the perspective of the second word we get the relation of ‘hyperonymy’ (or ‘hy‐ pernymy’), i.e. the relation of a superordinate term to a subordinate term: an animal is a hyperonym (or hypernym) of mammal, mammal is a hyperonym of dog, dog is a hyperonym of golden retriever. If translated verbatim from the Ancient Greek roots (see above), a hyponym is an ‘undername’, a hyperonym an ‘overname’. Admittedly, these terms are not actually real English terms and they sound a bit strange, but the German terms Unterbegriff and Oberbegriff do not. Yet another example of relational oppositeness are the two sense relations ‘holony‐ my’ and ‘meronymy’. The first describes the case when one word (the holonym) is a whole to another word (the meronym) that is the part of that whole, whereas the second describes the same situation from the opposite perspective, i.e. one word (the meronym) describes the part of a second word (the holonym) that describes the whole. For in‐ stance, the word arm is a meronym to the holonym body, the word hand is a holonym to the word finger, which is one of its meronyms, and so on. On the basis of metaphorical extension we can describe the member-of-relation as an example of meronymy, e.g. the Prime Minister as a member of the government. The term ‘polysemy’ refers to one word that has two or more related meanings. An example would be the word neck which has three meanings, namely ‘the neck of the body’, ‘the neck of a bottle’ and ‘the neck of a shirt’. The latter two are semantically related to the first meaning, since a bottle is similar in shape to the torso of a body which makes the narrow part of a bottle the neck of the bottle. The neck of a shirt is related to the first meaning for the same reason. In addition, the neck of a shirt is that part of the shirt which covers the neck of the body. Another example of polysemy is that of the word head, which describes a body part but is also used to refer to the person or persons in charge of an organisation. Note that the semantic relatedness of the different meanings needs to be obvious to the ‘normal’ present-day language user. The word pupil, for example, has two meanings, namely ‘the hole in the middle of someone’s iris’ and as a synonym of student. The two meanings are etymologically related since you can see a small reflection of yourself in someone’s pupil. However, nowadays language users are not any more aware of this connection, so we wouldn’t speak of a case of polysemy but of homonymy. 5.3 Stay connected - links in the mental lexicon 93 <?page no="95"?> homonymy homophony and homo‐ graphy ’Homonymy’ is not strictly a sense relation as meaning is not involved to create a link. Instead spelling and pronunciation play a central role. Two words are homony‐ mous if they have the same spelling and the same pronunciation but have two different meanings that are not related, at least in the eye of the typical language user. Whereas in the case of polysemy we said that one word has two or more related meanings, in the case of homonymy we would say that two different words (with different meanings) happen to be written and pronounced the same. The string bat, for example, can refer to the animal or to the implement that you use when you play baseball. The form bank relates to two words, namely bank as bank of a river or bank as the financial institution. In the case of homonymy two or more words happen to be written and pronounced the same. With homophony we have two or more different words that sound the same but are written differently such as right and write. In the case of homography the spell‐ ing is identical but the pronunciation differs as in the examples (the) tear and (to) tear or in the case of present tense (to) read and simple past read. Before you read on… … take a look at the following dialogue taken from the TV show How I Met Your Mother. Robin has just come back from Argentina where she met her boyfriend Gael and explains to her friends Barney, Lily and Ted how her life with Gael has changed her. This is what happens next: B: Please, vacation romances have an expiration date. Gael’s got a best-if-banged-by sticker on him. Once your romance starts to stink, you’ll dump his ass down the drain like sour milk and go back to being ‘unevolved Robin’, the one we actually like. Back me up here, Ted. T: I’m just happy Robin’s happy. B: I’m telling you, within three days… L: Ooh, here he comes. Switch to big words. [Gael comes in] B: Within a triad of solar periods, you’ll recognize your dearth of compatibility with your paramour and conclude your association. R: My journey was transformative and I reassert my commitment to both the aforementioned paramour and the philosophies he espouses. G: What are we talking of ? Baseball? B: This is all gonna return to masticate you in the gluteals. Support my hypothesis, Ted. T: I’m just jubilant my former paramour’s jubilant. Take a look at the exchange after Lily says “Switch to big words” when Gael comes in. What do you think Barney, Robin and Ted want to say? How would they normally phrase it? Which of the sense relations from above is exploited here for comical effect? 94 5 Nodes and links - the mental lexicon <?page no="96"?> When Lily tells her friends to switch to big words she has words in mind that are part of a much more formal register and are not used in everyday language and, therefore, will most likely not be understood by Gael, who does not speak English very well. In everyday English the dialogue might have been something like the following. B: Within three days you’ll realize that you and your lover do not fit and you’ll end your relationship. R: My journey has changed me. I really love Gael and the way he thinks. G: What are we talking of ? Baseball? B: This’ll all come back and bite you in the ass. Back me up here, Ted. T: I'm just happy Robin’s happy. A comparison of the two versions makes it clear that the writers of this episode played with synonymous terms and expressions. Before you read on … … take a look at the four cards that you have created at the beginning of section 5.3 and see if you can identify the relations that your taboo words have to the word on top. 5.3.4 Collocation and other relations in language use The defining features and the encyclopaedic relations discussed above have to do with what we know about the objects, states and actions that a word refers to. We have seen that sense relations are abstract relations that words enter into with other words in the language system. The relations we will be looking at now are established on the basis of language use, for instance because of the fact that two words are used together a lot. With the advent of modern computer technology linguists have become able to look at extremely large amounts (millions of words) of authentic language text all at once. This new approach to language brought with it new insights into the nature of language. More specifically, linguists found that to a large extent language use shows many patterns of co-occurring words or structures. Above, for example, we have already glanced at the two adjectives pretty and handsome. They both mean more or less the same i.e. ‘good-looking’ and, being adjectives, would usually be used to premodify a noun. Still, we would generally speak of a handsome man rather than a pretty man and of a pretty woman instead of a handsome woman. These are two patterns of language use that cannot really be explained, but they occur if we look at huge amounts of authentic language-use data at once. Native speakers are aware of these patterns on some level since that is the way they use their language. Language learners, in contrast, even if they are very advanced, are often not aware of this. As a consequence, the way they use a foreign language, though not incorrect in the strict sense, sounds strange and not idiomatic. We will take a closer look at what idiomaticity means in the next chapter. For now we will focus on a few of these patterns that we 5.3 Stay connected - links in the mental lexicon 95 <?page no="97"?> collocation find in authentic language use and we will see that words not only have preferences for words but also for semantic features or grammatical categories and grammatical structures. Before you read on … … take a look at the utterances below. Focussing on naked eye, what kinds of regularities can you see in the words that occur on its left, i.e. what can you say about the first word to the left, what about the second word? Can you find regularities if you move further to the left (not necessarily fixed to a particular position or restricted to words)? The mite is just visible to the naked eye. The fine silt particles being indistinguishable to the naked eye. The whiskers were too small to see with the naked eye. Many of these parasites cannot be seen by the naked eye. This behaviour is almost too quick to see with the naked eye. The exercise above is based on a discussion of naked eye by John Sinclair (1996). The first two words are easy. The first word to the left is always the definite article the. When you think about it, this is really strange. Why shouldn’t we find my naked eye, their naked eye or Barbara’s naked eye. All of these would be perfectly grammatical noun phrases, we just don’t find them. Not everything that is grammatically possible is also natural. Before the definite article we would usually have a preposition, in the examples above to, with, and by. Further to the left we find the words visible, indistinguishable, see and seen. All of these relate to visual perception. Still further to the left we find words or phrases that express that it is not easy to perceive the thing that is encoded as the subject of the sentence. The mite is just visible, the whiskers are too small, the behaviour too quick and the parasites cannot be seen. In one case this notion of difficulty is expressed by a morpheme: the silt particles are indistinguishable. The first kind of relation, the co-occurrence of a word with another word, is referred to as ‘collocation’. John Rupert Firth is usually credited with drawing our attention to the relevance of the concept for linguistics (roughly 70 years ago). “[Y]ou shall know a word by the company it keeps” (1957: 11) is a quote of his that you would typically find in any publication on the topic. The term ‘collocation’ itself is of Latin origin and derives from com- (meaning ‘together’) and locare (‘to place’), so it means ‘placing things together’, in our case words. By the way, do you notice the assimilation that leads to collocation instead of comlocation? The concept has been around for quite some time now and there have been many attempts at coming up with definitions. I will spare you unnecessary detail and, instead, will just hint at a couple of aspects that will help you grasp the concept. As was said above, collocation is about the co-occurrence of words (or phrases). This does not mean that the words that form a collocation need to be next to each other. Both, the handsome man and this man is 96 5 Nodes and links - the mental lexicon <?page no="98"?> colligation semantic preference semantic prosody very handsome would equally be regarded as instantiations of the collocation of handsome and man. In addition, the co-occurrence of the two words needs to be noteworthy somehow, it needs to hit the eye. Some researchers have claimed that this co-occurrence needs to be characteristic, which does not necessarily mean that it has to be highly frequent. For example, the phrase my father is a lot more common in authentic language use than the phrase handsome man, still the first would not be considered a collocation, since the determiner my is found very frequently with a very large number of nouns. Handsome, in contrast, is far less frequent but it occurs very typically with a small number of nouns, man being one of them - more of that later. That collocations are ‘real’ becomes particularly obvious when we compare the collocations of different languages. In German, for instance, we say ein starker Regen, which indicates that the adjective stark is a collocate of the noun Regen. In English, the typical collocate of rain is heavy, not strong. Similarly, German students ‘make’ their homework whereas English students do their homework. The second kind of pattern in the dataset above, the co-occurrence of a word and a gram‐ matical category or structure, is called ‘colligation’. In the data set above we see that naked eye will often be preceded by a preposition on its second position to the left. Other exam‐ ples are the verbs to say and to tell: the first colligates with the structure ‘NP + PP’, ‘say something to someone’, while the verb to tell chooses ‘NP + NP’, ‘tell someone something’. In one of the previous chapters we talked about the noun assumption. This noun frequently co-occurs with a that-clause, e.g. the assumption that linguistics is difficult. The other two relations are semantic in nature and we see that the words naked eye frequently occur with verbs or adjectives that involve visual perception, i.e. we could say it co-occurs with the semantic feature [+visual perception]. This semantic feature is neutral in the sense that its absence or presence would not be regarded as positive or negative. The co-occurrence of a word with a neutral semantic feature is called ‘semantic preference’. Another example is the verb to undergo, which likes to take nouns that refer to medical terms: you undergo a treat‐ ment, an operation or surgery. We could, therefore, say that undergo has a semantic prefer‐ ence of [+medical]. The semantic feature that is shown if we move further to the left, [+difficult], is not neutral because humans usually do not like the things that are difficult. This co-occurrence of a word with an evaluative semantic feature is called a ‘semantic pro‐ sody’. More specifically, we are talking about a negative semantic prosody in the case above. A similar negative semantic prosody can be found with the verb to cause (or its Ger‐ man counterpart verursachen). You cause accidents, a disaster, outrage or catastrophes, all of which are negative things. You cannot cause happiness or bliss. The verb to cause, we say, has a negative semantic prosody of [+negative]. In this context, Firth’s idea that “you shall know a word by the company it keeps” (1957: 11) expresses an interesting insight: the words that usually co-occur with a given word contribute to the meaning of that word. In and of itself the verb to cause is ‘harmless’ since, in principle, it could be used for positive as well as negative things. However, since it is typically used with negative things, this negativity reflects onto the meaning of the verb itself. Whenever we hear it being used we will expect something bad. Figure 5.7 sums up the insights we gained from the naked-eye exercise. 5.3 Stay connected - links in the mental lexicon 97 <?page no="99"?> Fig. 5.7: Four kinds of patterns with naked eye. In the next chapter we will see that the patterning of language use even includes much larger and more complex structures. These, too, are stored somewhere in our mental lexicon but we will devote an additional chapter to them. For now, table 5.1 sums up the possible relations in the mental lexicon that we have discussed so far. Please note that when the table uses the word ‘entity’ this stands as a proxy for objects, feelings, actions and anything else that we may encounter in the world around and within us. Relation Description Examples defining features the attributes you would list if you wanted to describe an entity. Typically, these would be attributes with a high cue validity. bird - feathers dog - barks winter - cold encyclopaedic knowledge knowledge about an entity that goes beyond the mere definition, e.g. interesting cultural or historical facts penguin - biscuit hardy - actor hardy - Venom synonymy sameness of meaning couch---sofa antonymy binary gradable directional relational oppositeness of meaning … that only has two states, also called ‘comple‐ mentary antonymy’ that allows in-between states that concerns opposite directions of move‐ ments or orientation that views the same situation from opposite perspectives dead - alive freezing - hot come - go buy - sell 98 5 Nodes and links - the mental lexicon <?page no="100"?> Relation Description Examples hyponymy one word, the hyponym, is subordinate to another word, the hypernym or hyperonym dog---mammal hyper(o)nymy one word, the hyper(o)nym, is superordinate to another word, the hyponym mammal - dog holonymy one word is the whole to another word that is the part/ a member body - head government - Prime Minister meronymy one word is the part to/ a member of another word that is the whole head - body Prime Minister - government polysemy one word with two or more related meanings neck (of body) neck (of bottle) neck (of shirt) homonymy two or more words that are written and pro‐ nounced the same bat (animal) bat (baseball) homophony two or more words that are pronounced the same (but written differently) right - write homography two or more words that are written the same (but pronounced differently) read (present) - read (past) collocation the typical co-occurrence of a word with an‐ other word handsome man heavy rain colligation the typical co-occurrence of a word with a grammatical category or construction say + NP + PP tell + PP + NP semantic preference the typical co-occurrence of a word with a neutral semantic feature undergo [+ medical term] large [+quantity/ size] semantic prosody the typical co-occurrence of a word with an evaluative semantic feature cause [+negative] provide [+positive] Table 5.1: A list of relations that a word may enter into. 5.4 What Taboo can teach us about teaching words I hope to have shown that words exist in large networks of connections within our mental lexicon. These connections, as we have seen, go far beyond mere definitional features and include all kinds of relations - some of a more linguistic kind, others of a more cultural nature. If we assume that new words are remembered better if we relate 5.4 What Taboo can teach us about teaching words 99 <?page no="101"?> them to huge numbers of other words in our mental lexicon, it makes sense to exploit all of the relations discussed above to the biggest extent possible. To some degree we can witness that if we take a closer look at the game Taboo. Initially, we might assume that taboo words would usually be those that are needed to define a word. While we do find such cards in the game, this is usually the case, if there are no other relations that can betray the word we are looking for. An example is the word Richter Scale. It is reasonable to assume that people are not aware of any encyclopaedic facts about the concept (at least I am not), there are no sense relations that the term enters into and there are no collocates that come to mind when thinking of the term. The situation is highly similar with the word gazump, about which the Collins English Dictionary has the following to say: “If you are gazumped by someone, they agree to sell their house to you, but then sell it to someone else who offers to pay a higher price”. So, in both cases we are left with defining features, which is what we see as taboo words in figure 5.8. By the way, this makes these cards particularly difficult cards in the game. Fig. 5.8: Richter Scale and Gazump. As soon as information beyond that which is represented by defining features can be expected to be available to most potential players, the respective words will become taboo words. We have seen this already in the case of the card for the word Penguin, where the taboo word biscuit keeps players from making use of their knowledge about sweets. The card for the word Panda is very similar in that it combines defining features with encyclopaedic knowledge. The Collins English Dictionary defines a panda as “a large animal rather like a bear, which has black and white fur and lives in the bamboo forests of China”. As you can see in figure 5.9 the first four taboo words are related to the definition of the term. The last taboo word, similar to the case of Penguin, makes reference to cultural background information, namely the fact that British police cars are by some referred to as panda or panda car. 100 5 Nodes and links - the mental lexicon <?page no="102"?> Fig. 5.9: Penguin, Panda and Boat Race. Sometimes encyclopaedic relations seem to be more important than the others, as shown on the right of figure 5.9 which shows the card for Boat Race. The term refers to an annual rowing competition on the River Thames between Oxford and Cambridge University. Obviously, the authors of the game thought that these terms were more of a giveaway than sense-related words that could be used in a description like “A famous competition where small ships from two famous universities run against each other”. In other cases, sense-related words seem to be particularly indicative, as the examples in figure 5.10 show. The five taboo words for obscure, for example, are all synonyms or near-synonyms of the word on top, which by the way is homophonous, making reference to the adjective obscure as well as to the verb to obscure. The card in the middle, too, shows a homophonous string on top. Present can either refer to something that you give to someone and would then be synonymous to the taboo word gift, or it refers to present as in the present moment. In relation to this sense we see two taboo words that are more or less synonymous, namely now and current. The words past and future are related to the word present in that all three refer to points in (or stretches of) time. All of them, we can say, are subordinate categories of the superordinate category ‘time’, which is why we say the three words are co-hyponyms. The card to the right is very similar. Tense is homophonous and either refers to a state of mind or to the way linguists and grammarians talk about time, as in tense and aspect. Nervous and uptight are near-synonyms of the word on top. The remaining three words are related to tense since all of them are specific kinds of tenses, i.e. they are hyponyms of the hyperonym tense, as in past tense, present tense and future tense. Fig. 5.10: Obscure, Present and Tense. 5.4 What Taboo can teach us about teaching words 101 <?page no="103"?> Before you read on … … take a look at the two taboo cards below. With regard to the card on the left, what can you say about the string BUSH and what kinds of relations can you identify? In a game of Taboo one of my team mates had to explain the word butterfly without using the five taboo words shown on the right. His way to solve this problem was “Mohammed Ali would float like one of these, as well as stinging like a bee”. Why does this give away the word on top and what kind of relation is involved here? As for the card BUSH, we see that the word on top is both, homophonous and polysemous. The form bush is used as a surname but also has a plant-related meaning, i.e. we have two forms that are spelt and pronounced the same but have two unrelated meanings. In addition, within the context of plants the word bush has two related meanings, namely that of a particular low-growing kind of plant (as opposed to flowers or trees) or that of an area of land (as opposed to the steppe and similar to the jungle). The word bush, then, can mean three things and we can see that the choice of taboo words addresses all of these three meanings. The taboo words President and George keep the players from exploiting encyclopaedic knowledge related to the surname Bush, namely that two US presidents were called George Bush. The term shrub is synonymous to the word bush in the sense of a special kind of plant, while the term outback is a synonym to bush as an area of land. In addition we see that the expression beat about is taboo. This way a possible collocational relation (based on the idiom to beat about the bush, meaning to avoid an unpleasant topic) is unavailable to the players. The butterfly card is a fairly simple card to solve since the term itself has relations to many other words and concepts. The taboo words caterpillar, insect and flower make reference to the animal itself. The words swim and nerves are not related to the insect in the strict sense. The first relates to a style of swimming that might be reminiscent of the way a butterfly moves while the second, completely unrelated to the swim sense and the insect sense (let’s for now ignore possible metaphorical relations), helps to express a feeling of nervousness or excitedness through the idiom to have butterflies in your stomach. Each of the three senses can be activated through a number of different words not listed as taboo, which makes the card a fairly simple one. The way this friend of mine solved the card was by reference to cultural background knowledge, i.e. encyclopaedic knowledge. The famous boxer Mohammed Ali was renowned for 102 5 Nodes and links - the mental lexicon <?page no="104"?> his elegance when boxing while at the same time being extremely dangerous for his opponents. It was, therefore, said of him that he floats like a butterfly and stings like a bee. Before you read on … … take a look at the four taboo cards that I asked you to create at the beginning of the previous section and compare them to the four cards from the game that we have just discussed. Which do you think are better in the sense that the five taboo words make it more difficult to explain the word on top - your own cards or those by the authors of the game? If one is better than the other, why is that? Are there any taboo words in the original game cards that you did not think of ? Can you think of a reason why that is? To conclude, I hope that this chapter has shown that there is indeed a wide range of relations between words in our mental lexicon. The more you make use of a large number of different kinds of relations when you teach new vocabulary, the better. New tools in your toolkit The mental lexicon as a network ▶ The insight into the nature of the mental lexicon as a vast network of interconnected units of different sizes suggests to teach lexis in a way that creates as many links as possible between new items and existing entries in the mental lexicon. Lexis is more than words ▶ The distinction of vocabulary and lexis draws attention to the fact that traditional views of what learners need to learn when it comes to words are insufficient. You do not teach words but lexical items, i.e. a word or larger pattern together with many different kinds of information including knowledge about co-occurrence patterns, semantic relations and associated background knowledge. Connections between lexical item ▶ You are aware of a large range of relations that might exist between lexical items and you know that each of them can contribute to the retention of new words and patterns. Diversity of connections ▶ You are aware that connections can be of very different kinds ranging from word to word and word to grammatical structure to word and semantic feature. If you want to know more A I T C H I S O N , Jean ( 4 2012).-Words in the mind: An introduction to the mental lexicon. Chichester: John Wiley & Sons. ▶ A classic about the mental lexicon. Also provides interesting descriptions of psycholinguis‐ tic experiments. A very enjoyable read. 5.4 What Taboo can teach us about teaching words 103 <?page no="105"?> K E R S T E N , Saskia (2010): The Mental Lexicon and Vocabulary Learning. Implications for the Foreign Language Classroom. Tübingen: Narr. ▶ Provides a concise overview of research and research methods on the mental lexicon. Contains a comprehensive overview of vocabulary learning and teaching. L A U F E R , Batia & I.S.P. N A T I O N (2012). “Vocabulary”. In: Susan M. Gass & Alison Makey (eds.): The Routledge Handbook of Second Language Acquisition. London: Routledge. 163-176. ▶ The paper provides an overview of vocabulary learning focussing on the questions of which words to teach, how to define and measure lexical knowledge and how new words are learned. Additional online content. 104 5 Nodes and links - the mental lexicon <?page no="106"?> 6 Think big! More than words [T]here exist a very large number of wellformed sentences which do not seem natural to the sensitive native speaker; (J O H N S I N C L AI R ) Well you could say that, but you wouldn’t! (S O M E T E A C H E R S O M E W H E R E ) What this chapter is about Much of this chapter is about raising your awareness of the fact that language use follows more or less fixed patterns, which over the last few decades has led to a new view on the nature of language itself. In addition, the chapter gives you an idea of the vast variety of the different kinds of patterns that we can find. Needless to say, these insights are relevant for the classroom and, accordingly, another focus of this chapter is to explore ways in which your knowledge of the patterned nature of language can be put to good use in the EFL classroom. Take a look at the following sentence. What do you think of it? (1) Through winter, rides between Oslo and Hammerfest use thirty hours up in a bus, though why travellers would select to ride there then might be pondered. (Hoey 2005: 5) When you look at the sentence in (1) you might feel that this sentence is somewhat odd, although you might find it difficult to say what it is exactly that seems strange. All the words are words used in English, there is nothing wrong with orthography or punctuation, and all the rules of grammar seem to have been followed as well. Still, it just doesn’t sound quite right, it would, as John Sinclair puts it, “not seem natural to the sensitive native speaker.” Similarly, I’m pretty sure you have had an experience of the following kind: you take a look at an essay that your teacher (either at school or at university) has graded and you notice the squiggly lines here and there in your text. When you ask them what is wrong here you might get an answer along the lines of “it’s not wrong in the strict sense, but we would not really say it that way.” <?page no="107"?> raising awareness of patterns Before you read on … … take the word goalless and think about what you would teach, if you had to teach your students this word. Finally, while you’re at it, what would be better ways of ‘saying it’ for the Hammerfest example above? (The solution will be provided later in this chapter.) All of the above has to do with a phenomenon that linguistics has managed to get a clear understanding of only over the last few decades due to the arrival of modern computer technology: ‘collocation’, i.e. the fact that words are a little like people; they are seen together with some words very frequently, while you will hardly or not at all see them with others. A classic example is that of the two adjectives pretty and handsome. They both mean more or less the same i.e. ‘good-looking’ and, being adjectives, would usually be used to premodify a noun. Still, as we have seen in the previous chapter, we would generally speak of a handsome man rather than a pretty man and of a pretty girl instead of a handsome girl. Admittedly, you might have been aware of this example before this book and many teachers might be as well. However, recent research has shown that in addition to collocations there is a large range of other patterns to be found in language use and, what is more, these phenomena are all-pervasive in language use. As a native speaker of your language you would usually be aware of the ways in which your mother tongue patterns, however, you would often not be able to verbalise that knowledge (as is normally the case if you are not a linguist). For example, while the first thing that comes to mind about pretty might be that it is an adjective it is actually much more frequently used as an adverb to amplify the meaning of an adjective, such as pretty dangerous, pretty harmful, pretty easy. This goes to show that frequently what we (seem to) know about how a word is used is often not the typical way in which it is actually used. Before you read on … … guess the word (1/ 6) This is the first of a series of tasks which are about identifying a word. As soon as you know (i.e. you should be certain) what the word is I’m looking for, take a mental note of how many clues you needed. Clue 1: … arrangements When it comes to raising awareness of patterns, the little goalless exercise from above can be very useful (see Hoffmann et al. 2008: 1-5). Whenever I do it in my seminars, my students would usually suggest that goalless means ‘having no purpose, no goal in life’ and that it would be translated into German as ziellos. A possible example sentence might be You’re living such a goalless life. 106 6 Think big! More than words <?page no="108"?> Before you read on … … and to see to what extent your intuition (in the sense of your knowledge of the English language) can differ from the way language is actually used, take a look at the authentic examples of goalless below. What can you see? Heart’s superiority in this- goalless affair. He played it cagey after the- goalless draw against Blackburn Rovers. … after the- goalless draw with Southampton at Anfield … Steve Pears contributed to a- goalless draw but … Roughwood played a- goalless draw at Farmers Arms and … the teams fought out a- goalless draw in grim weather at Ayresome Park Inter Milan were held to a- goalless draw at home to Sampdoria … he performed so well in the- goalless draw with Italy last month. Chester played out a- goalless game at Bristol City. Anyway the first half ended- goalless . The data set in the exercise above is taken from the British National corpus, a collection of one hundred million words of authentic English text. The corpus contains eighty-six occurrences of the word goalless neither of which is used in the sense of ‘having no goal in life’. All of the occurrences are used in the sports-related sense of ‘no goals being scored’. So, what I assume your intuition about the meaning of the word (as well as mine) was does not coincide with what we see in the data. In addition we also see a typical collocate of the word goalless, namely draw, but the patterned behaviour does not stop there: goalless draws are often with or against someone at a particular location. All in all, we can conclude that the most typical use of goalless could be described as shown under (2): (2) a/ the goalless draw with/ against X at Y The patterned nature in language is not only restricted to the co-occurrence of one word with a set of other words but often involves the co-occurrence of a word together with a particular syntactic category or construction. These syntactic restrictions can be very strict and would be considered an error if they were not followed. The two verbs to say and to tell, for instance, occur in different patterns: you say something to someone but you tell someone something, and the two patterns are not interchangeable. This is not the norm. Usually both patterns are possible with so-called ditransitive verbs, i.e. verbs that take two objects, but there might still be clear preferences. The verb to give, for instance, would usually favour a construction with two noun phrases, i.e. to give someone something. The verb to sell, in contrast, would rather choose a pattern with a noun phrase and the prepositional phrase, i.e. to sell something to someone. In both cases, however, the alternative construction would also be possible, albeit less likely. So, when we talk about the patterned nature of language this would usually not be in terms of right or wrong, but in terms of typical and untypical, or, in the words of 6 Think big! More than words 107 <?page no="109"?> the open-choice principle slots and fillers John Sinclair, in terms of what is natural and unnatural. The following two sections will explore what this means. Before we come to this, below you can find a more natural version of the Hammerfest example from above. More specifically, the original version from the beginning of Bill Bryson’s Neither Here nor There. Travels in Europe is given below together with the not so natural one from the beginning of this chapter (cf. Hoey 2005: 5). (3) In winter, Hammerfest is a thirty-hour ride by bus from Oslo, though why anyone would want to go there in winter is a question worth considering. (3’) Through winter, rides between Oslo and Hammerfest use thirty hours up in a bus, though why travellers would select to ride there then might be pondered. 6.1 Why tea can be powerful … and why it shouldn’t In the section above we have already seen that well-formedness or correctness is not the same as naturalness or idiomaticity. The idea of well-formedness is captured really well by the open-choice principle: This is a way of seeing language text as a result of a very large number of complex choices. At each point a unit is completed (a word or a phrase or clause), a large range of choice opens up and the only restraint is grammaticalness. […] It is often called a 'slot-and-filler' model, envisaging texts as a series of slots which have to be filled from a lexicon which satisfies local restraints. (Sinclair 1991: 109) It might appear a little cryptic at first but it’s actually not that complicated. It is maybe best understood if we start with the notion of a ‘slot-and-filler’ model. What does that mean? From a traditional perspective the syntactic structures of a language, i.e. phrases and clauses, have usually been analysed in separation from the vocabulary i.e. the words of the language. The syntactic part of a person’s linguistic knowledge provided rules for the structures that are possible in that language. With regard to these rules the structure of the sentence, as we will see in a minute, can be described as something that looks a little bit like an upside down tree. The syntactic component of a language provides the stem and the branches of that tree, and the vocabulary component of a language provides the leaves that are attached at the endpoint of the branches. These endpoints are the slots that Sinclair has in mind, and the leaves/ words are the fillers to these slots. 108 6 Think big! More than words <?page no="110"?> phrase structure rules or re-write rules Before you read on … … guess the word (2/ 6) Clue 1: … arrangements Clue 2: … reference to One way to describe the syntactic rules of the language are through what is called phrase structure rules or re-write rules. The second term is particularly apt since it describes what is actually going on. For example, a sentence (S) can be rewritten as a noun phrase (NP) followed by a verb phrase (VP), S → NP + VP. A noun phrase can be rewritten as a sequence of an optional (expressed through brackets) determiner (Det), followed by an optional adjective (Adj), followed by an obligatory noun (N), NP → (Det) + (Adj) + N. A verb phrase can be rewritten by a verb (V) followed by a noun phrase, VP → V + NP. With this small set of rules, the sentence structure shown in figure 5.1 can be created. Fig. 6.1: An example of a syntactic tree, the rewrite rules that created it and possible words to be filled in. As you can see, the upside down tree in figure 6.1 provides slots that still need to be filled. This is done with the help of the lexical component, an example of which you see on the bottom right corner of figure 6.1. Together, the rewrite rules and the lexical component make a grammar that can be used to create, or more technically: generate, sentences. Examples that would be generated by the set of rules in figure 6.1 are the girl drank, the dog sleeps the car, the girl drank the stone, woman gives my powerful tea, or the powerful stone sleeps my strong dog. As you can see, not all of these sentences make sense and are proper English sentences but the grammar that we see in figure 6.1 generates them. In contrast to that, sentences like the nice dog sleeps or the strong and powerful woman drank tea cannot be generated. Nice is not given in the lexical component and the structure of the NP does not allow conjoined adjectives, nor is the conjunction and part of the lexical component. 6.1 Why tea can be powerful … and why it shouldn’t 109 <?page no="111"?> sub-catego‐ risation rules By the way: Challenge accepted - or not … If you want a bit of a challenge, see if you can find out which of the following sentences are in line with the rewrite rules and the lexical component shown in figure 6.1. For those sentences that are not, where is the problem? It does not matter whether the sentence makes sense. Also, only those words are allowed that are listed in the lexical component. For instance, you can use sleeps but not slept or sleep. a) the dog sleeps the car b) my cold stone ate tea c) his car saw the girl d) the girl drank the stone e) hot tea gives car to girl f) my hot tea drinks her green girl When trying to find out whether the sentences are in line with the grammar or not, please keep in mind that it is irrelevant whether a sentence makes sense. The focus is on the structures that can be generated. Sentences that can be generated by the grammar in figure 6.1 are a), and d). Example b) is fine as far as the structure is concerned but cold is not among the list of adjectives. If the sentence was my hot stone ate tea it would be alright. The problem with c) is that his is not in the list of determiners. The structure, again, is alright. Sentence e) cannot be generated by the above grammar since the rewrite rule of the VP does not include a PP, e.g. VP → V + NP + (PP), and it does not include the class of prepositions in the lexical component. Example f) cannot be generated because the lexical component does not contain the verb drinks. If the sentence was my hot tea drank her green girl, everything would be alright because the structures in that sentence are in line with the rewrite rules. Two kinds of additional rules would prevent sentences like the dog sleeps the car or the girl drank the stone from happening. The first are so-called ‘sub-categorisation rules’. These are rules that make sure that a verb will have as many or as few constituents in a clause as it needs. For example, the verb to sleep is an intransitive verb, which means that it does not take any objects. For this special class of verbs the rewrite rule would, therefore, not show any NPs. A monotransitive verb, i.e. a verb that needs one object, would come with one NP in its respective rewrite rule, and a ditransitive verb would have two NPs, as shown in (4). (4) VP → V intransitive VP → V monotransitive + NP VP → V ditransitive + NP + NP 110 6 Think big! More than words <?page no="112"?> selectional restrictions the idiom principle semi-pre‐ constructed phrases While this refinement of our first set of rewrite rules would, thus, rule out sentences like the dog sleeps the car, it would not rule out sentences like the girl drank the stone. For this we need a second set of additional rules, which are called ‘selectional restric‐ tions’. These ensure that only those words are combined which can be combined se‐ mantically. The verb to drink, for example, should only have as a direct object a liquid, because we can only drink liquids. An adjective like educated would only go together with nouns that refer to entities that can be educated, e.g. a selectional restriction would rule out an NP like the educated stone. These two kinds of rules is what John Sinclair has in mind when he talks about ‘local restraints’, in the quote above: “slots which have to be filled from a lexicon which satisfies local restraints.” Selectional restrictions can only go so far. This becomes apparent when two words basically mean more or less the same, as for instance the word pair powerful and strong. Even if we make sure that all the sub-categorisation rules and all the selectional restrictions are followed, we might still create sentences that sound odd to the native speaker, like the girl drank the powerful tea. To repeat the quote from Sinclair (1983: 203) at the beginning of this chapter: “there exist a very large number of well-formed sentences which do not seem natural to the sensitive native speaker”. Again, recall our discussion of errors in chapter 2. If a sentence does not seem natural to a native speaker, they would usually not use it, which means that if we use it, it qualifies as an error in the sense of Lennon. So, not only should sentences be well-formed, they should also sound natural. Sinclair suggests the ‘idiom principle’ to account for that. A language user has available to him or her a large number of ‘semi-preconstructed phrases’ that constitute single choices, even though they might appear analyzable into segments. (Sinclair 1991: 110) The idiom principle emphasises three important aspects. First, instead of creating every sentence from scratch, the idiom principle sees sentences as a combination of ‘semi-preconstructed phrases’, i.e. phrases that are partly fixed and partly variable. These phrases, second, are represented as one unit in the mental lexicon of the speaker. They are just like words in that respect. Thirdly, of course, we can analyse these units according to the rules of grammar, but that does not mean that they are always created from their individual segments, when we use them. They are like already existing templates which we adapt to a given situation. Before you read on… … take a look at the utterances below and think about in which sense we can say that all of them are instantiations of semi-preconstructed phrases. Which elements are more or less fully variable, which elements can be varied within fairly tight constraints and which are fully fixed? Are there elements that are optional? a) Who do you think you are? b) Who the hell does Professor Kreyer think he is? 6.1 Why tea can be powerful … and why it shouldn’t 111 <?page no="113"?> c) Who the fuck does Peter Parker think he is? d) Who did they think they were? e) Who does this terrible woman next door think she is? Starting with the fixed elements first, we see two words that occur in all of the examples above, who and think. As for elements that are variable but only to some extent, we see that each question contains a form of the verb to do and a form of the verb to be. In front of the latter we always find a personal pronoun. This personal pronoun refers back to a noun phrase, within which we have full variability: some of these noun phrases are just realised as pronouns, others as proper names, still others as even more complex noun phrases. Examples b) and c), in addition, show expletive expressions, namely the hell and the fuck. These are obviously optional as they do not occur in the other examples. A structure that describes all of the examples above is shown in (5) (cf. Pawley and Syder 1983: 211): (5) Who (expl. expr.) D O NP i think PRON i B E ? Example (5) uses different notation conventions to express the degrees of variability in the structure above. If something occurs in brackets, this means it is an optional element, as is the case with the expletive expression. If something is in italics, it means that this word occurs exactly as it is, it is fully fixed. For instance, all possible instantiations of that structure have to use the word form think, it cannot be thinks or thought or thinking. The elements in S MALL CA P S refer to what in linguistics is called a ‘lemma’. A lemma is all the different grammatical forms in which a word as found in the dictionary can occur. For the verb to do, for example, these are to do, do, does, did, done, and doing. An element that is written in ALL CAPS refers to a grammatical class. In the case above these are the class of pronouns, which is variable but fairly restricted, and the class of noun phrase, which is fully variable and allows for an infinite number of instantiations. The little subscript ‘i’ expresses that the pronoun always refers back to the noun phrase. We see that all of the examples in the task box above are realisations of the structure described under (5). Pawley and Syder (1983: 191-92) call structures like these ‘lexicalized sentence stems’. In our context, lexicalized sentence stems can be regarded as a subtype of semi-preconstructed phrases, meaning that they consist of a given sequence of obligatory and optional elements which are fixed or variable to different degrees. The structure under example (5) would, therefore, satisfy the first criterion that Sinclair posits with regard to the idiom principle. 112 6 Think big! More than words <?page no="114"?> single choice analysable into segments formal aspects polywords Before you read on … … guess the word (3/ 6) Clue 1: … arrangements Clue 2: … reference to Clue 3: … enquiries The second criterion demands that these semi-preconstructed phrases constitute single choices, similar to the single choice of the word dog when we want to refer to this kind of animal or similar to an utterance like Hi, how are you? which we pick as one single unit when we want to greet someone. In which sense could the structure under (5) be understood as a single choice? The answer to that question lies in whether we can find a function that this structure fulfils, just like Hi, how are you? functions as a greeting. The function of example (5) is to express someone’s anger at the person(s) referred to in the noun phrase. For this case the English language provides this whole structure as one single choice. You do not create that structure from scratch with the help of the rules of syntax but you pick it as one unit and fill in the details to make it fit the given situation. That is what Pawley and Syder (1983) mean when they describe such struc‐ tures as ‘lexicalized’, i.e. as “regular form-meaning pairings” (192). This satisfies the second criterion of the idiom principle. The third criterion is simple, as the structure conforms to the general rules of ques‐ tion formation including all the elements that we need to form a question in English. So, the structure under (5) is analysable into segments, but, and this is really important, we do not analyse it into segments, it is one entry in our mental lexicon, much like a single word. The idiom principle helps us understand what makes language use natural. As we have seen, it places emphasis on linguistic units that consist of more than one word but still are stored as one mental unit in the language user’s mind. In the next section we will explore the many different forms such units can take. 6.2 Lexical phrases The concept ‘lexical phrases’ was introduced by Nattinger and DeCarrico (1992). A lexical phrase is a pattern of language use that fulfils a particular communicative func‐ tion. Nattinger and DeCarrico distinguish four formal categories of such patterns. They are distinguished on the basis of their size and the variability they allow. The first formal class of lexical phrases is ‘polywords’. The term itself is interesting and, in my view, well-chosen. It expresses that, basically, polywords should be treated or understood as one single word even though they consist of more than one word. Polywords are short phrases of usually around three or four words that do not allow any variability and do not allow any interruptions. An example would be in a nutshell, which has the function 6.2 Lexical phrases 113 <?page no="115"?> institutional‐ ised expres‐ sions phrasal constraints sentence builders of introducing a summary of what has been said before. Notice that this function of in a nutshell cannot be guessed from the meaning of the individual words. So, teaching your students that the English word nutshell translates as German Nussschale might make sense, but your students will benefit a lot more, if you teach them how it is used idiomatically in the phrase in a nutshell. Another example of a polyword would be by and large. It serves as a qualifier and expresses that something is true generally even though there might be rare instances when it is not. The next group of lexical phrases are ‘institutionalised expressions’. Like polywords they cannot be varied and are usually continuous, however, they are much longer and constitute a whole utterance, such as How do you do? as a greeting formula or Give me a break! as a way to object to something. Again, the latter example makes clear that the meaning and the usage of this phrase has to be learned independently and cannot be derived from the meaning of its individual parts. Before you read on … guess the word (4/ 6) Clue 1: … arrangements Clue 4: … an impression Clue 2: … reference to Clue 3: … enquiries While the previous two formal categories do not show any variation with regard to their wording, the following, both, have variable slots. ‘Phrasal constraints’ are similar to polywords in that they are rather short. The variable string ‘as far as I ___’, for example, might be realised as as far as I can tell or as as far as I know but in all cases functions as a qualifier. The phrasal constraint ‘see you ___’ is a useful phrase when parting from someone and might be instantiated as see you, see you later or see you next week. The lexical phrases of the last category, ‘sentence builders’, are similar to the previous ones in that they allow for variation but they differ in that they provide a framework for full sentences. The lexical phrase ‘it’s only in X that Y’ and an instantiation like it’s only in linguistics that you will find true fulfilment is a case in point. It serves as a qualifier and expresses that true fulfilment cannot be found anywhere but only in lin‐ guistics. The lexical phrase ‘let me start with X’ would be used to introduce a new topic, as in let me start with collocations. As you can see, each of the ‘completed’ sentence builders are, in fact, complete sentences, hence the term. Table 6.1 provides an overview of the four formal categories discussed above. 114 6 Think big! More than words <?page no="116"?> functional aspects social inter‐ actions Pattern type Description Example polyword institutionalised expression phrasal constraint sentence builder a lexical phrase of around 3 to 4 words that is continuous and shows no variability a continuous and invariable lex‐ ical phrase that constitutes a whole utterance a short lexical phrase that is mostly continuous and contains (usually) one variable slot a lexical phrase that can often be discontinuous and contains one or two variable slots; it serves as a framework for a whole sen‐ tence by the way so far so good in essence How do you do? Nice meeting you be that as it may a ___ ago (week/ month/ year/ …) as I was ___ (saying/ mentioning/ …) as far as I ___ (know/ can see/ …) I think that X Let me start with X not only X, but also Y MODAL VERB you VP for me? Table 6.1: An overview of formal kinds of lexical phrases (based on Nattinger and DeCarrico, 1992: ch. 2). In addition to these four formal categories, Nattinger and DeCarrico also focus on the functions in which lexical phrases are used in authentic discourse. It is because of these that they can become useful “as practical instruments for the classroom” (59). They distinguish three functional groups. For example, lexical phrases can be used to organise social interactions, such as how to say hello and goodbye, how to start and end a conversation, how to shift a topic, how to make sure that the intended meaning came across, how to ask for something or offer something and so on. Table 6.2 provides some examples of that category. Most of these will not come as a surprise, and maybe you know some of them and have even been taught them as useful formulae. The important point, here, is that we start to see these lexical phrases as single (sometimes partly flexible) units that are stored as such in the native speaker’s mind together with their respective functions. They are not created from scratch but picked as a whole and adapted to the given situation. 6.2 Lexical phrases 115 <?page no="117"?> necessary topics Function Description Example summoning getting someone’s attention Hey/ Hi/ Hello, (Name), how are you? Hi, I think we haven’t met yet topic establishing introduces a topic (X) into the discourse Have you heard about X? What do you think about X? topic shifting shifts to another topic (X) in the discourse Oh, that reminds me of X This might be a bit off topic, but what about X? clarifying asking someone to clarify something What did you mean when you said X? Could you repeat that again, please? making sure you have been understood So, what I’m trying to say is X Let me repeat: X taking a turn reserving the right to speak, taking the floor Excuse me, please. Might I just say something here? requesting asking someone to do some‐ thing for you Would you mind doing X? May I ask you to do X? offering offering to help someone Can I help you? Would you like X? refusing refusing to do something you are asked for Sorry, I’d rather not I’m sorry but I think that X asserting expressing that you believe something is the case I think/ believe that X I heard (somewhere) that X disagreeing Expressing that you do not share someone’s view Yes, but I think that X I don’t really agree with you/ what you’ve said/ X Table 6.2: An overview of selected functions of lexical phrases related to social interaction (based on Nattinger and DeCarrico 1992: 60-63). The second category is that of necessary topics. Here, lexical phrases are used to “mark topics about which learners are often asked, or ones that are necessary in daily con‐ versations” (63). These topics include information about the speaker, things they like or do not like, times and locations. Take a look at table 6.3. Again, some of these will look familiar, but as before, the point is that we start to see them as one single unit. Consequently, they also should be taught as one single unit. This is already done to some extent in many textbooks, e.g. in sections like ‘Useful expressions to talk about what you like or dislike’ or ‘Going to the restaurant’. 116 6 Think big! More than words <?page no="118"?> discourse devices Function Description Example autobiography talking about facts about your life my name is X I’m X years old I’m from X language talking about the language(s) you or someone speaks do you speak X? I don’t speak X very well. how do you say X? quantity talking about quantities and amounts how much is X? that seems a bit much for an X? lots of X time talking about time what time X? a(n) X ago in YEAR location talking about locations where is X? to the right/ left of X? how far is X (from here)? likes and dislikes talking about things you like I like X (a lot) X is lots of fun talking about things you do not like I don’t like X that much I’m not too keen on X shopping and food asking for prices, for what’s on offer, for the bill etc., order food, … how much is X? do you have X? the bill, please! Table 6.3: An overview of selected functions of lexical phrases related to frequent topics in conversa‐ tions (based on Nattinger and DeCarrico 1992: 63-64). I assume you know most of the above expressions and I also assume that you have already used them frequently. This shows that at least some of them are stored in your mental lexicon as a single unit. For instance, while the lexical phrase ‘my name is X’ could also be a word-by-word translation from German, the simple ‘I’m from X’ cannot be explained that way, since a translation would usually yield I come from X. So, somewhere in your mind there is a link from the communicative function of talking about your place of birth or residence to the lexical phrase ‘I’m from X’. Similarly, almost any German learner of English at some point in their life will say something like that makes a lot of fun, since this is what a verbatim translation from German would result in. It is only when learners know the relevant lexical phrase ‘X is lots of fun! ’ that they will not make this mistake anymore. To give a last example, German learners of English might say in the year 2022, because they do not yet know that the correct lexical phrase in English would just be in 2022. The last functional category the authors suggest is that of discourse devices. Lexical phrases of this kind are used to organise discourse, such as establishing logical or tem‐ poral connections, introducing examples or summarising what has been said. These 6.2 Lexical phrases 117 <?page no="119"?> kinds of lexical phrases are a little more advanced than the two types discussed before. See table 6.4. Function Description Example logical connection establishing a logical link between two aspects, such as causation or condition because (of) X Y unless X no Y in spite of X exemplifier introducing an example it’s like X for instance, X to give you an example, X qualifier modifying the degree to which a statement is valid or to which you are certain about its validity it depends on X it’s only in X that Y it might be the case that X evaluator making a statement about the truth or falsehood of a claim as far as I know X there’s (absolutely) no doubt that X I’m not quite sure but X summarizer introducing a summary of aspects of the previous discourse my point (here) is that X summing up, X in a nutshell, X Table 6.4: An overview of selected functions of lexical phrases related to discourse devices (based on Nattinger and DeCarrico 1992: 64-65). The above is not by any means a comprehensive list of lexical phrases. Rather, the examples are meant to, generally, make you aware of lexical phrases and to give you a feeling of their pervasiveness in everyday discourse. In addition, we have seen that they are relevant for the classroom because all of them fulfil an important communicative function. In the next section we will see that an awareness of the patterned nature of language and having a large range of patterns available at any time is also beneficial for other reasons. 6.3 Patterns and their relevance for the classroom Firth’s claim that we know words if we know the company they keep can also be interpreted with a teaching perspective in mind: You know only very little about a word if you know its general meaning. What you need to know on top of that is how the word is typically used. This becomes particularly apparent with (near-)synonyms. Large and tall are very similar with regard to their meaning and there are things that can be both, large and tall, such as a tree or a man, but while we talk about large sums of money we would never use the adjective tall in this context. If you are not yet convinced, maybe the next example will be helpful. The German phrase deine Hausaufgaben machen would translate word-by-word to to make your homework, which 118 6 Think big! More than words <?page no="120"?> the lexicon is not arbitrary fluency in comprehen‐ sion and production complex ideas are expressed lexically would be wrong as in English you do your homework. So, teaching your pupils that to make is translated as machen in German provides them with a little bit of information, but it is far from enough to put them in the position where they can use the word appropriately. Rather, you should teach every word together with the words, phrases and structures it typically occurs with. While an unawareness of specific patterns may result in incorrect language use (as is the case with to make your homework), often the result will merely sound strange, it will be something a native speaker would not use. Pawley and Syder (1983: 193) claim: “[…] native speakers do not exercise the creative potential of syntactic rules to anything like their full extent, and […] indeed, if they did do so they would not be accepted as exhibiting nativelike control of the language.” Even worse, if you recall our discussion from chapter 2, this would count as an error. It is only if you use the right collocations that you will sound natural and idiomatic. In short you should make sure that your pupils develop what Hill (2000) refers to as ‘collocational competence’. In addition to sounding idiomatic, your students will benefit from this in many other ways, as we will see presently. Before you read on … … guess the word (5/ 6) Clue 1: … arrangements Clue 4: … an impression Clue 2: … reference to Clue 5: … amends Clue 3: … enquiries We have already seen that “[t]he lexicon is not arbitrary” (Hill 2000: 53), i.e. when it comes to idiomatic language use, it is not the case that anything goes. Some words like to occur with particular other words, phrases or structures and your teaching should reflect that. The fact that the lexicon is not arbitrary makes language predictable to some extent. When you hear the word handsome a good guess of what follows would be man. If an English-speaking parent asked their child “Your homework, have you …”, you know the next words will be done it? Obviously, this increases the speed and the ease of processing. Hill even claims that “[o]ne of the main reasons the learner finds listening or reading difficult is not because of the density of new words, but the density of unrecognised collocations” (2000: 54; emphasis in the original). In addition to improving processing, a large stock of language patterns will also improve fluency in production. Just like it makes under‐ standing faster if you can work in patterns instead of word by word, production will become faster if you can rely on those patterns: “[…] the use of formulaic sequences accounts for a great proportion of natural speech and nativelike fluency” (Götz 2013: 32). The above becomes particularly obvious if we understand that “[c]omplex ideas are often expressed lexically” (Hill 2000: 55). For example, a student might come up with 6.3 Patterns and their relevance for the classroom 119 <?page no="121"?> saving brainspace too much choice a non-idiomatic sentence like his disability will continue until the day he dies because he or she lacks the collocation permanent disability, which would enable the student to produce the much more idiomatic he has a permanent disability (49-50). Instead of having to create a long and complicated sentence word by word, the student can pick ‘to have a permanent disability’ from their stock of patterns and adapt it to the current situation. Not only would this result in faster production but it would also decrease the likelihood of errors. If your students do not have to use complex structures and complex wording, there is just less opportunity for things to go wrong. Before you read on … … take a look at the two examples below which express complex ideas in complex sentences. Can you find a simpler, more lexical way to express the same content? (a) I was going to the bus stop but when I got there, the bus had already gone. (b) My father - he’s getting better - he had a big operation. (Hill 2000: 64) Maybe the best lexical ways of expressing the same contents would be I missed the bus and my father is recovering from surgery/ a major operation. Each of these involve a collocation, namely to miss the bus and to recover from surgery/ an operation, respectively and your students would be at an advantage if they had these at their disposal. Not only do the examples in the task box sound cumbersome and unnatural, but they are also more error-prone because they are so much more complex than their collocation-based counterparts. Related to that is the fact that patterns “make thinking easier”. They allow us “to name complex ideas quickly so that we can continue to manipulate the ideas without using all our brainspace to focus on the form of words” (55). A last reason for teaching collocations and patterns is the problem of too much choice, as formulated by Wray (2002: 206): Where a native speaker would draw an expression whole from the store of formulaic sequences, the non-native has to compose it out of individual words, and this offers too much choice: there are plenty of comprehensible and grammatical ways of expressing an idea, but only those combinations which are stored formulaically by native speakers will be received as idiomatic […]. Wray illustrates this problem with the collocation major catastrophe. Let us assume your students want to translate ein schreckliches Unglück into English. Without an awareness of patterns and collocations they would usually work on a word-by-word basis. Looking up schrecklich in a dictionary, they might find words like dreadful, terrible, unholy, frightful, grim, horrible, horrid and shocking. As possible translations for Unglück the dictionary might yield disaster, bad luck, suffering, misfortune, trouble, mishap and catastrophe. Being unaware of collocations and assuming that anything 120 6 Think big! More than words <?page no="122"?> lexical approach a new view on the men‐ tal lexicon lexico-gram‐ mar goes, they end up with the vast range of 8 x 8 = 64 possible combinations, only some of which, such as major catastrophe or terrible disaster, might actually be possible. Having (hopefully) established the relevance of patterns for the classroom we will conclude this chapter by taking a look at how to make your students aware of these patterns and how to teach them. 6.4 Teaching patterns Much of the work and research with regard to taking patterns seriously for the EFL-classroom has occurred in the context of the so-called ‘Lexical Approach’ to for‐ eign language teaching. It involves a shift away from the commonly assumed, tradi‐ tional view on how sentences are created, as for instance formulated in the open-choice principle. Recall that according to this approach sentences are the result of a two-step process. First, the language user forms an abstract sentence frame on the basis of syn‐ tactic rules, which, second, is filled with words from the language user’s mental lexicon. That is grammar comes first, lexis comes second. Lewis (1993: 89) advocates a different order, putting lexis first: […] language consists of grammaticalised lexis, not lexicalised grammar. Lexis is the core or heart of language but in language teaching has always been the Cinderella. Fundamental to the lexical approach is a reconceptualisation of the mental lexicon. Traditionally, the mental lexicon is the store of the words that the speaker knows, the so-called ‘vocabulary’. ‘Knowing a word’ in this approach means, basically, knowing the meaning, the spelling, its pronunciation and some necessary syntactic features, such as the word class or the number of obligatory constituents it takes. Proponents of the lexical approach would argue that the mental lexicon contains a lot more than that. The individual words are still at the basis but the information connected to these words is far more comprehensive. On top of what traditionally would be stored with each word, the lexical approach would assume that the mental lexicon would also hold in‐ formation on the collocates of the words and all the different kinds of patterns that the word occurs with, including everything that we have discussed above. In addition, the mental lexicon would also contain units that are larger than the word, such as the different kinds of lexical phrases. Instead of the traditional ‘vocabulary’ as a store of words, the lexical approach talks of ‘lexis’ as a store of lexical items which includes individual words as well as larger semi-prefabricated units consisting of more than one word and variable slots. With this understanding of the nature of lexical items, some syntactic information becomes part of the lexical item. While the traditional understanding of language draws a sharp distinction between grammar and vocabulary, the lexical approach rather sees a continuum of ‘lexico-grammar’, as shown in figure 6.2. 6.4 Teaching patterns 121 <?page no="123"?> Fig. 6.2: The traditional conception of grammar as separated from vocabulary and the concept of lexico-grammar. Against the background of this view, utterances are not usually created by first forming an abstract sentence frame on the basis of syntactic rules, which is then filled with words from our lexicon. Words and syntactic structures are co-selected and the sentence is the result of interlocking larger lexical items. Take a look at the sentence under example (6) together with the list of lexical items it consists of. (6) Prof. Kreyer strongly opposes the widely held view that linguistics is useless in comparison to literary studies when it comes to teaching English at school. a. sentence builder: X opposes Y b. collocation: oppose + strongly c. phrasal constraint: the view that X d. collocation: view + widely held e. sentence builder: X is ADJ in comparison to Y f. phrasal constraint: when it comes to doing X g. phrasal constraint: do X in/ at Y The whole sentence is an instantiation of sentence builder a. The main verb to oppose is modified and instantiates collocation b. The Y of a is instantiated by the phrasal constraint c, where, again, a collocation, d, serves as the basis for the premodification of the noun phrase head view. The X in the phrasal constraint c is realised by sentence builder e. The phrasal constraint f provides circumstantial information for the situation described by sentence builder e. Phrasal constraint g is a possible realisation of ‘doing X’ in f. Patterns of this kind do not just exist in made-up example sentences living in the minds of linguists. Take a look at the conversation between Stephen Colbert and Barack Obama (www.youtube.com/ watch? v=zhnPAHQFs_g; 1: 20 - 2: 20) in table 6.5. The left-hand column shows the conversation, the right-hand column provides the patterns that can be found. 122 6 Think big! More than words <?page no="124"?> the impor‐ tance of awareness Example Pattern SC: Well, president Obama, thank you so much for joining me today. Thank you (so much) for X BO: It is wonderful to be here. Thank you for having me. It is wonderful to do X Thank you for having me SC: It’s good to see you. It’s good to see you. BO: It is good … It is good to be seen. I’ve been seeing you on television, but it’s good to see you in person. to see someone somewhere on television to see someone in person SC: That’s nice. How old are you? Are you … Are you… 18 and 54. BO: I’m not the demographic you are looking for. […] X BE not the Y Z BE looking for Michelle says hey. X SAY Hi/ Hey SC: Oh, good. Good. Good. BO: You know she loves you. X KNOW (that) Y SC: Well, I’ve really enjoyed spending some time with her over the last four years. X ENJOY Y X ENJOY doing Y X SPEND time with Y over + [period of time] BO: I know. And she adores you. She thinks the world of you. I think you’re okay. to think the world of X SC: Can we just take a moment… Can I just… I don’t want to talk. I just want to take a moment to drink you in for just a moment, because I’m having to get used to looking at the president… take a moment to do X drink X in just for a moment to have to get used to doing X BO: laughs SC: … again. You know, I’ve gotten out of… I’ve gotten out of the habit. I’ve got to warm up for Joe Biden. I don’t want to pull anything when I see him take the Oath of Office. You have to ease me into this a little bit. to get out of the habit to warm up for X to pull something/ a muscle to see X do Y to take an oath to ease X into Y Table 6.5: Examples of patterns in authentic conversation. When I do exercises of this kind with my students, they are usually astounded by the large number of patterns that can be found. Admittedly, identifying patterns is not trivial but with a little bit of practice it is doable. An important first step in this respect and when it comes to teaching patterns is raising awareness of the existence of patterns and of the pervasiveness of them in authentic discourse. Unfortunately, this will not happen automatically. More than 40 years ago, Stephen Krashen has drawn our atten‐ 6.4 Teaching patterns 123 <?page no="125"?> input is not intake noticing awareness through near-syno‐ nyms awareness through translation tion to the distinction between input and intake (e.g. 1981: 46). Input is any language material that the language learner encounters in the course of their life. If input was all it took to learn a new language, all English teachers by now would be out of the job, since schoolchildren and university students have unlimited access to authentic input through YouTube, Netflix and so on. Usually, input just washes over the language learner and does not automatically increase foreign language competence. Input needs to become intake for learning to happen. A pre-requisite for that is that learners notice a particular word, a particular pattern, a particular structure, a particular rule and so on. When it comes to patterns in language use, noticing is difficult even for advanced learners at university level and there is often no guarantee that it will happen. However, teachers can create favourable conditions that increase the likelihood of noticing. As said above, the first step is to raise awareness of the fact that authentic language use does actually show patterns of various kinds. Often, providing students with some examples will be helpful, similar to what I have done in this chapter. A good way in, from my experience, is to take a look at near-synonyms, i.e. words that mean more or less the same, such as pretty and handsome or tall and large and help your students notice, that they differ with regard to the words that co-occur with them. We have also seen above that near-synonyms might pick different syntactic structures, as is the case of to tell someone something as opposed to to say something to someone. Another possible way in is through (problems of) translation. It can be helpful to choose patterns of the L1 that, if translated word by word, will result in an L2-expres‐ sion that is either incorrect or not quite idiomatic. An example of the first is that of German Hausaufgaben machen and English to do your homework, an example of the latter is German die Zähne putzen, which would usually be translated as to brush one’s teeth even though the much less frequent to clean one’s teeth would be possible. Inter‐ estingly, German die Zähne reinigen would only be used in the context of professional dental hygiene. As a last example consider prepositions, which often show differences in their patterns of co-occurrence: while both German and English use the preposition in in the context of newspaper, the two languages differ when it comes to TV: etwas im Fernsehen sehen as opposed to to see something on TV. Before you read on … … guess the word (6/ 6) Clue 1: … arrangements Clue 4: … an impression Clue 2: … reference to Clue 5: … amends Clue 3: … enquiries Clue 6: … up your mind By now you will most likely have guessed that the word that I was looking for is the verb to make: you make arrangements, you make reference to, you make enquiries and so on. A verb like to make lends itself particularly easily to an exercise of the above 124 6 Think big! More than words <?page no="126"?> awareness through pre‐ dictability awareness through per‐ vasiveness recognising patterns through translation kind since it is what is called a ‘light verb’. Light verbs are verbs that only have a very general and usually very vague meaning, such as to give, to do, to take and so on. Because of their little semantic content, they are quite versatile and can occur in a large number of collocations and patterns. Usually, these patterns cannot be translated word by word into another language as you can see in table 6.6. make arrangements make reference to make enquiries make an impression make amends make up your mind Vorkehrungen treffen auf etwas Bezug nehmen Nachforschungen anstellen einen Eindruck hinterlassen etwas wiedergutmachen sich entscheiden Table 6.6: Collocations with to make and their German counterparts. Another way to make learners aware of the patterned nature of language is through focusing on the predictability of language, as has been done in this chapter with the ‘guess the word’ exercise. Good starting points in that respect would be combinations of verb + noun as above or adjective + noun, such as heavy which has a set of typical collocates including rain, metal, traffic, drinking, breathing, weight or burden. Obvi‐ ously, there are a lot more approaches that you can take from a didactic perspective, but the goal here merely is to give you a first idea of how awareness of the patterned nature of language can be raised. The next step would be to make your students aware of the pervasiveness of patterns in language use. An obvious method, of course, is to present your students with an authentic text and show them how many patterns you can find in this text, similar to what I did in table 6.5 above. Usually, students are surprised when they first see an example of that kind and it does not take much more (if anything at all) to convince them that patterns really are everywhere in authentic language use. In a next step, you could provide your students with a text and send them out to ‘hunt’ for patterns themselves. As said above, recognising patterns is not trivial. An easily applicable first criterion that might help to identify patterns, again, is that of translation. Possible guiding question would be the following: If you think something might be a pattern, think about how you would get the same meaning across in German. 1. Would you use a completely different sentence altogether? 2. Would you use structures that do not show in the English pattern? 3. Would you use words that are not a translation of the words in the English pattern? The first question would be particularly useful to identify larger patterns. ‘X S AY Hi’, for instance, would qualify as a pattern since the usual way to express the same in German would be Viele Grüße von Michelle or Michelle lässt Dich grüßen. The verbatim translation Michelle sagt Hallo is still rather rare but is found in badly dubbed American or British TV series and movies. Thank you for having me would be translated as Vielen 6.4 Teaching patterns 125 <?page no="127"?> recognising patterns through function Dank für die Einladung and she thinks the world of you would translate to something along the lines of Sie findet Dich ganz toll, Sie hält viel von Dir, or Sie hält große Stücke auf Dich. Accordingly, both ‘Thank you for having me’ and ‘to think the world of X’ would qualify as patterns. The second question would also serve to identify larger patterns. An example is ‘Thank you (so much) for X’ since the German counterpart has a different structure with regard to the position of the intensifier, e.g. ‘Vielen (herzlichen) Dank, dass …’. Another structural difference would be the complementation by ‘for + ing-clause’ instead of a subordinate clause introduced by dass in German. Similary, ‘X E N J O Y doing Y’ is noteworthy as the German counterpart would suggest a to-infinitive clause. The third question would primarily identify collocations, as in the already discussed difference between on TV and im Fernsehen or in the contrast between ‘to drink X in’ as opposed to ‘X in sich aufsaugen’. Other examples in the conversation between Obama and Colbert are the differences between to pull a muscle and sich einen Muskel zerren or between to take an oath und einen Eid schwören. Another useful criterion that makes reference to the concept of lexical phrases is that of function. It asks the question whether a string of words might be particularly useful in a certain situation. Thank you for having me will come in handy when you meet the host of a party you have been invited to. The idiomatic way to say that some‐ one says ‘Hi’ is to use the pattern ‘X S AY Hi’ (and not ‘X greets you’ or ‘many greetings from X’) and ‘over + [period of time]’ is one idiomatic way to express that something has happened for a period of time. The final step would then be to let your students discover collocations and patterns on their own. Obviously, your students will spend a lot of time with authentic English language when listening to music, when reading English books and when watching clips and films on Youtube, TikTok, Netflix and other platforms. If they used the potential hidden in that vast amount of authentic material, their stock of patterns would increase dramatically over the years. If they only acquired one new pattern every day from grade 5 to grade 13, this would result in the staggering number of more than 3000 useful patterns. Again, this needs some guidance, particularly at the beginning. Even though your students will have a good understanding of the nature of patterns if you have done the steps described above, they might still be overwhelmed if you tell them to find one pattern a day in the media they consume. Light verbs are one good option. For instance, you might ask your students to write down all the things that can be made for two weeks (make a mistake, make friends, make an effort), or all the things that can be given (give a chance, give advice, give access). Similarly, light nouns and the question of what verbs precede them can be interesting. For example, you spend, waste, devote, save time. With increasing awareness of the patterned nature of language you can, then, encourage your students to become freer and let them note instantiations of abstract combinations, such as verb + noun (give presentation, guarantee safety, question loyalty) , adjective + noun (huge amount, terrible accident, major problem) , noun + verb 126 6 Think big! More than words <?page no="128"?> (accident happen, problem arise, situation deteriorate), adverb + adjective (particularly important, completely different, hardly surprising) and so on. With all of these, students should be encouraged to search for as big a pattern as possible. For example, the collocation make + enquiries comes with a number of larger patterns of complementa‐ tion, such as ‘ MAK E enquiries about NP’, ‘ MAK E enquiries as to NP’, ‘ MAK E enquiries as to wh-clause’ or ‘ MAK E enquiries with regard to NP’. The collocation major problems comes in interesting patterns like ‘major problems faced by/ facing/ for X’, where X is the one affected by the problems and ‘major problems in/ with X’, where X describes the area in which the problems occur. The steps described above are summed up in the figure 6.3. Fig. 6.3: Possible steps to increasing your students’ pattern noticing ability. Obviously, the suggestions above are not exhaustive and I am certain that you can come up with a large range of other ways to 1) make your students aware that language use is patterned, 2) make them aware that these patterns are everywhere, and 3) help them find ways in which they themselves can create and increase their stock of readily available patterns. Whatever it is you do, it is important that we can address the fundamental problem so aptly described by Wray (2002: 206): Where the first language learner starts with large and complex strings, and never breaks them down any more than necessary, the post-childhood second language learner is starting with small units and trying to build them up. […] The result is that the classroom learner homes in on the individual words, and throws away all the really important information, namely what they occurred with. 6.4 Teaching patterns 127 <?page no="129"?> The quote makes clear that we need to get back to looking at larger units and we need to help our students do this as well. One useful tool in that respect is a corpus, as we will see in the next part of this book. New tools in your toolkit Open-choice vs. idiom principle ▶ Your understanding of these two fundamentally different approaches to view language enables you to better understand that natural language to a large extent consists of more or less prefabricated patterns that native speakers use regularly. Lexical phrases ▶ You now have an even better idea of what patterns of language use can look like. Their focus on functions make lexical phrases particularly useful for fostering communicative competence in your students. Relevance of patterns ▶ You now know many respects in which patterns make language processing and production easier and more native-like. The problem of too much choice ▶ Your awareness of this problem stresses the importance of thinking bigger than the word when it comes to teaching lexical items. Whenever possible support your students to store larger expressions in their mental lexicon rather than try and create them from individual words. Lexico-Grammar ▶ Your knowledge about this concept helps you to appreciate that your mental lexicon contains units that combine lexical information with grammatical informa‐ tion. Words and grammar are more linked and interdependent than traditionally assumed. Noticing patterns ▶ In the present day and age you and your students have access to an unlimited amount of authentic language data. Since you now know that input does not automatically become intake you are aware that you need to help your students notice collocations and other patterns. The suggestions made in this chapter can help to increase your own noticing abilities and, more importantly, it can help you help your students. In addition, it can serve as a starting point for you to find better ways of supporting your students. If you want to know more B O E R S , Frank & Seth L I N D S T R O M B E R G (2009): Optimizing a Lexical Approach to Instructed Second Language Acquisition. London: Palgrave Macmillan. ▶ The monograph explores important aspects when it comes to putting the Lexical Approach into practice. These include which patterns to choose for the L2 classroom and how to make sure that they are retained in memory. C A R T E R , Ronald ( 2 2012): Vocabulary. Applied Linguistic Perspectives. London: Routledge. ▶ Provides a thorough introduction to the study of lexis, from words to patterns and discourse. Also provides a chapter on learning and teaching vocabulary. 128 6 Think big! More than words <?page no="130"?> L E W I S , Michael (1993): The Lexical Approach. The State of ELT and a Way Forward. Boston, MA: Thomson Heinle. ▶ A seminal publication that triggered a new way of thinking about teaching the English language. W R A Y , Alison (2002): Formulaic Language and the Lexicon. Cambridge: CUP. ▶ An excellent monograph devoted to formulaic sequences. In addition to formulaicity in adult language, first language acquisition and aphasic language, Wray also devotes a substantial part of her monograph to formulaic sequences in a second language. Additional online content. 6.4 Teaching patterns 129 <?page no="132"?> Abstract Contents 7 The theory of practical corpus analysis 7.1 The corpus as a collection of authentic language 7.2 Words and their contexts 7.3 More than words - annotation provided in corpora 8 A more practical introduction to corpus analysis 8.1 Wild cards and corpus queries 8.2 Corpus queries - exploiting lemmatization and tagging 8.3 Corpus queries - collocations 8.4 Corpora in the classroom 8.4.1 Benefits and opportunities 8.4.2 Problems and pitfalls The previous part has emphasised how important it is to think big when it comes to teaching vocabulary and patterns. This part introduces corpus linguistics as a very helpful tool to explore the patterned nature of language. It shows how corpora can give us insights into how language is actually used and it suggests ways of using corpora in the EFL classroom. The corpus in the classroom The language looks rather different when you look at a lot of it at once. (J O H N S I N C L AI R ) <?page no="134"?> why use a corpus? 7 The theory of practical corpus analysis Every corpus I have had the chance to examine, however small, has taught me facts I couldn’t imagine finding out any other way. (C H A R L E S F I L L M O R E ) What this chapter is about The chapter heading might sound a little strange and, maybe, even contradictory. However, a thorough understanding of theoretical aspects of corpora and their design is extremely useful if you want to do your own corpus analyses. You can be much more specific and efficient in your work with any corpus if you base your queries on your linguistic expertise and on a general understanding of what a corpus is and of the kind of information that a corpus may contain in addition to the mere words that you can find in it. This way, this chapter prepares the ground for the more practical chapter 8. Corpora (the plural of corpus) are large collections of authentic text that are used for linguistic analysis. Corpus linguistics, i.e. the branch of linguistics that bases its analyses on such large text collections, led to a real paradigm shift: A language does indeed look very different when you look at a lot of it at once. For instance, corpus linguistics helped us understand that language, to a very large extent, is driven by patterns that are stored as semi-prefabricated units inside the minds of the language users (see section 5.3.4 and chapter 6). In this chapter and the next chapter we want to explore ways in which corpora can be useful for you as a student of the English language and as a future teacher of English. You might ask yourself why you should bother working with corpora in the first place. After all, in this modern day and age you have all kinds of resources, e.g. Youtube or Netflix, at your fingertips just one mouse click away. You have access to a range of very good online dictionaries, both monolingual and bilingual. Some of these even provide a fairly large number of example sentences taken from authentic language use. On top of that the web now even offers collocation dictionaries, which give you the basic facts about how a given word is used. All of these are truly excellent resources and the question “Why should I bother learning how to use a corpus? ” is a valid one. <?page no="135"?> incidental learning providing convincing answers a corpus for teaching and learning corpus representa‐ tiveness In my opinion there are three main reasons. First, even though modern online re‐ sources are really good, the content they provide is always the result of choices some‐ one else has made. This means useful and interesting information that can be found in the original data may be ignored. For instance, let us assume one of your students produces a sentence like She is married with Peter. As a competent speaker of English you, of course, know that the correct preposition is to, and you ask your student to find the correct preposition in a (collocation) dictionary, which they will be able to do with‐ out any problems. However, what the dictionary will (most likely) not tell them is that the preposition with does occur with married in sentences like she is married with three sons. This is an example of incidental learning, where you learn something in addition to what you were looking for. A corpus provides endless opportunities for this kind of learning, both, for you and your students. The second reason relates to giving your students convincing answers to questions regarding the use of English. We have all had the experience that our teacher told us something along the lines of “Yes, you could say that, but I wouldn’t. Just doesn’t sound right.” I, personally, never thought this was particularly convincing. In addition, being a (university) teacher myself, I am now very glad that I can use corpus data to show my students that, even though their language use is not incorrect, native speakers would have put it differently. Finally, a corpus can be used as a resource for teaching and learning, i.e. you can use a corpus to create authentic learning materials and you can even provide your students with access to corpora. It does not take much to get a grasp of the available corpus resources and with only a little expertise on how to search a corpus, corpora can become an invaluable resource both, for you and your students. This chapter is about discovering the potential of that resource. It first makes sense, of course, to take a look, in the next section, at what corpora are and what kind of information they contain. After that we will explore how we can use this information to find out what authentic English looks like. The chapter will conclude with opportunities and problems regarding the use of corpora in the classroom. 7.1 The corpus as a collection of authentic language A corpus is a large collection of authentic texts that can be used for linguistic research. Corpora (that’s the plural) today are usually computerised, which is a good thing when it comes to analysing them. They are usually representative of something, i.e. they are a sample that stands for a whole. For one research article I compiled a hip-hop corpus of US-American hip-hop. While the language represented in this corpus, obviously, is authentic English, it does represent a particular kind of language use and, generally, a particular variety of English, often referred to as AAVE or African-American Vernac‐ ular English. The Corpus of American Soap Operas (Davies 2011-; available online at www.english-corpora.org/ soap), to give another example, contains 100 million words of ten American soap operas from 2001 to 2012. If you were interested in informal 134 7 The theory of practical corpus analysis <?page no="136"?> spoken language use, this would be probably a good source. However, it wouldn’t be particularly useful if your focus was on how to write academic texts. The Corpus of US Supreme Court Opinions (Davies 2017), accessible through the same platform, would come closer to academic texts but might be too narrow since it represents the very special register of legal language. Many corpora want to represent a whole variety. The British National corpus (BNC; BNC Consortium 2001, Davies 2004), for example, is a sample of the English language as it is spoken and written by British language users. The corpus of contemporary American English (COCA; Davies 2008-) wants to represent present-day American English. There are many aspects that are relevant when it comes to questions of representativeness (which kind of novels should be included, which kinds of newspapers, which kind of spoken texts, should pop-song lyrics be included and so on), but they need not concern us here since this chapter is not about corpus compilation. However, what is of concern if you want to use corpora in language teaching is that you are aware of what it is that the corpus does actually represent. The original British National Corpus (BNC Consortium 2001), for instance, might be considered outdated to some extent since the data are from 1960 to 1993. The compilation of a corpus with newer data, BNC2014, is ongoing at the moment of writing (the Spoken BNC2014 (Love et al. 2017) is already available). So, if you wanted to find out how a word like muggle, smartphone or Corona is used, the original BNC could not provide you with the relevant information. The BNC2014 could at least help us to see how muggle or smartphone are used. Questions of text type also become important when using a corpus. Written language, generally, differs a lot from spoken language (see section 11.1). In addition, written texts might differ drastically depending on the topic or the envisaged reader‐ ship: an essay in a journal on philosophy would usually read a lot different from a news item in a tabloid. As regards spoken language, the language we use in a prepared speech would usually differ from the language we use in a conversation, where we would, again, find huge differences if it was a conversation among friends, among employees and their bosses, or among students and their professors. To make things worse, often it is difficult to tell whether language is written or spoken. Is the language of text messaging written language or spoken language? In addition, language use, even if among the same group of people and the same mode (i.e. spoken or written), can be expected to differ in a face-to-face conversation as opposed to a telephone call. Demonstrative pronouns, for example, do not make much sense if the interlocutors do not share the same space: how am I supposed to know what this in Do you see this? refers to if I do not know what my friend is looking at? The above should have made clear that language can differ a lot with regard to the situation it is used in (see chapter 11 for details). Different corpora take this into account to differing degrees. The COCA, for instance, distinguishes between spoken, fiction, magazines, newspapers, academic, web (both general and blog) and TV/ movies. Interestingly, if you wanted to find information on how to use language in informal conversations the spoken component would not necessarily be the genre to look at since it consists of conversational 7.1 The corpus as a collection of authentic language 135 <?page no="137"?> concord‐ ance sentenceand KWIC-dis‐ play data from TV and radio programmes, which are maybe more on the formal side of language use. The TV/ Movies part of the COCA might be more suitable for you as the language there is more informal. When you work with a corpus, either for yourself or with your students, you need to be aware of such aspects. The BNC might be more helpful in finding the right genre as it distinguishes between 46 written genres and 24 spoken genres. The former contains, for example, six kinds of academic prose (e.g. humanities, medicine and natural sciences), the latter features spontaneous face-to-face conversations, job interviews or religious sermons. The downside of this, however, is that the number of words for these individual genres might become rather small. This is not the time and place to go into more detail here. For teaching or classroom purposes, it usually suffices to be aware of the variety represented (e.g. British as opposed to American English) and to be able to access those genres that you are interested in. 7.2 Words and their contexts When it comes to language teaching a very important tool that a corpus can provide is that of concordance. A concordance, in its most basic sense, is a resource that provides you with the context for any word in a given text. Concordances have been around even before the advent of corpus linguistics and computers. A Bible concordance, for example, would provide you with a list of all the verses that a given word occurs in so that you could then look up what the Bible has to say about this word. Modern computer technology makes concordances a much more user-friendly ex‐ perience since the search term is already provided together with the context in which it occurs. Two kinds of display are usually offered, a sentence display and the KWICdisplay. KWIC stands for ‘Keyword in Context’ and presents all occurrences of the search term under one another. Consider table 7.1, which shows a sentence and a KWIC-display of the same five sentences. 1) Because there are children and therefore the woman has obviously been part of a couple at some point, there is an assumption that in her single state she is critical, cynical and undermining of couples around her. 2) As well as being a cry for security, however, it is also an assumption that liberalism has failed. 3) The error lay in a very simple but quite reasonable assumption which everyone made. 4) Yet as Shaw (1980) has argued the assumption that mixed schooling is preferable on academic grounds is not well supported by educational research. 5) That the surface of a rotating earth would experience a great wind rested on the assumption that the air could not rotate with the earth. 136 7 The theory of practical corpus analysis <?page no="138"?> sorting the output 1) … couple at some point, there is an assumption that in her single state she is… 2) … for security, however, it is also an assumption that liberalism has failed. 3) … in a very simple but quite reasonable assumption which everyone made. 4) Yet as Shaw (1980) has argued the assumption that mixed schooling is… 5) … a great wind rested on the assumption that the air could not rotate… Table 7.1: Examples of sentenceand KWIC-format. While both kinds of display have their own advantages and disadvantages, the format most suitable for language-teaching purposes is the KWIC format. In particular, patterns in which the keyword is used become obvious very quickly. The fact that assumption is frequently followed by a that-clause is far more obvious in the bottom half than in the top half of table 7.1. When it comes to finding patterns, sorting the output is very helpful. Sorting works with an invisible grid that is put on top of the raw data and that distinguishes between the different positions on the left and right of the search term. The position L1, for example, relates to the first word on the left of the search term, R3 is the third word on the right and so on, as you can see in table 7.2 (note that the wider context has been omitted). L3 L2 L1 Node R1 R2 R3 1) there is an assumption that in her 2) is also an assumption that liberalism has 3) but quite reasonable assumption which everyone made 4) has argued the assumption that mixed schooling 5) rested on the assumption that the air Table 7.2: The node and other positions in the KWIC-format. If you sorted the data in table 7.2 on R1, line number 3 would be shifted to the end of the output since the letter ‘w’ comes after the letter ‘t’ in the alphabet. Sorting on L1 would not change anything, since the initial letters of the words immediately to the left of assumption are already in alphabetical order. Before you read on What would the order of the five examples above be if you sorted on R3? 7.2 Words and their contexts 137 <?page no="139"?> The alphabetical order of the five words in R3 is air - has - her - made - schooling, which puts line 5) first, followed by lines 2) and 1). Lines 3) and 4) would be put at the end. The resulting sequence would be 5 - 2 - 1 - 3 - 4. Most corpus tools would also allow you to sort with regard to several positions at once. Table 7.3 shows what would happen if we sorted the data above on R1 first and R3 second. L3 L2 L1 Node R1 R2 R3 5) rested on the assumption that the air 2) is also an assumption that liberalism has 1) there is an assumption that in her 4) has argued the assumption that mixed schooling 3) but quite reasonable assumption which everyone made Table 7.3: Sorting on R1 and then on R3. The first sort would result in shifting example 3) to the end of the list. The second sort, on R3, would sort the output within the different R1-groups. The fifth line now containing example (3) would stay the same since we only have one sentence in the w-group in R1. The four sentences that show the word that in R1 will be rearranged alphabetically according to the third word to the right i.e. air, has, her and schooling. In the context of the lexical approach we maintained that it is useful to think big, i.e. to find larger patterns in language use. Sorting can help: We have already talked about the fact that the noun assumption is often followed by a complementizing that-clause. Looking up assumption that and sorting on L1 can give us an idea of adjectives that often appear in this context. The BNCweb (Hoffmann and Evert 2006) returns 1368 instances of assumption that. Figure 7.1 provides an overview of the most frequent instantiations (each adjective occurs at least five times) of the larger pattern ‘ADJECTIVE + assumption + that’. Fig. 7.1: The most frequent adjectives with assumption + that-clause. 138 7 The theory of practical corpus analysis <?page no="140"?> What I have described above is the simplest form of analysis that can be done with a corpus but it can already be very useful. Since the process itself is so simple and does not need a lot of expertise there is no problem in letting your students work with a corpus that way. For example, as a result of German mit jemandem verheiratet sein a student of yours might produce a sentence like She is married with Tom in one of their exams. This is an opportunity for your students to do some corpus work: searching for married, sorting the output on R1 and taking a look at the preposition with will quickly make clear that this is not the correct translation for mit jemandem verheiratet sein, as the examples under (4) show: (4) Sgt Forth from Sunderland was married with two children … Bill is married with two young children … Now 43, she is married with a two-year-old daughter … . He is married with a 5 year old child HASAN GULUNAY: Married with four children… Married with always occurs with information on the number of children that the married person had and would be the correct translation of German verheiratet, X Kinder. Identifying the correct preposition should be easy as the preposition to is by far the most frequent one and should be very obvious when the data is sorted on R1. Another frequent problem, even at university level, is the use of example for when example of would be correct. The problem lies in the fact that both phrases do exist in English and both would be translated in German as Beispiel für. However, we will see that the noun changes its meaning in dependence of the preposition that follows. The process, here, is a little more complex since example is a highly frequent word with 36007 occurrences in the BNC. It is possible to create random subsamples, say 1000 hits, of the query but the combination example for is very rare in the corpus data, namely only 109 hits. The first step, then, should be a query for example for. It yields examples like the following: (5) Charles was keen to use his estate at Highgrove as an example for the rest of the Duchy. That scoundrel sets a bad example for the other young men of the parish. Try to set the example for the rest of your organization. […] he is a man of the utmost integrity who is a good example for kids today. […] he was a bad example for young players to follow. A look at the sentences above reveals that example in example for is used in the sense of role model (German Vorbild) or its opposite (schlechtes Beispiel). It is possible to use it in English but not in the ‘neutral’ sense, i.e. as an instance of a category. Fortunately, the correct phrasing in this case, example of, is very frequent and even a small random sample of 100 instances provides many good examples. (6) Figure 8 is an example of such a study a twelfth-century example of this deadly weapon the much-quoted example of the parks in colonial Shanghai This is a good example of a local commemoration ‘Muzik Fantastique’ is another example of their hidden talents 7.2 Words and their contexts 139 <?page no="141"?> the problem of homo‐ nymy lemmatiza‐ tion lemma The examples of married with/ to and example for/ of show the potential that even the most basic corpus queries can have when it comes to discovering the ways words are used. This potential increases massively with more elaborate queries, as we will see in the next chapter. 7.3 More than words - annotation provided in corpora Modern corpora usually provide more than just mere texts. Additional information would typically include information on the language user (e.g. age, gender or level of education), on the medium that is used (e.g. writing or speech), the situation of use (e.g. conversation among friends as opposed to written academic article), and so on. This kind of information will help us to focus on that part of the corpus that is most likely to provide a useful answer to a question that we might be having. For example, if we wanted to find out how people start and end a conversation, obviously, we will not find an answer to that in the written academic subsection of the corpus but only in the conversation subsection. In addition to this information about the texts that are stored in a corpus, modern corpora would usually provide information with regard to the words in the texts them‐ selves. For instance, if you were interested in how the verb to bottle was used, you would have to be careful for two reasons. If you search for bottle, you would only get this form but not bottles, bottled or bottling. Your best option, in this case, would be to use the pipe character to search for alternatives, i.e. ‘bottle|bottles|bottled|bottling’ (more of that below). But even then you’d run into another problem. The form bottle is homonymous. It can be the form of a verb or of a noun. The same is true for bottles. The solution to both these problems is lemmatization. A good way to understand what that means is by taking a look at dictionaries. Let’s say you read an English text and stumble upon a sentence like And here I was, procrastinating again. If you were not familiar with the word procrastinating and wanted to look it up in a dictionary you would probably not expect to find this exact form but you’d look up the word procras‐ tinate since you know that the -ing form is one of the forms that a verb can take. In lexicography (the study of how to write a dictionary) the forms under which you can look up words are called ‘citation form’ or ‘lemma’. So, to find out the meaning of any of the forms of the verb to procrastinate, you would look up the citation form (to) procrastinate in the dictionary. In linguistics the term ‘lemma’ is understood quite similarly. It is the set of all the forms that a word (as you would find it in a dictionary) can take. For example, the verb to bottle in addition to its citation form (to bottle) occurs in the forms bottle, bottles, bottled or bottling. The lemma of the verb B O TTL EV , thus, has five forms, namely to bottle, bottle, bottles, bottled and bottling. The noun lemma B OTTL EN has four forms, namely bottle, bottles, bottle’s and bottles’. The first two being singular and plural nom‐ inative (and dative and accusative), the last two being the respective genitive forms. 140 7 The theory of practical corpus analysis <?page no="142"?> tagging Lemmatization is a process whereby the different forms that a word can take are grou‐ ped under a head word, typically written in small caps. Consider the second paragraph of example (7), where all forms of bottle have either been assigned to the verb or the noun. (7) I have been bottling water all day and I’m sick and tired of bottles. Why don’t you pick up a bottle and bottle some water. You have not bottled any water yet. (7’) I have been B O T T L EV water all day and I’m sick and tired of B O T T L EN . Why don’t you pick up a B O T T L EN and B O T T L EV some water. You have not B O T T L EV any water yet. Lemmatization is typically automatised and maybe you can imagine that it is not an easy task. For instance, how would a lemmatizer decide whether the form bottle is a verb or a noun? Another automatic process, tagging, is extremely useful here. A tag, generally, is a piece of information that is attached to any kind of object. In the context of corpus linguistics it is usually information on the word class of a given word, consider example (8). You can see that in the second line a little tag is attached to each word (and also to the punctuation mark), specifying the word class, e.g. ‘D’ for determiner, ‘N’ for noun, ‘V’ for verb and so on. (8) this book on linguistics is the best book of many good books on the topic. (8’) this_D book_N on_P linguistics_N is_V the_D best_A book_N of_P many_D good_A books_N on_P the_D topic_N._. By the way: Tagsets and levels of detail Word-class tagging depends on the set of tags that is used. Some tagsets show more detail, some less. For example, many tagsets do not just distinguish verbs from other word classes, but they would also distinguish different kinds of verbs, such as forms of the major auxiliaries to be, to have and to do. Similarly, tagsets might distinguish between the ‘normal’ form of adjectives and their comparative and superlative forms. Compare the original tagset from example (8’) above to the much more detailed one in (8’) below. (8’) this_D book_N on_P linguistics_N is_V the_D best_A book_N of_P many_D good_A books_N on_P the_D topic_N._. (8’’) this_DT0 book_NN1 on_PRP linguistics_NN1 is_VBZ the_AT0 best_AJS book_NN1 of_PRF many_DT0 good_AJ0 books_NN2 on_PRP the_AT0 topic_NN1 ._PUN The original in (8’) only distinguishes between the main lexical and grammatical word classes. As a consequence, there is no difference between words in singular or plural, like linguistics, topic and book or books. Similarly, we are not told that the verb is the third person present tense singular form of to be (VBZ). The tagging in (8’) does not tell whether we are dealing with the base form of an adjective or with its comparative or superlative form, as opposed to (8’’) which distinguishes 7.3 More than words - annotation provided in corpora 141 <?page no="143"?> between AJ0 (good) and AJS (best). Another distinction that the much more fine-grained tagset applied in (8’’) makes is that between the preposition of (PRF) and other prepositions (PRP). Tagging is done automatically and it yields suprisingly good results of around 95 % correctly tagged words (accuracy can vary considerably depending on the corpus). Take a look at the tagged version of example (7’), here repeated as (9) for convenience. (9) I have been bottling water all day and I’m sick and tired of bottles. Why don’t you pick up a bottle and bottle some water. You have not bottled any water yet. (9’) I_PNP have_VHB been_VBN bottling_NN1 water_NN1 all_DT0 day_NN1 and_CJC I_PNP 'm_VBB sick_AJ0 and_CJC tired_AJ0 of_PRF bottles_NN2 ._PUN Why_AVQ do_VDB nt_XX0 you_PNP pick_VVI up_AVP a_AT0 bottle_NN1 and_CJC bottle_VVI some_DT0 water_NN1 ._PUN You_PNP have_VHB not_XX0 bottled_VVN any_DT0 water_NN1 yet_SENT ._PUN The tagging in (9’) is quite accurate. Note, for example, that the tagger correctly assigns the tag NN1 to bottle in pick up a bottle whereas the same form in and bottle some water is correctly tagged as a verb in its unmarked form, i.e. VVI. Also note that the form can is interpreted correctly as the modal verb and not the noun (as in a can of Coke) nor the main verb (to can, meaning ‘to put in a can’). The wrongly assigned tag NN1 for bottling is most likely due to the fact that the tagger understood the sequence bottling water as a compound noun. You can see that automatic tagging is surprisingly accurate. When it comes to searching a corpus, tagging provides us with a lot of options that are extremely useful for language teaching, as we will see in the next chapter. New tools in your toolkit Representativeness ▶ You are aware of the fact that a corpus, as a sample of authentic language use, is always representative of either a whole variety or a particular text type/ register. This enables you to choose the right corpus/ sub-corpus for any question that you might have. Incidental learning ▶ You have found out that you and your students will benefit from working with corpus data since you will often learn things on top of the answers you were looking for. Concordance ▶ You now have a basic corpus-linguistic research method at your disposal which allows you to perform very simple but still very effective queries by sorting the output skilfully. Tagging ▶ You have learned that corpora usually consist of annotated text. Because of tagging it is now very easy for you to distinguish between homophonous word forms, such as the verb to help and the noun help. 142 7 The theory of practical corpus analysis <?page no="144"?> Lemmatization ▶ Corpora are usually lemmatised, i.e. the different forms of the word are grouped together under a common head word. So, you can search for all the forms of, say, a verb and know that you will not miss any relevant data. If you want to know more O’K E F F E E , Anne and Michael M C C A R T H Y (2010): The Routledge Handbook of Corpus Linguistics. London: Routledge. ▶ A very good resource that covers many aspects of corpus linguistics. Also has a section on corpora and language pedagogy and a section on designing corpus-based learning and teaching materials. T R A C Y -V E N T U R A , Nicole & Magali P A Q U O T (2020). The Routledge Handbook of Second Language Acquisition and Corpora. London: Routledge. ▶ Covers many aspects of corpus linguistics and its relations to second language acquisition. A useful resource if you are interested in learner corpus research. http: / / ucrel-api.lancaster.ac.uk/ claws/ free.html ▶ A freely available online tagger that was used to tag the original BNC and the BNC2014 as well as all the English corpora in the BYU corpora provided by Mark Davies. It’s really fun to play around with it. Additional online content. 7.3 More than words - annotation provided in corpora 143 <?page no="146"?> 8 A more practical introduction to corpus analysis Every student a Sherlock Holmes. (T IM J O H N S ) What this chapter is about This chapter introduces the basics of corpus queries and shows how to get the most information out of unannotated as well as annotated texts. In addition, the chapter will suggest ways in which corpora can be used in the classroom. They provide lots of opportunities and working with a corpus can be highly rewarding, both for you as a teacher and for your students. However, working with huge amounts of natural data comes at a price: corpus data will always contain a certain amount of junk or noise. It is important that you are aware of that whenever you use a corpus. The previous chapter has provided us with some insights into the nature of corpora and the kind of data and information that we would usually find in a corpus. This chapter aims to introduce a more practical perspective. However, it will still be theoretical in some sense. Firstly, there are many different kinds of corpus tools available. All of these do things a little differently and, therefore, it does not necessarily make sense to focus on one corpus tool and teach you how to do analyses with this particular tool. Maybe you’re not interested in using that tool or, being web-based, the tool might no longer be available in five or ten years (although most likely it will be). Secondly, from my own experience as a user of different corpus tools, I would usually forget how exactly things are done in a particular corpus tool. However, I know precisely which kind of query I want to perform and I know where to find the help file or a YouTube video that informs me about the more technical details. So, this chapter will provide you with a thorough understanding of corpus queries in general and it will, thus, put you in the position to work successfully with any of the major corpora available. 8.1 Wild cards and corpus queries In the previous chapter we have already looked at queries on the basis of word forms. The computer searched for an explicit sequence of letters and provided us with <?page no="147"?> wild cards the matches of that search. An intelligent sorting of the output, then, gave us the answers we were looking for. In the cases of married with/ to or example of/ for, for instance, this was possible since the phrases we were looking for contained exactly these sequences of letters. With assumption + that-clause this approach was also okay since the string assumptions that is a lot less frequent than assumption that (128 as opposed to 1368) and, in addition, the that in this case is usually a relative pronoun introducing a relative clause as for example in any technical assumptions that have been made. That-complement clauses with plural assumptions as in assumptions that she is driven by pity usually annoy her are the exception rather than the rule - it is the other way around for singular assumption. In many cases, however, limiting ourselves to one particular word form only might lead to skewed results. By the way: Relative clauses and complement clauses with that From my experience students sometimes struggle with the distinction of relative clause and complement clause. The easiest way to determine whether you are dealing with the first or the second is by checking whether you can insert is between the head noun and that. If this is not possible the postmodifying clause is a relative clause, if it is possible it is a complement clause. the assumption that has been made → *the assumption is that has been made the assumption that this is difficult → the assumption is that this is difficult Above we looked at adjectives that frequently co-occur with assumption that. We might also be interested in the more general question “what are adjectives that work well with the noun assumption? ” In that case there is no reason to restrict ourselves to the singular only. Similarly, you might find that your students lack variability when it comes to adjective-noun combinations. They might always be talking about big problems or about difficult or important questions. A concordance can be useful to increase their range of adjectives, but it does not make sense to restrict our query to one word form (usually the nominative singular). A first way out is the use of wild cards. The term ‘wild card’ originated in card games, where it is used to describe a playing card that does not have any given value in itself but can substitute for any other playing card, such as the joker card. In the context of corpus linguistics the term describes a character that can be used to represent one or more other characters. For example the question mark character ‘? ’ is used to substitute for one single arbitrary character, i.e. searching for ‘characteri? e’ in your corpus would yield characterise as well as charac‐ terize as results. Unfortunately, different corpus softwares and interfaces may use dif‐ ferent wild cards and it will always be necessary to take a look at the manual or the help section to find out how things are done with the interface you are currently 146 8 A more practical introduction to corpus analysis <?page no="148"?> the ‘? ’ wild card the ‘*’ wild card working with. But if you have a basic understanding of the functions and the potential of wild cards, coming to terms with these idiosyncrasies will not be any problem. This is why we will tackle the problem of wild cards from a functional perspective. In some cases it might be necessary to keep one slot in a string of characters variable. Let us assume you want to take a look at how women are characterised in the English language. In this case it does not make sense to restrict your search to the singular woman only. Instead of doing two individual queries you can combine them through the ‘? ’ wild card. It substitutes for any conceivable character, which is why the query ‘wom? n’ will find all instances of woman as well as women. When it comes to orthog‐ raphy British English and American English spelling is often difficult for learners to keep apart. A useful exercise in the classroom might be to have your pupils search for different words ending in -ize/ -ise, such as characterise, emphasise, sympathise etc. to let them find out on their own. Again, the ‘? ’ wild card would come in handy. Another wild card, the asterisk ‘*’, substitutes for zero or more characters. A search for *man, for example yields man, woman, chairman, policeman, human, Chapman, Roman, Rahman etc. A combination of the two is possible: the query *m? n finds the above but also the plural forms men, women, policemen, gentlemen etc. While these might be forms you are interested in you will also find common and uncommon or admin. This makes clear that wildcards have to be used with care lest you increase the amount of junk, i.e. search results that you are not interested in, beyond reasonable limits. More of that below. Before you read on … … think about the following problem. You might be tempted to make your search for differences regarding British and American English spelling with verbs in -ise/ -ize more efficient through the use of the search string ‘*i? e’. Which junk word forms would be produced that way? As I was writing this chapter I actually wanted to use the string *i? e as an example to illustrate how we can make the query for verbs ending in -ise/ -ize more efficient. This sounded like a really good idea in theory but it did not work in practice. The query was not specific enough and produced too much junk, such as provide, choice, office or vaccine. Making mistakes like these is important und useful because they give you a feel for problems and pitfalls in the context of corpus queries. The asterisk wild card is particularly helpful for morphological studies. Say, you wanted a list of all the adjectives that end in -ful. The query *ful will provide you with all of them, e.g. successful, useful, beautiful or powerful, but, as always, also a little junk like the quantity nouns handful, spoonful or teaspoonful. If you wanted to find all the adjectives ending in -able, such as reasonable, comfortable or acceptable, *able will do the trick. 8.1 Wild cards and corpus queries 147 <?page no="149"?> the pipe character ‘|’ Beyond morphology the asterisk is also very useful if we want to look for larger patterns. Above we talked about typical adjectives that go with nouns. Let us assume we want to take a look at the adjectives that are commonly used to premodify woman/ women and compare these to what we find in front of man/ men. Instead of searching for the nouns in question and then sorting the output on L1 we could also look for the string ‘* wom? n’ and compare these results to ‘* m? n’. In this special case, a pattern query does not have any advantages over sorting concordance lines but pattern queries become more useful if the patterns themselves become more complex. Before you read on … … let us assume you wanted to find the most frequent adjectives that occur in the frame ‘really ADJ woman/ women’. What is the pattern you would look for? To find out which adjectives are used in the frame ‘really ADJ woman/ women’, the pattern query ‘really * wom? n’ will be your best choice (for the time being). The asterisk substitues for zero or more characters and would, thus, identify all strings of letters between really and either woman or women (recall that the question mark substitutes for one single character, including ‘a’ and ‘e’. This pattern query would be much faster than sorting concordance lines with regard to the different positions. A last wild card that can be very useful is the pipe character ‘|’. It is not always easy to find on a keyboard since the key may show different symbols such as two small vertical lines on top of each other ‘¦’ or a pipe character in a circle. On a German keyboard layout (QWERTZ) you can find the character on the same key as the ‘smaller than’ and ‘larger than’ symbols and you access it by pressing ‘Alt Gr’ together with this key. English keyboards (QWERTY) would usually have the character together with the backslash key ‘\’ and you access it by holding down the shift key while pressing it. The pipe character enables us to specify alternatives in our patterns. Above we said that men would usually be described as handsome and women as pretty. We could check whether this is actually true by searching representative corpora for the patterns ‘pretty m? n|handsome m? n’ and ‘pretty wom? n|handsome wom? n’. The pipe character is also extremely useful if you want to make sure that you get all the forms of, say, a given verb. Let’s assume you want to find all occurrences of the verb to help. In addition to a search for help you would also have to do one for each of the other forms, i.e. helps, helped, helping. Using the pipe character we only have one instead of four queries, namely ‘help|helps|helped|helping’. Most corpus interfaces would make use of the wild cards discussed above in the same way. However, when queries become more complex, the query syntax, i.e. the way in which queries have to be formulated, might differ, as can be seen in table 8.1 below. 148 8 A more practical introduction to corpus analysis <?page no="150"?> AntConc pretty man|handsome man www.english-corpora.org pretty|handsome man BNCweb (CQP-Edition) (pretty|handsome) man Table 8.1: Examples of differences in query syntax. AntConc (Anthony 2022) does not provide the option of specifying alternatives for a given slot, instead we have to search for whole alternative queries, i.e. ‘pretty man’ as opposed to ‘handsome man’. Both www.english-corpora.org as well as the BNCweb work with alternative fillings of the same position. But while the former will understand any word next to the pipe character as one option for a given slot the latter demands that we put a bracket around the alternatives. None of this is particularly problematic but as users we have to be aware of the idiosyncrasies of the query syntax of the different corpus interfaces. It is always useful to find the help files or pages that come with interfaces (e.g. BNCweb or www.english-corpora.org) and check how things are done. In AntConc the relevant information would be found in the ‘Global Settings’ tab. It also makes sense to explore the help sections a little, since some softwares or interfaces offer options that others do not. We have seen above that the ‘? ’ wild card substitutes for a single arbitrary character. AntConc, in addition, provides the ‘+’ wild card, which stands for zero or one character. This does not seem to make an awful lot of sense at first but it can make your life easier in some cases. For example, American and British English have a fairly consistent contrast between -bor and -bour, e.g. labor-labour, neighbor-neighbour or harbor-harbour. A search for all words ending in either -bor or -bour could be done with the help of the wild card, i.e. ‘*bo+r’. Further relevant differences become obvious if we move beyond simple text-based queries. 8.2 Corpus queries - exploiting lemmatization and tagging In the previous chapter we have introduced tagging and lemmatization. The main advantage of a tagged and lemmatized corpus is that our queries can become a lot more specific. For example, if we wanted to identify adjectives that go well with the noun help (and all its forms) we would have to search for ‘help|helps’ and then sort the output on L1. You might wonder at this stage whether this would not ignore the (admittedly very rare) genitive forms of the noun help. This is a valid question. However, many taggers would not treat a genitive form as one word but as a combination of a word plus a possessive marker. The genitive form help’s, for example, would be separated into help + ‘s and tagged as ‘help_NN1 ‘s_POS’. Still, after our search we would be faced with two further problems. Our results would contain a lot of junk, since two of the noun forms are also verb forms, namely help and helps. It takes a lot of effort to 8.2 Corpus queries - exploiting lemmatization and tagging 149 <?page no="151"?> get rid of all of these. In addition, we would have to go through all of the concordance lines to see which of the words on L1 actually are adjectives and which of these are particularly frequent. What can we do? Before you read on … … take a look at the eight tagged lines from the BNC below. Using your knowledge of wild cards and lemma queries, which query would you have to use to identify all instances of a form of the noun help preceded by an adjective? a. Every_AT0 donation_NN1 does_VDZ help_VVI b. That_DT0 help_NN1 makes_VVZ all_DT0 the_AT0 difference_NN1 c. as_AV0 far_AV0 as_CJS possible_AJ0 help_VVB them_PNP d. Get_VVB any_DT0 medical_AJ0 help_NN1 available_AJ0 e. This_DT0 food_NN1 was_VBD an_AT0 invaluable_AJ0 help_NN1 f. the_AT0 lack_NN1 of_PRF environmental_AJ0 helps_NN2 g. You_PNP will_VM0 get_VVI better_AJC help_NN1 h. the_AT0 great_AJ0 army_NN1 of_PRF unpaid_AJ0 helpers_NN2 First, let’s find ways to identify the noun help. Several options come to mind. The first treats the annotated text like any other kind of text. So, to get all the forms of the noun help you could search for ‘help*_N*’, which would identify lines b, d, e, f and g. This query translates as ‘Find all nouns that begin in help’ and, accordingly, would also find line h, since helpers is also a noun and satisfies the restriction ‘help*’. If your query syntax had the wild card ‘+’, i.e. zero or one character you could use the query ‘help+_N*’. This would give you both the forms help and helps as nouns and would not retrieve helper or helpers. With a combination of lemma and wild card none of these problems matter anymore. The query ‘{help/ N}’ in the BNCweb interface or ‘HELP_nn’ in www.english-copora.org will yield all occurrences of all forms of the noun help. The only thing left to do is to specify that we want this noun to be preceded by any adjective. The tagset used to tag the corpora at www.english-corpora.org and the BNCweb distinguishes unmarked adjectives from comparative and superlative adjectives, i.e. AJ0, AJC and AJS, respectively. A possible query to identify all occurrences of the noun help preceded immediately by an adjective would, therefore, be ‘*_AJ* {help/ N}’ (in the BNCweb). As you can see in example (9) this string matches all of the bold expressions. By the way, in www.english-corpora.org the whole procedure is a bit different and the string ‘ADJ HELP_nn’ would yield the required results. (9) d. Get_VVB any_DT0 medical_AJ0 help_NN1 available_AJ0 e. This_DT0 food_NN1 was_VBD an_AT0 invaluable_AJ0 help_NN1 f. the_AT0 lack_NN1 of_PRF environmental_AJ0 helps_NN2 g. You_PNP will_VM0 get_VVI better_AJC help_NN1 150 8 A more practical introduction to corpus analysis <?page no="152"?> Another advantage of any kind of pattern query is that you can also see all the different realizations of that pattern at one glance. Figure 8.1 for example shows the first ten items of the hits that the above query produces. Fig. 8.1: The top ten results of the query ‘ADJ HELP_nn’ in www.english-corpora.org (with kind permis‐ sion of Mark Davies). As regards the use of corpora in the classroom it is easy to see how we can benefit from lemmatization and tagging. In the previous chapter we explored the differences between married to and married with as well as between example of and example for. We had to take a detour to arrive at the results. With married with, for instance, it was easy to find that this is no translation for verheiratet mit, but to find that married to would be correct involved scanning a full concordance of the word married. The same holds true for Beispiel für. Using tags makes our life a lot easier by merely looking for ‘married *_PR*’ or ‘example *_PR*’, respectively. The queries translate to ‘Look for any occurrence of the string married/ example immediately followed by another string that has any tag attached that starts in PR, i.e. either PRF (the preposition of) or PRP (any preposition other than of).’ In the previous chapter we tried to identify things that you make. This can now be done very easily through the query ‘MAKE *_N*’. In the previous chapter we also talked about near-synonyms, such tall and large. Again, with the help of tags it is a lot easier now to find out which nouns go with tall and which go with large just by using the search strings ‘tall *_N*’ or ‘large *_N*’, respectively. Take a look at table 8.2 for the top 10 most frequent noun forms with the two adjectives. tall *_N* large *_N* tall man large number tall figure large numbers tall order large part tall trees large scale 8.2 Corpus queries - exploiting lemmatization and tagging 151 <?page no="153"?> tall *_N* large *_N* tall boy large proportion tall buildings large amounts tall glass large quantities tall woman large amounts Tall ships large extent tall windows large areas Table 8.2: The 10 most frequent noun collocates of the adjectives tall and large in the BNCweb. All in all, you can see how tagged and lemmatized corpora can make it very easy for you to identify patterns in language use. One problem that still remains is that patterns are often not very flexible (depending on the interface you use). For example, the things that you make do not always occur immediately after the verb, e.g. make a scene, make a promise, make a mistake. We would at least need two queries. ‘MAKE *_NN*’ will identify all patterns where the noun follows the verb immediately, ‘Make * *_NN*’ would allow for one intervening word between the verb and the noun. Maybe a third query, ‘MAKE * * *_N*’ might be necessary to catch cases like I made a terrible mistake. To find out whether handsome is indeed an adjective that goes together with man, a simple query like ‘handsome {man/ N}’ or the more general ‘*_AJ* {man/ N}’ would not detect strings like those shown in (10) below. (10) a. a man as handsome, as powerful, and as … as … b. a handsome and elegant man c. one young man, very handsome …, sat among them amiably d. three very handsome young men e. a handsome and well-muscled man f. a handsome, well-groomed young man g. other handsome, rich young men h. a handsome grey-haired, grey-eyed man The problem with pattern queries in general is that you need to have some idea as to what possible implementations of the pattern at issue look like. It would be nice to have an automatic way that would provide us with more flexible queries. Fortunately for us such an automatic way exists. 8.3 Corpus queries - collocations Most corpus interfaces and softwares make it very easy for you to get an idea of co-occur‐ ring words since they enable you to search for collocations automatically. Above we have seen various definitions of the term ‘collocation’, the general idea being that two words are 152 8 A more practical introduction to corpus analysis <?page no="154"?> characteristic co-occur‐ rence collocates if they co-occur together characteristically or typically. How do we tell a com‐ puter what characteristic or typical co-occurrence is? Usually this is done through various kinds of statistical measures. The details of these are not relevant here but it makes sense to have some basic understanding of the underlying rationale. ‘Characteristic co-occurrence’ in the statistical sense means that two words occur more frequently than we would expect them to co-occur if their co-occurrence was merely random. For example, above we looked at patterns in which naked eye occurs. The word naked itself is a collocate of eye, even though the phrase naked eye only occurs 148 times in the BNC. The string my eye can be found 266 times in the BNC but my is not as strong a collocate of eye as naked is, since my occurs 145,250 times in the BNC, whereas naked is only found 1878 times. Since my is al‐ most 80 times as frequent as naked, the probability of my to occur with eye is a lot higher than that of naked to co-occur with eye, which makes its actual co-occurrence not particu‐ larly typical, characteristic or special. Put differently, only 0.18% of occurrences of my occur in front of eye, whereas 7.9% of occurrences of naked can be found in that position. The word eye seems to be special for naked but not for my. By the way: A bit of the maths - comparing expected and observed frequencies There is a large number of statistical measures to find out how characteristic the co-occurrence of two words is. By way of illustration let us take a look at expected and observed frequencies to get an idea of how degrees of characteristic co-occurrence can be calculated (usually other measures are used). The BNC has a total of 100,000,000 words. Within these 100 million words, naked occurs 1,906 times. The noun eye in the singular occurs 8,949 times in the BNC. So, if naked occurs 1,906 times in the whole BNC, how often would we expect it to be found in front of the noun eye (if naked was distributed completely randomly)? The solution is simple and involves the rule of three: 1, 906 100, 000, 000 = X 8, 949 X = 1, 906 · 8, 949 100, 000, 000 ⇔ X = 0.171 As you can see, the expected frequency of 0.171 times is a lot lower than the actual frequency which is 148 times. If we divide the observed frequency by the expected frequency we get the following score S. S = f obs f ex p ≈ 865 In this case the score is roughly 865. The word my, in contrast, occurs 146,573 times in the whole BNC. If we substitute this number for the 1,906 occurrences of naked we get an expected frequency of 13. Dividing the observed frequency of 268 occurrences of my eye by the expected frequency of 13 we get a value of roughly 21, which is a lot lower than the score of 865, which we had for naked eye. 8.3 Corpus queries - collocations 153 <?page no="155"?> The search for collocations is not restricted to the L1 position but can be applied to any window around a given search term. You could, for example, ask your interface to search for collocates in a window from L2 to R3 or from R1 to R4 and so on. In addition, if the corpus is tagged you would also have the option to restrict your search to collocates of a specific word class. In our handsome man example we would primarily be interested in the adjective collocates - the interfaces provided in both, the BNCweb and the www.english-corpora.org allows you to restrict your search accordingly. You can even be more specific and restrict your search to adjectives in their basic form only, or those in comparative or superlative form. Your options are almost endless and it depends on you to make the most out of it. Figure 8.2 gives you an impression of those options in the BNCweb and in www.english-corpora.org. In both cases we look for adjective collocates in the window of three words to the left and right of the string man. Fig. 8.2: Specifying collocation parameter for man in BNCweb (top) and in www.english-corpora.org (bottom) (with kind permission of Sebastian Hoffmann and Mark Davies). 154 8 A more practical introduction to corpus analysis <?page no="156"?> data-driven learning step 1: corpus as advanced ‘dictionary’ 8.4 Corpora in the classroom Hopefully, the chapter so far has given you a good idea of how you (as a language learner) could benefit from using a corpus. We will now turn to ways in which corpora can enrich your teaching. Basically, there are two ways in which corpora can be useful: 1) you can give your students access to corpora, and 2) you can use corpora as a very helpful resource to create authentic learning material. In both cases, corpora provide lots of benefits and opportunities. At the same time, however, there are also problems and pitfalls that you need to be aware of. 8.4.1 Benefits and opportunities One of the major and maybe most rewarding opportunities a corpus can offer is that it can give learners of a language access to vast amounts of authentic language data. The idea of having students use corpora to find out about language use is maybe best captured in Johns’ concept of ‘data-driven learning’ (DDL): My approach was rather to confront the learner as directly as possible with data, and to make the learner a linguistic researcher. The metaphor I use with my students is that of the detective, ‘Every student a Sherlock Holmes! ’. ( Johns 2002: 108) Students become researchers themselves with activities ranging from the analysis of pre-edited concordance lines to Bernardini’s (2002: 166; see also 2000) idea of corpus-linguistic ‘discovery learning’, where “learners browse corpora much in the same way as they would explore an unknown land”. The advantage of this approach over the more traditional teaching approach is primarily that students take more ownership of their own learning process. It makes sense to assume that information that students have found on their own is more likely to be kept in mind than facts that have just been presented to them. The road to discovery and the act of discovery itself leads to a stronger contextualization resulting in a firmer mental integration of the discovered information. In addition, as we have seen above, working with corpora provides ample opportunities for incidental learning, i.e. students will find interesting and useful facts beyond the answer to their original question. Still, as a teacher you have to be aware that corpora can be a little overwhelming at first and it will take time to come to terms with the variety of options that a corpus interface provides. It is important, therefore, to start with simple tasks. A good way in is to let students check whether a phrase they have used does actually mean what they intend it to mean. The corpus, in a sense, would be used like an advanced dictionary. An example would be the string married with. A simple concordance of the string itself will make it clear that it is no translation of verheiratet mit. To help them find the correct translation you could ask them, in a second step, to search for married, sort the output on R1 and find out which preposition is the correct one in this case. This simple con‐ cordance approach is useful in all those cases where your students use an uncommon 8.4 Corpora in the classroom 155 <?page no="157"?> step 2: wild-cards and the pipe character step 3: lemma queries step 4: collocation queries or erroneous combination of words. To give but one further example: A student of mine once wrote a harsh change. A simple search for harsh change in the BNC quickly re‐ vealed that this phrase cannot be found in the 100 million words of the corpus. A simple follow-up concordance of change and sorting on L1 provides better adjectives such as drastic, fundamental, major or significant. All of this can be easily done and does not need any particular expertise with corpora and their interfaces. Even more complex queries can be introduced without much effort if they are treated like simple phrase queries. For instance, you could make your students acquainted with the use of wild cards by asking them to type in ‘a * change’ in the query window of your interface and let them explore the output. The query ‘(handsome|pretty) man’ (in the BNCweb) would let your students discover that the first adjective is used far more frequently than the sec‐ ond (81 as opposed to 1) and it would introduce the pipe character at the same time. Moving away from the most basic queries, in a next step students could be introduced to lemma queries. As before, it makes sense to first provide the queries for your students. For example, the query ‘_{A} {change}’ (in the BNCweb) would find all instances of an adjective immediately followed by any form of either the noun or the verb change. This is not a prob‐ lem in this context since the noun change will much more frequently be preceded by an adjective than the verb to change. A restriction to nouns only would be achieved by ‘_{A} {change/ N}’. It is up to you to decide at which point you want to introduce your students to more complex queries and when you want to explain to them what the syntax actually means and what it does and why it is necessary to make those specifications. As a next step your students could be made familiar with the search for collocations. As described above, this would free your students from the restrictions that a pattern query entails and often provides the necessary information in a few mouse clicks only. For exam‐ ple, after you have chosen the BNC in www.english-corpora.org a click on ‘collocates’ pro‐ vides you with one field for the word or phrase you are looking for and another for the collocates you are interested in. Note that the syntax differs from the one of BNCweb. If you type in ‘CHANGE_nn’ in the first, ‘ADJ’ in the second field and then specify a search win‐ dow (e.g. L2 to R3), a click on ‘Find collocates’ will lead you to a list of all adjectives that collocate with the noun change in the specified window (see figure 8.3). Fig. 8.3: Looking for adjective collocates of the noun change in a window of L2-R3 in www.english-cor pora.org (with kind permission of Mark Davies). 156 8 A more practical introduction to corpus analysis <?page no="158"?> In a similar fashion we can find the nouns that typically occur as the direct object of a given verb - you may recall the little exercise in chapter 6 where you had to guess the verb to make. Often this is very useful when dealing with light verbs, i.e. verbs that only have a very general meaning. All you have to do is search for ‘{make/ V}’ and {do/ V}, respectively, set the correct collocation window span and filter the results by the tag ‘any noun’, as shown in figure 8.4. Table 8.3 shows the first 15 entries of the noun collocates of the verbs to make and to do from R1 to R3 in the BNCweb. Fig. 8.4: Looking for noun collocates of the verb to make in a window of R1-R3 in the BNCweb (with kind permission of Sebastian Hoffmann). Noun collocates of to make (R1 - R3) Noun collocates of to do (R1 - R3) sense job difference things use harm mistake thing decisions work contribution good decision bit progress homework effort lot debut justice love washing statement trick profit utmost 8.4 Corpora in the classroom 157 <?page no="159"?> step 5: larger patterns Noun collocates of to make (R1 - R3) Noun collocates of to do (R1 - R3) mistakes shopping way talking Table 8.3: The top 15 noun collocates (R1 to R3) of to make and to do in the BNCweb. As you can see, not all the results in the right-hand column fit: bit and lot are mostly used as quantifiers, as in I was doing a bit of surfing myself or a lot of homework but still, a simple query like this gives you a very good idea of the things that you do and the things that you make. You may recall from chapter 7 that the native speaker has available to them larger chunks or prefabs of language. As a consequence, students should develop an aware‐ ness of the general patterned nature of language and they should develop an ability for identifying patterns. Starting off from the list above, students could set out to search for larger patterns with chosen verbs or nouns. Looking at individual concordance lines they can identify patterns such as ‘ MAK E use of X’ or ‘ MAK E an effort to do X’. A look at the occurrence of to do + harm will even reveal quite a large number of patterns: ‘ DO harm to X’, ‘ DO PersProN harm’, ‘something DO PersProN no/ any harm’ or ‘ DO more harm than good’. The five steps of copus queries sketched out above are summed up in figure 8.5 below. Fig. 8.5: Five steps to introduce working with a corpus to your students. 158 8 A more practical introduction to corpus analysis <?page no="160"?> creating cor‐ pus-based materials For you as a future teacher, the results from your corpus query can serve as a useful resource to create your own authentic corpus-based learning materials. In the case above, the concordance lines for the individual collocates can become part of a fill-in-the-gaps exercise as shown in figure 8.6. Fig. 8.6: A simple corpus-based fill-in-the-gap exercise on the light verbs to make and to do. The same data set can be used to create an exercise for noun collocates of the verb to make, as shown in figure 8.7. Fig. 8.7: A simple corpus-based fill-in-the-gap exercise on noun collocates with the light verb to make. 8.4 Corpora in the classroom 159 <?page no="161"?> recall Picking useful concordance lines to make such exercises obviously involves some skill and might need a bit of training. Generally, corpus data is usually not simple or ‘clean’ data. Actual language use is far more complex than that shown in made-up examples from textbook writers. In addition, since a lot of corpus annotation is done automatically, you will find mistakes, examples being the occurrence of bit and lot in the list of noun collocates of to do. Finally, corpus queries can be problematic since any natural language comes with its ‘problems’: a word, for instance, might be ambiguous or two different words might happen to be written the same way. All of these make a corpus a somewhat advanced tool to use in the classroom and it is important that you as a teacher are aware of problems and pitfalls that you might encounter when working with a corpus yourself or when providing your students with access to original corpus data. 8.4.2 Problems and pitfalls At various places in this section we have already hinted at the benefits of using corpora in the classroom. A very general advantage was that corpora provide you and your students with access to unfiltered data, basically: they are the real thing. Even though unfiltered data can be overwhelming, on the plus side, they will usually provide more authentic and much richer language than that found in textbooks. Consequently, it will offer far more opportunities for incidental or serendipitous learning - both, for you and your students. For you as a teacher the corpus, in addition, can be a useful way to substantiate claims that you make about the English language. Where before you would have had to tell your students “It just doesn’t sound right”, you can now show them results from a corpus that indicate that some expression, though not incorrect, is very rare or restricted to a particular text type. Your students might benefit from working with a corpus since it lets them discover facts about language on their own, leading to stronger memory traces than if these facts had just been presented to them alongside the many other facts that they have to digest over the course of a school day. All of these benefits, naturally, come at a cost. More natural data is more complex data and sometimes may be overwhelming. In what follows we will explore this aspect in more detail. With any kind of corpus query two aspects become relevant, namely the recall of the query and the precision of the query (Ball 1994). The term ‘recall’ refers to the percentage of relevant cases that you identify. The term ‘precision’ refers to the amount of junk that a query produces. Both concepts are interdependent in the sense that a high precision will usually result in a lower recall, while a high recall will often produce a lot of junk, i.e. will lead to low precision. Whenever you use corpora in the classroom you will want to find the ‘sweet spot’ where a reasonably precise query corresponds with a reasonable recall. How can this be achieved? 160 8 A more practical introduction to corpus analysis <?page no="162"?> Before you read on … … let us assume you want to find out which nouns typically occur as the direct object of the verb to say. How could you find out with the help of a representative corpus? Direct objects, typically, would occur after the verb. So, one possible solution would be to run the query ‘say’ and then sort the output. A potential pitfall in this scenario is that you sort on R1 only. True, you would find some nouns this way (e.g. say goodbye), but often the object would be a noun phrase that contains a determiner (say a prayer) or more complex premodification (say a well needed prayer). So, if you only looked at R1, you would miss all of the nouns that occur in R2 and R3. Your whole approach would have a very low recall. Increasing the recall when using concordance lines involves time consuming consecutive sorting with regard to these positions. A much better approach, therefore, would be to work with the automatic retrieval of collocations by looking for noun collocates in the window R1 to R3. This way it would be a lot easier for you to get a better recall regarding nouns in these positions. Before you read on … … riddle me this! Searching for collocations of say in R1 to R3 might relieve you of the tedious sorting procedure but you would still have an extremely low recall. Do you know why? The problem I have in mind is a typical rookie mistake, which I would assume any person who has ever worked with a corpus has made at least once. If you run the query ‘say’ and then search for collocations in R1 to R3, you would have missed almost 80 % of the relevant data, since the form say is only one possible form of the verb to say (others are said, says, saying) and accounts for roughly 21 % of all cases (the most frequent form is said with 61 %). A way to avoid this problem is to use lemma queries, e.g. {say/ V} in the BNCweb, and then identify noun collocates in the relevant window. Generally, performing simple queries and then sorting the output with regard to different positions can lead to low recall if the co-occurring words that you are looking for are parts of structures with some degree of complexity. Again, your linguistic expertise will be useful to assess whether sorting can be useful or not. What linguistic aspects do we have to keep in mind when we are looking for nouns that typically occur as the direct object of the verb to say? Firstly, we are looking for the direct object of a verb. Where does the direct object occur relative to the verb? If our sentence is an active sentence, the direct object would usually follow the verb immediately. However, if our sentence is a passive sentence, the direct object would precede the verb (including an intervening form of to be). So, if we focus on a window on the right of the verb, we will ignore the direct objects in the passive sentences. This is not a problem generally 8.4 Corpora in the classroom 161 <?page no="163"?> precision since the active voice is a lot more frequent than the passive voice, but it is good to be aware of this issue. Secondly, the direct object of the clause is often instantiated by a noun phrase. Usually, this noun phrase consists of more than just the noun as we see in the simple example He reads an interesting book on linguistics. If we are looking for noun collocates as direct objects of verbs, what we are actually looking for are the heads of the noun phrases that instantiate these direct objects - book in the example sentence. Because of the different structural realizations of a noun phrase these head nouns can occur in different positions with regard to the verb as we have seen above. As a consequence, performing a simple query and then sorting the output is not the best approach. However, if structures are not as complex, sorting concordance lines might be very efficient. For example, if you want to know which adjectives go well with a given noun (e.g. handsome man or pretty man), a simple query for the noun followed by sorting on L1 would usually do the trick. This is because, in contrast to noun phrases, adjective phrases are a lot less complex, frequently consisting of the adjective only. The same holds true for adverb phrases as premodifiers of adjectives. In both cases recall would be fairly high and you would not have to bother your students with the technicalities of more complex pattern queries or collocation queries. Again, linguistic expertise will prove useful. Precision, the other relevant aspect of any corpus query, can be seen as the counterpart to recall. Precision has to do with the amount of junk that a corpus query produces. Generally, we can say that high precision leads to a low amount of junk but may also run the risk of reducing recall. On the other hand, maximizing recall might come at the expense of reducing precision, which will lead to a larger amount of junk. Whereas the degree of recall can often be estimated at least with some degree of accuracy (if you are in possession of the relevant linguistic expertise) to predict the amount of junk will often be more difficult. What would be needed is an intuition for the frequencies with which words and structures occur in language use. Unfortunately, we usually lack this intuition. Before you read on … … let us assume you want to find out how the adjective pretty is used. How could you find out with the help of a representative corpus? Maybe you thought of the same approach that I tried out first: search for the word pretty in your corpus and then sort the output on R1 to identify typical nouns that are premodified by the adjective. What you will find is that the word pretty is most frequently used as an adverb premodifying another adjective and it is only half as frequently used as an adjective itself. With regard to your original question, then, you see that the results of our query contain a lot of unwanted concordance lines, i.e. a lot of junk. 162 8 A more practical introduction to corpus analysis <?page no="164"?> By the way: Learning the hard way - present perfect with since I stumbled across a problem with junk when I tried to create a corpus-based exercise that should lead students to ‘discover’ the use of perfect aspect with the preposition/ conjunction since, as in I haven’t seen him since I left school. My first attempts were so disheartening that I decided to take a closer look at the problem and wrote a paper (Kreyer 2010a) about it. In this paper I analysed a random sample of one hundred concordance lines of the search term since and found that only thirty-nine of these would have been suitable for (advanced) students as in example (a) below. In forty-two cases since was used as a causal conjunction similar to because (b), and in nine cases since introduced a postmodifying prepositional phrase (c). Six of the instances showed a finite verb without perfect aspect (d) and four occurred in verbless clauses (e). a. The NRA has been closely monitoring the situation since the mine closed. b. Since the three theories are rather different, it may be helpful to bring out here to common themes. c. Miss F, a London student, is now in her first relationship since the assault. d. Probably had ten more kids since you were last here. e. Long time since we did fall. All in all, more than 60 % of the results were junk and did not even illustrate the phenomenon I was looking for. To make the situation worse, not all of the remaining concordance lines would necessarily be suitable for learners of English since some of the language used might be too complex and advanced for them. However, even if things do not fall nicely into place from the start, often a little bit of linguistic expertise can go a long way. Most of the problems were due to the homonymy of since and the fact that 42 % of all occurrences used causal since, i.e. similar to because. If we ignore these cases, we find that 39 out of 58 concordance lines are actually what we are looking for, and that is an impressive two thirds (Kreyer 2010a: 168). Even less advanced students would be able to weed out the cases of causal since (“Ignore all those sentences where you could substitute since by because! ”) and would, thus, have a good change of discovering the use of present perfect aspect with since. Still, as before with regard to recall, even simple corpus queries can have a high precision. A case in point is the position of the adverb already. In a study from 2010 (the same as the one discussed in the box above) I looked at 100 random cases of already and found that 75 % of the concordance lines showed what we would expect from textbook rules: already is either placed directly after a form of to be (11a) or after the auxiliary, immediately preceding the main verb (11b). 8.4 Corpora in the classroom 163 <?page no="165"?> (11) a. Though he is already aware of speech through sound, amplified vibration provides reinforcement. b. The success confirmed what his instincts had already told him. A high precision can be expected with pattern queries that contain alternatives and no fully variable slots. For example, your students could very easily find the difference between married to and married with on their own. A search for ‘married (to|with)’ in the BNCweb, for instance, will produce almost no junk and the difference between the two phrases becomes apparent almost immediately. Similarly, false friends can easily be explored: for example, a search for handy will make it clear very quickly that this word is no translation for the German word Handy. To conclude, corpora are an excellent resource that you and your students can benefit a lot from. However, corpora do not give away the wealth of information that they contain willingly and easily. It is up to you as a language learner yourself to find ways to get the most out of any corpus and it is your responsibility as a teacher to help your students become corpus-proficient. True, this will take some time and effort but you will provide your students with access to an excellent learning resource that has the potential to take your students to the next level. New tools in your toolkit Wild cards ▶ You have learned how to use the most important wild cards and the pipe character to make your corpus queries efficient and effective. Annotation ▶ You have learned how to exploit the annotation that comes with corpora, in particular tagging and lemmatization. In addition to making basic queries more efficient and effective this also allows you to search for larger patterns like the ones we have discussed in section 6.2. Collocations ▶ You have learned that many corpus interfaces provide you with the option to search for collocations directly. This enables you to find co-occurring words very easily. Five steps of corpus queries ▶ You have learned about one possible way to take your own students from a mere awareness of the patterned nature of language to being able to freely discover this patterned nature by themselves. Corpus-based learning materials ▶ Your corpus linguistic expertise enables you to use concordance data to create those learning materials that you think are optimal for your students. Pitfalls ▶ You are now aware that corpus analysis is not trivial and that at times original and non-edited corpus data might be overwhelming for your students. You have also learned that your linguistic expertise can help you prevent some of the problems that come with corpus analysis. Precision and recall ▶ You are now aware that any corpus query can be described by these two interdependent variables. This helps you to form the corpus query that is most useful for your purposes. 164 8 A more practical introduction to corpus analysis <?page no="166"?> If you want to know more Lindquist, Hans & Magnus Levin ( 2 2018): Corpus Linguistics and the Description of English. Edinburgh: EUP. ▶ Provides a hands-on introduction to corpus-based analysis with the help of a variety of different corpus resources. Covers a broad range of research questions from lexis and collocations to metaphor and gender aspects. Additional online exercises available. S Z U D A R S K I , Pawel (2018): Corpus Linguistics for Vocabulary. A Guide for Research. London: Routledge. ▶ An introduction to the use of corpora for lexical research, including aspects such as frequency, phraseology, formulaicity as well as corpora and teaching vocabulary or corpora and discourse analysis. Also provides in-depth discussions of representative studies. T I M M I S , Ivor (2015): Corpus Linguistics for ELT. Research and Practice. London: Routledge. ▶ Provides an overview of topics such as corpus compilation, corpus research into lexis and grammar, or the use of corpora in the classroom. W E I S S E R , Martin (2016): Practical Corpus Linguistics: An Introduction to Corpus-Based Language Analysis. Hoboken, NJ: Wiley-Blackwell. ▶ A really practical introduction to corpus linguistics, which will be very helpful for creating your own datasets as well as cleaning and analysing them. Also provides information on how to use BNCweb, the BYU corpora and AntConc and features an online student companion site. Additional online content. 8.4 Corpora in the classroom 165 <?page no="168"?> Abstract Texts - - Dear Jake, This is the safety protocol - Sincerely, Raymond Holt - - You don’t have to sign your name on texts. - - Dear Jake, suggestion noted. Sincerely, - Raymond Holt - (F R O M A T E X T - M E S S A G E E X C H A N G E B E T W E E N J A K E P E R A L T A A N D R A Y M O N D H O L T ) Contents 9 What makes a text a (good) text 9.1 What makes a text a text - standards of textuality 9.2 The relevance of text structure 9.2.1 Themes and rhemes 9.2.2 Rhetorical Structure Theory 10 Texts and the relevance of sentence structure 10.1 Principles of text processing 10.1.1 The given-before-new principle 10.1.2 The principle of end-focus 10.1.3 The principle of end-weight 10.2 Helpful sentence structures 10.2.1 Non-canonical clause patterns 10.2.2 Secondary clause patterns 10.3 From text to good text - an example 11 Texts and the relevance of the situation 11.1 Spoken and written English 11.2 The interpersonal dimension - politeness The field of text linguistics does not necessarily take center stage in the linguistic sciences. However, knowing how texts work and what criteria have to be met to make a text a good text is highly relevant for the classroom. This involves an understanding of text structure and knowledge about how texts are processed as <?page no="169"?> well as about the sentence-level options that the English language provides to make a text easier to read. Finally, good texts are always appropriate for the situation in which they occur. 168 Texts <?page no="170"?> 9 What makes a text a (good) text ? (V I C T O R H U G O T O H I S P U B L I S H E R ) ! (T H E P U B L I S H E R T O V I C T O R H U G O ) What this chapter is about Although all of us have a basic idea of what a text is, it is not trivial to define the notion of text scientifically. The present chapter presents an approach that suggests seven important standards of textuality. When it comes to the question of what makes a text a good text, aspects of text structure become important. This chapter discusses two useful ways in which text structure can be described. This way the present chapter prepares the ground for the next chapter, which explores how English sentence structures can be used to create text structures that are easy to process. The text ? is a telegram Hugo sent to his publisher to find out about how his novel Les Miserables was selling. The short answer of the publisher, ! , was meant to show that sales were good. I stumbled across these two telegrams when I wanted to check something else which I recall from my Latin lessons. My teacher told me about the shortest letter ever. It was written as a father’s response to his son. The son wanted to know whether he should go to war or not. The father replied with a brief I, which is the Latin imperative form of the verb ire, meaning ‘to go’. It goes to show that letters and texts can take forms that we would usually not expect, but they would still be understood. What, then, makes a text a text? We’ll find an answer in the following section. 9.1 What makes a text a text - standards of textuality Before you read on … … take a minute and think how you would define the term ‘text’! Please write your definition down. <?page no="171"?> text cohesion Most likely your definition of ‘text’ involved features like ‘written’, maybe even ‘printed’ and something to the effect that it is longer than only one sentence. Maybe you also added something about different kinds of texts, such as short stories, novels or academic papers and your definition also contained information about the internal structure of texts, for instance that it consists of an introduction, a main part and a conclusion. It might also be that according to your definition texts can be either written or spoken, that they would usually consist of letters and numbers but might also contain icons or emojis. Maybe your definition even suggested that a text can have more than one author (as for instance the Bible) or that it can be the product of two or more authors working together simultaneously (for instance in a conversation). Maybe a song counts as a text as well, or a roadsign or even a painting or a movie. This little exercise was supposed to show that, when it comes to a linguistic definition of ‘text’, a lot of aspects need to be taken into consideration. One of the most influential definitions of ‘text’ is over forty years old and was suggested in 1981 by Robert-Alain de Beaugrande and Wolfgang Dressler: “a T E XT will be defined as a C OMMUNICATIV E O C ‐ C U R R E N C E which meets seven standards of T E XTUALIT Y ” (1981: 3), namely cohesion, co‐ herence, intentionality, acceptability, informativity, situationality, and intertextuality. Even though not all of these seven criteria are currently regarded as being defining for the concept ‘text’, they all point to important aspects of texts and will, therefore, be discussed in the following. The first standard, cohesion, has to do with the surface of the text. In a non-technical sense cohesion describes a state in which elements are connected or cling together to form a single whole. It is easy to see that with regard to text these elements are words (and also larger units) and the cohesive links between individual words make the in‐ dividual sentences of a text hang together. Take as an example the previous three sen‐ tences. They show a number of cohesive links. One link is created by the use of the word cohesion in sentences 1 and 2 and the morphologically related word cohesive in sentence 3. Similarly, the repeated use of the word text as well as elements and words contribute to our feeling that these three sentences somehow create a single unit. In addition to repetition of content words cohesion can also be (and very frequently is) created by grammatical words, in particular pronouns. An example can be found in the following two sentences from above: Take as an example the previous three sentences. They show a number of cohesive links. This is an example of what Halliday and Hasan (1976) described as cohesion by reference. The pronoun they functions as a signpost. It urges the reader to search the preceding text for a noun phrase that describes some‐ thing in the third person plural. In the above example that would be the previous three sentences. This example nicely illustrates a feature that is central to Halliday and Ha‐ san’s understanding of cohesion, namely that cohesion happens when the interpreta‐ tion of one part of a text is dependent on another part of the same text: “where the interpretation of any item in the discourse requires making reference to some other item in the discourse, there is cohesion” (11). In the example above we could understand the sentence They show a number of cohesive links even if the sentence occurred in 170 9 What makes a text a (good) text <?page no="172"?> coherence inten tionality accepta bility isolation. However, we would not be able to fully interpret the sentence if we did not have the previous sentence to find out what is meant by the personal pronoun they. While cohesion focuses on the surface of the text, the next standard of textuality, coherence, concerns what we find under the textual surface, namely the ‘world’ that the text creates. Basically, a text has coherence if it makes sense. Take a look at example (1). (1) The girl shook the tree. An apple fell down. The sentences above make sense since they can be aligned with the knowledge that we have of our world: one kind of tree is an apple tree and in autumn, when the apples are ripe, shaking the apple tree might result in apples falling down. Note that there are no cohesive ties between these two sentences, but the reader still tries to draw a connection between these two sentences if at all possible. In this case the reader will supply an interpretation of causality, i.e. because of the shaking of the tree the apple fell down. De Beaugrande and Dressler already observed that “[p]eople will supply as many relations as are needed to make sense out of the text as it stands” (1981: 4). Let us assume the second sentence in example (1) was An angry fairy fell down. This sentence does not make sense in our world, but it would make sense in a fairy story or a story placed in a fantasy setting. Even if it was not stated explicitly, readers would usually make that assumption, so that the modified version of example (1) makes sense. Cohesion and coherence are features of the text as such, which is why de Beaugrande and Dressler describe them as “text-centred notions” (1981: 7). In addition the authors suggest a number of “user-centred notions” (1981: 7) which describe standards of tex‐ tuality that concern the act of communication through text. As for the text producer, for instance, the authors suggest that a text needs to have intentionality, i.e. the pro‐ ducer of the text intends it to be cohesive and coherent and to fulfil some goal with regard to the receiver, for instance adding information to his or her stock of knowledge. The text you are reading at the moment, for example, has intentionality. I work very hard to make it cohesive and coherent, so that it can fulfil its goal, to show you that linguistics will be useful for you as a future teacher, as effectively as possible. Con‐ versely, the receiver of the text generally assumes that the text makes sense and he or she will try to find out which goal the producer had in mind with the text. In short, a text needs to be an acceptable attempt at communication from the point of view of the receiver. Acceptability also concerns the amount of effort the receiver has to put into decoding the message. Usually, the receiver will be willing to invest considerable mental (and other) resources to find the message. Just half an hour ago, for example, my wife asked me “Did you put this on here, yesterday? ” She was in the kitchen and I was in the living room, so I had no way to know either what this referred to or what here meant. Obviously, I still regarded her attempt at communication as acceptable since I got out from my comfy chair, went to the kitchen and found out that she was referring to the goat’s cheese and the small plastic ‘plate’ it comes with. While acceptability is tightly connected to the amount of effort that has to be invested on the part of the 9.1 What makes a text a text - standards of textuality 171 <?page no="173"?> informa tivity situatio nality intertex tuality receiver, the next standard, informativity, concerns the content of the message itself. It has to do with “the extent to which the occurrences of the presented text are expected vs. unexpected or known vs. unknown […]” (9). The relevance of informativity becomes particularly obvious when a text does not provide any new information. Let us assume you are sitting next to the window in one of my seminars. It is summer and it is really hot. If I say to you “It’s really hot in here”, you will know immediately that this text does not have a high degree of informativity. After all, we are in the same room together. Still, you would assume that I intend this text to be informative for you and you will search for the information that I want to transmit and maybe conclude that I want you to open the window. We can stay with this example to illustrate the next standard of textuality, namely situationality. Situationality demands that a text be appropriate for the situation in which it is used. The reason why I would choose the indirect “It’s really hot in here” over the more direct “Open the window, please” might be found in the situation in which this text is uttered. For instance, I might consider the direct order a little impolite and, given the hierarchical assymmetry between me (as professor) and you (as student), accordingly, opt for the more indirect and, hence, more polite state‐ ment (see section 11.2 for more details on this). The last standard of textuality, inter‐ textuality, can be understood in two ways. Firstly, it can relate to the relations that a text has to already existing texts. For example, if you write a paper about Hamlet’s madness, your paper will be intertextually related to Shakespeare’s tragedy Hamlet. Whoever reads your paper can only fully assess its content if they know Shakespeare’s text. Speaking of Hamlet and, by association, of Shakespeare: In the video game God of War: Ragnarök, Mimir tells a story that shows very strong intertextual ties to Shake‐ speare’s Macbeth. Not only does the recognition of these ties make you feel well edu‐ cated, it also allows you to come up with hypotheses about the period of time in which the story of the video game itself is set. The last research paper I wrote was called “Make political discourse rational again”. Everyone who reads this title might have some idea of what to expect in the paper. However, the intertextuality of the title, i.e. its relation to former US President Trump’s campaign slogan “Make America Great Again”, would help to provide a clearer idea of the contents of the paper. In addition to this first kind of intertextuality, the term can also be understood as referring to the fact that texts are generally processed as an instance of a particular text type. For example, if you know a piece of writing is a research article in a scientific journal you will approach this text with particular assumptions and you will read it differently than the newest novel by your favourite author. Not all of the above standards are necessary conditions that have to be met to make a text a text. Some scholars claim that only the standard of coherence needs to be met for a piece of language use to qualify as a text. Nevertheless, all of the seven standards point to important features that texts might show to varying degrees and they will serve as a guideline for the remainder of this chapter not only to describe what makes a text a text but also to describe what makes a text a good text. We will come back to individual standards at various points in what follows. 172 9 What makes a text a (good) text <?page no="174"?> the impor‐ tance of background knowledge 9.2 The relevance of text structure Before you read on … … please read the following text and try to make sense of it. The procedure is actually quite simple. First you arrange things into different groups depending on their makeup. Of course, one pile may be sufficient depending on how much there is to do. If you have to go somewhere else due to lack of facilities that is the next step; otherwise you are pretty well set. It is important not to overdo any particular endeavour. That is, it is better to do too few things at once than too many. In the short run this may not seem important, but complications from doing too many can easily arise. A mistake can be expensive as well. The manipulation of the appropriate mechanisms should be self-explanatory, and we need not dwell on it here. At first the whole procedure will seem complicated. Soon, however, it will become just another facet of life. It is difficult to foresee any end to the necessity of this task in the immediate future, but then one can never tell. The text above is taken from Bransford & Johnson (1972), who describe an experiment where three groups of test subjects listened to the text above and later had to recall the content of the text. One of the groups were told before what the text will be about, the second group was told what the text was about after they had heard the text and the third group was not told what the text was about (similar to you). The results showed that the first group was much better in recalling the text than the other two groups. This should not surprise you: after all, I assume you have found it difficult to make sense of the text in the first place, let alone recall it. I use this little text a lot in my university teaching to illustrate the importance of background knowledge. What often happens when I tell my students that the text is about washing clothes is that you can almost see how suddenly things start to fall into place and how the whole text starts to make sense. I assume that you have just had this very same experience. The information that the text is about washing clothes activates knowledge structures that exist in your mind and the structures help you interpret the ambiguous or vague information presented in the text itself. In another experiment Anderson et al. (1977: 372) gave their informants the following text: (2) Rocky slowly got up from the mat, planning his escape. He hesitated a moment and thought. Things were not going well. What bothered him was being held, especially since the charge against him had been weak. He considered the present situation. The lock that held him was strong, but he thought he could break it. This time you might not find it that difficult to make sense of the text but notice how the story changes if you read it under the title “A prisoner plans his escape” as opposed to 9.2 The relevance of text structure 173 <?page no="175"?> the title “A wrestler in a tight corner”. Many of the words used in the example (2), such as mat, being held, charge and lock, are ambiguous and they are disambiguated differently depending on the title. Again, this goes to show that background information is very important when it comes to interpreting text. How is all of this relevant when it comes to writing a good text? When I talk of text in the present context I think of informative or argumentative texts, i.e. texts like a seminar paper at university or texts that your students will write in their later grades in school. The aim of such texts is to increase the amount of information or knowledge in the reader’s mind. However, the reader’s mind is not empty but it already contains some pieces of information. If at all possible, your text should try to connect to this stock of knowledge. This will make it a lot easier for the reader to interpret and understand what you want to tell him or her with your text. The experiments above have shown that the activation of previous knowledge structures have a huge impact on the interpretation and understanding of the whole of the text. However, and this is where linguistics comes in, these insights are also relevant when we consider the individual sentences that make up a text. As the reader reads your texts, each sentence that they read (hopefully) becomes part of their stock of knowledge. Each new sentence in your text should link up to the readers’ existing stock of knowledge in the best way possible. In a way a text can be compared to a situation where you, as the author, take your readers by the hand and guide them through the textual world that you create and in the process draw their attention to those parts of the textual world that you think are especially important. If a text is written well, the reader will never lose track of what it is you want to show them. Linguistics provides us with the means to describe texts and text structures in general. To stay in the image of the author taking the reader by the hand, linguistics helps to give us a general idea of what a good tour through the textual world will look like. In addition, linguistics can also be helpful when it comes to forming our sentences in a way that they will be easy to understand for the reader while at the same time making it easy to recreate in their mind the textual world that you created for them. So, in the following we will focus on two aspects: the remainder of this chapter will discuss text structure and ways of creating text structure that facilitates comprehension. The next chapter, then, will focus on how we can form sentences in such a way that your texts can be understood easily. 9.2.1 Themes and rhemes The idea that texts have a particular structure is most likely not new to you. However, non-linguistic approaches to text structure are usually of a broadly functional type, in the sense that they break texts down into fairly large bits, such as ‘introduction’, ‘body’ and ‘conclusion’ or ‘introduction’, ‘previous research’, ‘methods and data’, ‘results’, ‘discussion’ and ‘conclusion’. Knowledge of these text structures can be very helpful both, when reading a text and when writing a text. For example, it often makes sense 174 9 What makes a text a (good) text <?page no="176"?> theme and rheme thematic progression linear progression to first take a look at the introduction and the conclusion of an academic research paper to find out whether this paper really is relevant for you. In addition, a first look at the introduction and the conclusion will help you create some relevant background knowledge in your mind which then makes it easier for you to understand the rest of the text. In the following section I would like to take a look at text structure from the per‐ spective of the sentence, i.e. I want to explore how sentence-level resources can help to create an easy-to-read text. A very basic distinction with every sentence is that between theme and rheme. The theme is what a sentence is about and often contains known information, the rheme is what we learn about the theme and would usually consist of new information. In example (3), for instance, Peter is the theme and bought a new car yesterday is the rheme, i.e. the sentence is about Peter and we learn about Peter that he bought a new car yesterday. (3) Peter bought a new car yesterday. Prague School linguist František Daneš (1974) applied this distinction to the description of text structure by looking at what he called ‘thematic progression’, i.e. the way themes develop within a text. The first basic type of thematic progression, linear progression, is illustrated in example (4). (4) 1. Peter bought a new car yesterday. - 2. The car is red. - 3. Red is the favourite colour of his girlfriend. - 4. She prefers going by bike. With linear progression the rheme of a given sentence (or a part of it) becomes the theme of the next sentence. The rheme of sentence 1 (R1), for instance, is bought a new car yesterday. A part of this rheme, namely the car, serves as the theme of sentence 2 (T2). Part of R2, red, becomes T3 and so on. The linear progression of example (4) is shown in figure 9.1. The theme and rheme for each sentence are in angular brackets. The bottom of figure 9.1 shows the development of themes and rhemes in a net‐ work-like fashion. As you see, the thematic progression forms one line, hence the term ‘linear progression’. 9.2 The relevance of text structure 175 <?page no="177"?> constant theme Fig. 9.1: An example of linear progression. The second major type of progression is that of ‘constant theme’, where all sentences of a text share the same theme. It is illustrated in example (5) and figure 9.2. (5) 1. Linguistics is the scientific study of language. - 2. It is one of the main study areas of English as a foreign language. - 3. It is often regarded as more difficult than other areas. - 4. Linguistics is similar to maths in that it makes use of many abstract concepts. Fig. 9.2: An example of constant theme progression. Usually, texts will show a mixture of both kinds of thematic progression as you will see in the exercise below. 176 9 What makes a text a (good) text <?page no="178"?> Before you read on … … take a look at the text below, adapted from a passage of J.R.R. Tolkien’s Silmarillion. For each of the sentences identify the theme and the rheme and provide a description of the whole text in terms of thematic progression. Please note that sometimes it makes sense to divide the rheme of a sentence into two sub-rhemes, for example when Tolkien first mentions Quenya and Sindarin. Each of the two languages will become themes of their own further below in the text. In addition, sometimes it makes sense to not get too fixated on sentence structure. The first sentence that has Quenya as its theme, for instance, can be divided in two (4. and 5.) as far as theme and rheme is concerned. 1. The Elves became divided into the West-Elves (the Eldar) and the East-Elves. 2. The languages of the latter kind are not important here. 3. Of the Eldarin tongues two are found in this book: Quenya and Sindarin. 4. Quenya was an ancient tongue of Eldamar, 5. the first to be recorded in writing. 6. It had become, as it were, an ‘Elven-Latin‘. 7. Sindarin was in origin akin to Quenya; 8. for it was the language of those Eldar that had lingered in Beleriand… To get an idea of how the themes are used and develop in the text above it is first important to identify the theme and rheme for each of the sentences. This is shown in figure 9.3. Fig. 9.3: Themes and rhemes in the Silmarillion text. 9.2 The relevance of text structure 177 <?page no="179"?> idea map As you can see in the figure above, natural texts are usually a little more complex. In the example above the rhemes of sentences 1 and 3 introduce two concepts each of which become themes in the following text. The Eldar of sentence 1, for instance, become the theme of sentence 3 and the East-Elves of sentence 1 become the theme of sentence 2. The rheme of sentence 3 can be divided into Quenya and Sindarin, which become the constant themes for three (4 to 6) and two clauses (7 and 8), respectively. As a consequence, the network representation of the thematic progression of this text is more complex than the one for the previous examples. As can be seen at the bottom of figure 9.3 it contains elements of both, linear and constant-theme progression. When it comes to working with texts an awareness of thematic progression will be useful as it increases our awareness of the way in which themes (or topics) develop and unfold in a text. This will be very helpful when reading a text. To stay with our image of the textual world: you will know exactly which object of the textual world you are currently looking at and, accordingly, you will know exactly where to store the new information that the text provides. While writing a text yourself, the awareness of thematic progression will help you to keep track of the aspects that you want to tell the reader about. It can keep you from being sidetracked or from skipping from one topic to the next and back. Or, in terms of the metaphor of the tour guide: you lead your tourists to one particular sight, tell them everything you want to tell them about that sight and then move on to the next. You might be familiar with the concept of an idea map or idea web. It is an approach that supports the planning of a text by organising all the different aspects that are relevant to a given topic. Each of the nodes in your idea map would usually figure as themes or rhemes (or both) in a thematic progression network. Figure 9.4 shows what an idea map for the text on the Elven languages might look like. Fig. 9.4: An idea web of the Silmarillion text. As you can see, each parent node in your idea web, i.e. each node that points down to another node, figures as a theme in the theme-rheme network of the text. Each 178 9 What makes a text a (good) text <?page no="180"?> hypertheme child node is introduced as (part of) the rheme of a sentence at some point. If a child node has children of its own, the node will be the theme of one or more sentences. Those nodes in the idea web that do not have children of their own will only figure as rhemes in the thematic progression. With regard to the seven standards of textuality discussed in the previous section it is the connections between those nodes and the corresponding connections between themes and rhemes that provide cohesion to a text. These connections make clear how the individual sentences of a text are connected to form one unit. So far, the thematic progression approach hinged on the recurrence of concepts in a text. For example, Quenya is part of the rheme in sentence 3 and becomes a theme of its own afterwards. The relevant connections can be drawn easily since Quenya is mentioned (or can be guessed) in all of the pertinent sentences and structures. This is also why it ties up so nicely with the idea of cohesive links in a text. However, as we have seen above, texts might not show any signs of cohesion but they might still make sense, i.e. have coherence. Daneš’ (1974: 120) concept of ‘hypertheme’ can account for this. Hyperthemes are not explicitly given in the text itself but they can be derived from the text. Before you read on … … let’s take a look at hyperthemes. The process of identifying a hypertheme of a stretch of text could be compared with guessing the heading of a text while reading the text. What could a possible heading for (and, hence, the hypertheme of) the following text (Daneš 1974: 120) be? 1. New Jersey is flat along the coast and southern portion. 2. The north-western region is mountainous. 3. The climate is usually very mild. 4. The leading industrial production includes chemicals and coal. 5. The most important cities are Newark, Jersey City, Camden … Possible hyperthemes that might be derived from the five sentences above include ‘The geography of New Jersey’ or ‘Geographical facts about New Jersey’ or something along those lines. In Daneš’ model, the individual sentence themes would be linked to the derived hypertheme and each of them would lead over to its individual rheme, as shown in figure 9.5. 9.2 The relevance of text structure 179 <?page no="181"?> nucleus and satellite Fig. 9.5: An example of hypertheme. With Daneš’ idea of hyperthemes we can identify connections between sentences even though no obvious links are given. However, in some cases, even hyperthemes will fail to account for linking which is apparent to the reader. This can be illustrated in example (1) here repeated as (6) for convenience. (6) The girl shook the tree. An apple fell down. The two sentences do not show cohesion but we would still draw a connection between the two, which results in the two sentences being a coherent text. In our understanding of the world surrounding us we can see a relation of causation between the two sentences: because the girl shook the tree, the apple fell down. The thematic-progression approach cannot account for that, but Rhetorical Structure Theory can. 9.2.2 Rhetorical Structure Theory Rhetorical Structure Theory (RST) is an approach to text structure, developed by Mann and Thompson (1988; see the RST website for an accessible introduction: www.sfu.ca/ rst/ index.html), that is particularly useful to describe the coherence of texts (beyond obvious cohesive ties). The basic unit is that of the clause. So, even if example (6) consisted of one sentence only, e.g. After the girl shook the tree an apple fell down, RST would still break this sentence down into its two clauses and try to determine how the two relate to each other. A basic distinction they make is that between ‘nucleus’ and ‘satellite’. The nucleus is the main message and the satellite ‘supports’ the nucleus. ‘Support’, here, is understood in a very wide sense, basically expressing that the satellite 180 9 What makes a text a (good) text <?page no="182"?> is functionally subordinate to the nucleus. For instance, in example (6) above, the main message of the little text is that an apple fell down. So, the second clause would be the nucleus of the text. The first clause ‘supports’ this main message in that it provides the cause for the apple falling down. In the RST framework relations of this kind are rep‐ resented in diagrams like the one shown in figure 9.6. By the way: Clauses and sentences At this point a little refresher might be in order regarding the distinction of clauses and sentences. A ‘clause’ is a configuration of words that contains at least a subject and a predicate but may also contain other obligatory and/ or optional elements. For instance, Yesterday, the girl shook the tree contains a subject (the girl), a predicate (shook) and another obligatory element called the direct object (the tree). In addition, the clause also has an optional adverbial (Yesterday). Which elements are optional or obligatory depends on the main verb of the clause, i.e. with the verb to shake you need someone that does the shaking and someone/ something that is being shaken. Information about when the shaking happens is optional. The clause above can stand on its own and would, therefore, also qualify as a sentence, more specifically a ‘simple sentence’. Clauses can also have functions within another clause, for example, Because the girl was hungry, she shook the tree. In cases like these we speak of a ‘complex sentence’. It consists of at least two clauses, one of which serves as a constituent in the other. In the example above, the first clause provides a reason and, thus, serves as an adverbial within the second clause. A ‘compound sentence’, in contrast, consists of two or more clauses that are on the same level of hierarchy, e.g. The girl shook the tree and an apple fell down. Fig. 9.6: An example of an RST diagram. The top-most node, in this case [1-2], represents the text as a whole. Each text, according to RST, can be reduced to one major message. In the present case, this would be the content of clause 2, i.e. the fact that an apple fell down. The (central) nucleus in RST diagrams is the one clause that all other clauses point to. In the example above, 9.2 The relevance of text structure 181 <?page no="183"?> there is only one other clause. It is the satellite to the nucleus and describes what caused the apple to fall down. RST provides a large number of relations that can hold between nucleus and satellite. A good overview can be found on the RST-website. Table 9.1 below (based on www.sfu.ca/ rst/ 01intro/ intro.html) provides a list of the most relevant relations for informative/ argumentative texts. Relation Nucleus Satellite antithesis ideas favoured by author ideas disfavoured by author background text whose understanding is being facilitated text for facilitating understanding cause a situation another situation which causes that one circumstance text expressing the events or ideas occurring in the interpretive context an interpretive context of situation or time concession situation affirmed by author situation which is apparently incon‐ sistent but also affirmed by author condition action or situation whose occurrence results from the occurrence of the conditioning situation conditioning situation elaboration basic information additional information evidence a claim information intended to increase the reader’s belief in the claim purpose an intended situation the intent behind the situation result a situation another situation which is caused by that one solutionhood a situation or method supporting full or partial satisfaction of the need a question, request, problem, or other expressed need Table 9.1: An overview of important relations between nucleus and satellite. When trying to identify nucleus-satellite relations in texts, it helps (at least at first) to stick to the wording provided in the table above. With regard to example (6) above, for instance, we see that the cause relation holds: the nucleus describes a situation and the satellite refers to another situation that caused the situation represented in the nucleus. This makes the second sentence (An apple fell down) the nucleus and the first sentence (The girl shook the tree) the satellite. 182 9 What makes a text a (good) text <?page no="184"?> Before you read on … … provide a description of the text below in terms of Rhetorical Structure Theory. That is, identify the one clause that is the nucleus (the central message) of the whole text and identify the satellites and their relation to the nucleus. Linguistics is a difficult subject for most first-year students. Most of the linguistic content has never been dealt with in school and linguistics often deals with highly abstract concepts. That is why, at first, some student may feel a little frustrated. Since RST wants to describe relations between clauses, it is necessary to identify the clauses first. This is simple: the first sentence and the last sentence are simple sentences, i.e. they consist of one clause each. The second sentence consists of two clauses, i.e. Most of … school and linguistics often … concepts. What is less simple is to identify the nucleus of the whole text. Actually, I think that two interpretations are viable. Either you consider the first clause as the main message, or you interpret the last clause as the central nucleus. Both is possible but it results in different RST-diagrams as you can see in figure 9.7. Fig. 9.7: Alternative analyses in Rhetorical Structure Theory. In the left part of figure 9.7, all arrows and lines in the diagram lead to clause 1, since this analysis interprets clause 1 as the main message and analyses the rest of the text as supporting information. Clauses 2 and 3 describe the cause(s) that lead to the situation described in 1 and clause 4 describes a result of the situation described in the first three clauses. We could sum the text up as follows: Linguistics is difficult (1) because of the situation described in (2) and (3). As a result of this situation students feel frustrated (4). The analysis on the right of figure 9.7 treats the last clause as the main message and, accordingly, all arrows and lines lead to clause 4. Clauses 1-3, now, become supporting satellites providing the cause for the situation described in 4. 2 and 3, just like before, 9.2 The relevance of text structure 183 <?page no="185"?> are in a cause relationship to clause 1. We could say students are frustrated (4) because linguistics is difficult (1), which is because of the situation described in (2) and (3). In addition to nucleus-satellite relations RST also suggests relations between nuclei. In a way this is similar to what you already know from traditional grammar and the distinction between hypotaxis and parataxis. Both terms are of Greek origin. Taxis relates back to the Greek term for order or arrangement. Together with the prefix hypo-, which as you may recall means ‘under’, the term describes a relation of subordination, in this case between clauses, e.g. a main clause and a causal subordinate clause as in an apple fell down because the girl shook the tree. The prefix para-, on the other hand, means ‘next to’ or ‘beside’. In the present context the term ‘parataxis’ denotes an arrangement of clauses, one next to each other without any of the two being superior or inferior. An example would be the girl shook the apple tree and the boy shook the pear tree. In RST relations of this kind are termed ‘multinuclear relations’. Three important ones are provided in table 9.2. Relation Nucleus 1 Nucleus 2 Symbol -contrast - -one alternate - -the other alternate - -joint - -(unconstrained) - -(unconstrained) - -sequence - -an item - -a next item - Table 9.2: Three important multinuclear relations in RST. The joint relation is exemplified in the exercise text (figure 9.7) above (albeit not made explicit in diagram, but it will be in figure 9.8 below). Clauses 2 and 3, together, provide causes for the situation described in 1. None of these causes is in any way superior to the other so that we are dealing with two nuclei. The example shown in figure 9.8 also illustrates the other multinuclear relations. The whole text shows a basic multinuclear relation of contrast. Clauses 1 to 7 make a statement about linguistics and contrast this with the situation of literary studies 184 9 What makes a text a (good) text <?page no="186"?> in clauses 8 to 9. With regard to literary studies the main claim is that students are less frustrated (9). The cause for that is given in clause 8. As for the part relating to linguistics, clauses 2 and 3 now show the joint relation from table 2. Together they provide the cause for the perception that linguistics is difficult (1). Clauses 1 to 3, as before, are the cause for a feeling of frustration on the part of the students expressed in clause 4. Taken together, the first four clauses describe a problem to which clauses 5 to 7 offer a solution. This solution consists of three successive steps, which is expressed through the sequence relation in figure 9.8. Note that because of the way that RST defines the solutionhood relation, the solution to a problem is always the nucleus to the problem, which is the satellite. Hence, the diagram in figure 9.8 makes the solution the main message of the whole text. Fig. 9.8: RST analysis with multinuclear relations. 9.2 The relevance of text structure 185 <?page no="187"?> The discussion of the two approaches, thematic progression and Rhetorical Structure Theory, is primarily meant to raise your awareness of the different kinds of structures a text might have. I do not think it is necessary that you create a thematic progression diagram for each text that you write, neither is it necessary to analyse all of your texts with regard to RST. However, being aware of these two aspects of text structure can be helpful while you are working on a text. Thematic progression can be supportive when it comes to making sure that your text is cohesive, i.e. it makes clear how the individual sentences hang together and form a whole. RST, in contrast, “offers an explanation of the coherence of texts” (RST website): One formulation of coherence is that it is the absence of non-sequiturs and gaps. That is, for every part of a coherent text, there is some function, some plausible reason for its presence, evident to readers, and furthermore, there is no sense that some parts are somehow missing. I find this formulation very enticing, particularly because often when I read student papers I come across a sentence for which I find it hard “to see a plausible reason for its presence”. I am not always convinced that the writer could tell me what exactly the function of this particular sentence would be. If the writer cannot answer this question for each of their sentences, their text does not do too well with regard to the standard of intentionality (and coherence for that matter). For me as the reader, the text loses acceptability. An awareness of relations between clauses in terms of RST, I believe, can support the whole writing process, thereby increasing the overall coherence, intentionality and acceptability of the text at issue. Just give it a try! New tools in your toolkit Standards of textuality ▶ Your awareness of the seven standards of textuality give you an idea of the many different ways in which a text can be good (or bad). Theme and rheme ▶ Your understanding of a general bipartition of a clause into theme and rheme provides you with a way of describing and assessing text structure with regard to cohesive links. Rhetorical Structure Theory ▶ This approach to the description of text coherence helps you to keep in mind that every clause within a text should have a function. In addition, RST provides you with a method to improve your own text-writing skills and with a way to assess the coherence of the texts of your students. 186 9 What makes a text a (good) text <?page no="188"?> If you want to know more D E B E A U G R A N D E , R. A., & D R E S S L E R , Wolfgang U. (1981): -Introduction to Text Linguistics-(Vol. 1). London: Longman. ▶ A very influential classic that provides an in-depth exploration of the seven standards of textuality. E S S E R , Jürgen (2009): -Introduction to English Text-linguistics. Frankfurt a.M.: Peter Lang. ▶ Discusses major concepts of text linguistics, such as cohesion and coherence, questions of presentation and psycholinguistic and cognitive aspects of texts and text processing. H A L L I D A Y , Michael A.K. & Ruqaiya H A S A N (1976). Cohesion in English. London: Routledge. ▶ Another very influential and insightful classic that provides a comprehensive account of the different linguistic means that contribute to the cohesion of a text. www.sfu.ca/ rst/ ▶ Provides a very good introduction to Rheotrical Struture Theory. Additional online content. 9.2 The relevance of text structure 187 <?page no="190"?> 10 Texts and the relevance of sentence structure No, please stop! You cannot teach them that! That’s wrong! (A T E A C H E R I N T H E B A C K O F H E R C L A S S W H E N I W A S T E L L I N G H E R P U P I L S A B O U T N O N - C A N O N I C A L W O R D O R D E R ) What this chapter is about In the last chapter we have seen how we can describe the structure of texts. This chapter explores how we can create a text structure which makes it easy for the reader to follow the writer’s train of thought. This involves, first, a discussion of the principles that govern text processing and, second, a look at how sentences can be formed that are in line with these principles. Throughout this chapter we will be working on the same text, which will get better and better as we move towards the end of this chapter. Knowing about ways to describe text structure is one thing, it is another thing to have an awareness of good text structure and still another thing to know how we can create a text that has a good structure. The text in the following task will accompany us through the whole chapter so, make sure that you get the most out of this first task. Before you read on … … take a look at the text below. With regard to text flow, how would you rate the text? What do you think could be improved? a. That linguistics is a difficult subject for most first-year students becomes apparent fairly early in their first semester. b. There are three reasons for this. c. The fact that linguistics is usually not taught at school is the first reason. d. In this respect, literature or cultural studies have a more privileged status. e. Teachers at school teach these aspects and the syllabus lists them as important. f. The fact that students have to learn many new terms in linguistics is related to the first reason. g. Because students do not have any experience with linguistic terminology and ap‐ proaches they often feel a lot of pressure to get a grip on all the new concepts. <?page no="191"?> h. The abstract nature of linguistics is another reason. i. Some students are reminded of maths by linguistics. j. Often they are scared off by this. k. To include aspects of linguistics into English language teaching at school would be a promising way of addressing these three problem areas. Reading the text above you might get the feeling that it is a little ‘bumpy’ and does not read as smoothly as it might. For instance, sentence b mentions three reasons and sentence c names the first reason but it does not tell us until we have reached the last three words of the sentence. We are told something about the fact that linguistics is usually not taught at school, but we are not told what to do with this information until we have reached the end of the sentence. Compare this to the alternative version provided below. (1) (b-c) There are three reasons for this. The fact that linguistics is usually not taught at school is the first reason. - (b-c’) There are three reasons for this. The first reason is the fact that linguistics is usually not taught at school. After having read the first three words of sentence c’ we already know where the information provided is supposed to go. In terms of thematic progression we see that the theme of sentence c is now a lot closer to the rheme of sentence b, where three reasons are introduced. Again, if we stay with our metaphor of a guided tour, what you do is essentially to first point your visitor to a particular sight, let’s say a building, and then tell them what you want them to know about that building. It does not make much sense for your visitor to listen to you if they do not know what it actually is you are talking about. If we look at the textual world that the text in the exercise is building we can say that the first sentence introduces some state of affairs or some problem. The second sentence then tells us that there are three reasons for this problem. An abstract representation of the textual world you have created so far can be seen on the left-hand side in figure 10.1. Fig. 10.1: The textual world after having read sentences a and b (left) and after having read sentence c (right). The right-hand side of figure 10.1 shows the textual world after sentence c. The original version of the text provides you with the information that linguistics is not taught at 190 10 Texts and the relevance of sentence structure <?page no="192"?> linking: known infor‐ mation school, but you do not know where to integrate that information in the textual world until you have read the whole sentence. It is only at the end of sentence c that you are told that “this is the first reason”. The alternative version in (1b-c’), on the other hand, tells you at the beginning of sentence c’ where this piece of information has to go with regard to the textual world: “The first reason is …”. This way it makes it easier for the reader to re-create the textual world that you have in mind. With regard to our discussion of textuality in chapter 9 we are talking about an increase in acceptability. Much of what we will discuss in the following can be viewed in this light when we discuss general principles of sentence and text processing. 10.1 Principles of text processing 10.1.1 The given-before-new principle We have already touched the given-before-new principle in our discussion of thematic progression, when it was stated that the theme typically contains old or known or given information while the rheme typically contains new or unknown information. The general idea is almost 160 years old and was suggested by the French-German linguist Henri Weil, who claimed that it is […] necessary to lean on something present and known, in order to reach out to something less present, nearer, or unknown. There is then a point of departure, an initial notion which is equally present to him who speaks and to him who hears, which forms, as it were, the ground upon which the two intelligences meet; and another part of discourse which forms the statement […], properly so called. ([1844] 1978: 29) Similar observations have been made by a number of other authors under a range of different names, e.g. the distinction of ‘psychological subject’ and ‘psychological predicate’ by Gabelentz (1901: 365-373), Behaghel’s (1932) Second Law, which claims that “[t]hat which is less important (or known to the reader) is placed before that which is important”, or in the Prague School idea of Functional Sentence Perspective. Yet another formulation of basically the same insight comes from Clark and Clark (1977), who describe it in terms of what they call the ‘given-new contract’: The speaker agrees (a) to use given information to refer to information she thinks the listener can uniquely identify from what he already knows and (b) to use new information to refer to information she believes to be true but is not already known to the listener. (92) In all cases the general idea is that clauses typically start with something that is known and proceed to something that is unknown, since this arrangement of information makes it easier for us to process a text. This is illustrated by the modified sentence c in example (1). In contrast to the original version, this version starts with something that is known, namely the first reason. After having read the first two sentences, the writer 10.1 Principles of text processing 191 <?page no="193"?> linking: points on a scale linking: sense relations and the reader share a common ground, namely that fact that there is a problem which can be explained by three reasons. The modified version of sentence c, begins with something that is part of this common ground, as it is part of the knowledge shared by writer and reader and then provides the reader with new information. The writer first provides a link and then tells the reader what should be attached to this link. This makes it a lot easier for the reader to follow than the original version of the text. A final comment is in order here regarding the nature of the link to the previous discourse. These can be established in ways other than repeating fully known infor‐ mation. Consider example (2) (taken from Kreyer 2010: 179): (2) An exemption clause as a term of the contract intended to exclude or restrict the liability of one of the parties, usually the seller. Sometimes a seller is so financially powerful that he can blatantly insist on such a clause being included in the contract; he can adopt a “take it or leave it” attitude. More common is the seller who, having inserted an exclusion clause into his conditions of sale, relies on his buyer not bothering to read (or not understanding) the small print; … (BNC-H7U: 310-312; my emphasis) In the example above the last sentence does not begin with given information in the strict sense. Nowhere in the previous text do we find a mention of another thing that is “more common”. However, we do find the adverb sometimes in the previous sentence. More common, just like sometimes, describes some kind of frequency with which something happens or something exists. The link, in this case, would exist in the fact that both expressions designate points on a scale of frequency (cf. Kreyer 2004: 75-80 for details). This will allow the reader to integrate the new information provided in the last sentence of example (2) in a meaningful way. Similarly, many of the sense relations explored in section 5.3 allow for establishing links between sentences without verbatim repetition of elements from the previous discourse. (3) Yesterday was my first day at university. My professors are really nice. Again, in a strict sense the second sentence contains all-new information, but since professors are known to be members or parts of a university, the noun phrase my professors can be understood as given information and a link between the second and the first sentence can be established by the recipient. As we have seen in our discussion of acceptability in section 9.1, readers are usually willing to invest considerable amounts of effort to make a message meaningful. With regard to the present discussion this means that readers will usually establish a link to the previous discourse if this is at all possible. Before you read on … … and to conclude this section, take another look at the text at the beginning of this chapter. Which of the sentences do not begin with known information or do not provide a link to the previous discourse at the beginning of the sentence? 192 10 Texts and the relevance of sentence structure <?page no="194"?> end-focus and given-be‐ fore-new a. That linguistics is a difficult subject for most first-year students becomes apparent fairly early in their first semester. b. There are three reasons for this. c. The fact that linguistics is usually not taught at school is the first reason. d. In this respect, literature or cultural studies have a more privileged status. e. Teachers at school teach these aspects and the syllabus lists them as important. f. The fact that students have to learn many new terms in linguistics is related to the first reason. g. Because students do not have any experience with linguistic terminology and ap‐ proaches they often feel a lot of pressure to get a grip on all the new concepts. h. The abstract nature of linguistics is another reason. i. Some students are reminded of maths by linguistics. j. Often they are scared off by this. k. To include aspects of linguistics into English language teaching at school would be a promising way of addressing these three problem areas. There are many sentences in the text that do not provide a link at the beginning, more specifically, sentences b, c, e, f, h, i, j and k. This is not surprising since the text was especially written to exemplify bad writing. However, this is not the only way in which the text is not optimal. There are two more principles that are relevant when it comes to writing a text that is easy to process and follow. 10.1.2 The principle of end-focus The principle of end-focus suggests that we place the most important information at the end of the sentence, since this is a place of particular prominence when it comes to sentence processing. This does not come as a surprise if we assume that language users follow the given-before-new principle. Given information cannot be particularly important to the reader since the reader already knows about this particular piece of information. New information, on the other hand, is important to the reader by the mere fact that the reader does not yet know about it. In a way language users are trained to expect important information towards the end of the clause or sentence. When it comes to writing a text, then, in a majority of cases, sticking to the given-before-new principle will automatically result in following end-focus. This is probably why Beha‐ ghel’s (1932) Second Law treats them as one: “[t]hat which is less important (or known to the reader) is placed before that which is important.” Still, the principle deserves a discussion in its own right. This becomes apparent if we look at sentences which occur at the beginning of a chapter, where we do not yet have any real kown information. The following example is taken from a text in the context of law. (4) POLICE MISCONDUCT Remedies for unlawful arrest and detention - A basic principle of the rule of law is that any interference with the liberty of the individual must be justified by law. Simply by virtue of his official position no police 10.1 Principles of text processing 193 <?page no="195"?> officer has the right to interfere with the person’s liberty unless he can point to legal authority to justify his actions. If a person is detained irregularly or his property or person searched without lawful authority, he has the following remedies: … (BNC-EVK: 1316-1321) The first sentence in the example above is an instance of full-verb inversion, a clause pattern that is non-canonical and would usually be regarded as erroneous in the school context. Note that the sentence is more or less all-new. This and its position immediately after the heading rules out the given-before-new principle as a motivation to choose a non-canonical clause pattern. The principle of end-focus, however, can provide an explanation. The clause pattern serves to put the phrase any interference with the liberty of the individual must be justified by law in the final position of the clause where it is given extra prominence. This makes a lot of sense since the following sentences discuss the liberty of the individual in more detail. This way, the first sentence serves to introduce a topic into the discourse, which is then explored for a portion of the rest of the text. In creative writing, end-focus can be used to introduce major characters. This is the typical way of starting a fairy story, e.g. once upon a time there was king. The example below shows the beginning of the narrative A Tale of Annabelle Hedgehog. Note how the chapter heading introduces the beast, which, therefore, is already given information. Still, the first sentence puts the beast in final position, thereby introducing it as the topic for the following sentences. (5) The Awful Beast - Standing before her across the water was the most enormous beast Anabelle had ever seen. It had large powerful muscles covered by a sleek golden coat, a massive boxy head with two huge floppy ears on either side, and a broad muzzle with a grand black nose, and two big, quick brown eyes. Its feet were the size of dinner platters, and its tail was as long and heavy as an oak limb. A vast pink tongue was hanging out of the creature’s mouth between a pair of the longest, sharpest teeth imaginable. It was a spectacular creature, an amazing animal, a monster by any measure. (BNC-CFJ: 1-6) The relevance of end-focus also becomes apparent if a clause, in terms of thematic progression, has a large and complex rheme which allows for alternative orders of elements. Before you read on … … take a look at the text below. How would you arrange the three reasons in the second sentence (This is due to …) if the rest of the text continued as shown under a), b) or c)? The first weeks at university are often tough for students. This is due to a change of loca‐ tion, a change of social environment and study programmes that might be overwhelming at times. 194 10 Texts and the relevance of sentence structure <?page no="196"?> a) The mere act of moving in itself can already be stressful. This text explores ways to deal with this and related problems. … b) Students will have to say goodbye to most of their friends. This text explores ways to deal with this and related problems. … c) Even in those cases where programmes seem to be identical to subjects in school, the approach to the same subject will usually be very different. This text explores ways to deal with this and related problems. … It makes sense to put the one reason at the end of the sentence that serves as the topic for the remainder of the whole text. If the text continued as shown in a), it would make sense to put a change of location at the end of sentence 2, in the case of b) sentence 2 should conclude with a change of social environment, and everything can remain as it is in the case of c). Before you read on … … as before, take another look at the text at the beginning of this chapter and identify those sentences that do introduce a new aspect into the text but do not follow the end-focus principle. a. That linguistics is a difficult subject for most first-year students becomes apparent fairly early in their first semester. b. There are three reasons for this. c. The fact that linguistics is usually not taught at school is the first reason. d. In this respect, literature or cultural studies have a more privileged status. e. Teachers at school teach these aspects and the syllabus lists them as important. f. The fact that students have to learn many new terms in linguistics is related to the first reason. g. Because students do not have any experience with linguistic terminology and ap‐ proaches they often feel a lot of pressure to get a grip on all the new concepts. h. The abstract nature of linguistics is another reason. i. Some students are reminded of maths by linguistics. j. Often they are scared off by this. k. To include aspects of linguistics into English language teaching at school would be a promising way of addressing these three problem areas. The whole text is about the fact that linguistics is difficult for most first-year students. Accordingly, it would make sense if the first sentence, following the principle of end-focus, put this piece of information in final position. The same holds true for sentence b, which introduces the three reasons, but does not follow the principle of end-focus. Similarly, sentences c, f and h would make the text more readable if each of the three reasons occurred in final position of the clause. Finally, the concluding sentence would stress the main point of the whole text (include aspects of linguistics into English language teaching at school) more, if it deferred it to the end of the clause. 10.1 Principles of text processing 195 <?page no="197"?> dative alter‐ nation heavy-NP shift 10.1.3 The principle of end-weight While the two principles discussed in the previous sections look at sentences with regard to their function in a text, the principle of end-weight concerns the sentence in isolation. It suggests that longer and more complex constituents are put at the end of the clause and has been formulated as early as 1909 by Otto Behaghel in his Law of Growing Elements: if possible, arrange syntactic elements in a way that shorter elements precede longer ones. Of course, English, being an analytic language (see below for details) has a rather fixed word oder, but in those cases where alternative arrangements are possible, the law can be seen at work. Take a look at the following example. (6) a) I sold the last copy of my really fascinating textbook Introduction to English Syntax that I published in 2010 to Peter. - b) I sold Peter the last copy of my really fascinating textbook Introduction to English syntax that I published in 2010. The two clauses above illustrate the phenomenon of dative alternation. The term ex‐ presses that we can choose between two constructions in the case that the construction has a dative (or indirect) object. The alternatives are as follows: (7) a) [I] S [sold] V [the last copy of my really fascinating textbook Introduction to English Syntax, which I had published in 2010] Od [to Peter] Oi . - - - Subject - Verb - direct Object - indirect Object - b) [I] S [sold] V [Peter] Oi [the last copy of my really fascinating textbook Introduction to English Syntax, which I had published in 2010] Od . - - - Subject - Verb - indirect Object - direct Object In cases like these, language users would tend to avoid alternative a) and prefer b), thus, following the principle of end-weight. The effect of end-weight can also work the other way, as you can see in (8). Version a) is what we actually find in the British National Corpus, b) is the modified version that does not follow end-weight. (8) a) [He] S [gave] V [the cash] Od [to the astonished owners of the house at Crosby, Merseyside] Oi . (BNC-CBF: 3757) - b) [He] S [gave] V [the astonished owners of the house at Crosby, Merseyside] Oi [the cash] Od . A last example of end-weight at work is shown in (9). (9) a) [I] S [looked] V [the price of Kreyer’s really fascinating textbook Introduction to English Syntax, which he had published in 2010] Od [up] Prep . - b) [I] S [looked up] V [the price of Kreyer’s really fascinating textbook Introduction to English Syntax, which he had published in 2010] Od . Example (9) illustrates a phenomenon called ‘Heavy-NP shift’, where heavy, i.e. long and complex, noun phrases are shifted to the end of the clause. Heavy-NP shift, as in the example above can occur with phrasal verbs but also in any case where another 196 10 Texts and the relevance of sentence structure <?page no="198"?> movable element occurs in the clause (some people would include dative alternation under this heading as well). So, we have seen a number of examples that illustrate the principle of end-weight, but what is the theory behind it? An excellent explanation can be found in Hawkins’ (1994: 57) Principle of Early Immediate Constituents words and constituents occur in the orders they do so that syntactic groupings and the immediate constituents (ICs) can be recognized (and produced) as rapidly and efficiently as possible in language performance. Different orderings of elements result in more or less rapid IC recognition. Let’s apply this to example (7). A competent speaker of the English language knows that the verb to sell needs a direct and an indirect object. In Hawkins’ terminology, the verb phrase consists of the immediate constituents (ICs) verb, direct object and indirect object. In case 7a) the reader has to read 21 words until they arrive at the preposition to, which tells them that they have reached the indirect object. In case 7b), the reader only has to read four words to be clear about the exact structure of the verb phrase and to know where all of the ICs begin. The earlier we are clear about the structures we are processing the better. In a less technical way, in case a) we are working our way through an extremely long and complex constituent without even knowing yet what to do with it. This puts a lot of strain on our working memory. It is basically the same argument that holds for the given-before-new principle: it takes effort for the human mind to keep something active without knowing what to do with it exactly. In case b), in contrast, the reader already knows after four words what structure they are dealing with and, accordingly, they know where the long noun phrase goes even before they have processed it completely. Before you read on … … again, please take a look at the text at the beginning of this chapter and identify those sentences that do not follow the principle of end-weight. a. That linguistics is a difficult subject for most first-year students becomes apparent fairly early in their first semester. b. There are three reasons for this. c. The fact that linguistics is usually not taught at school is the first reason. d. In this respect, literature or cultural studies have a more privileged status. e. Teachers at school teach these aspects and the syllabus lists them as important. f. The fact that students have to learn many new terms in linguistics is related to the first reason. g. Because students do not have any experience with linguistic terminology and ap‐ proaches they often feel a lot of pressure to get a grip on all the new concepts. h. The abstract nature of linguistics is another reason. i. Some students are reminded of maths by linguistics. 10.1 Principles of text processing 197 <?page no="199"?> analytic and synthetic languages j. Often they are scared off by this. k. To include aspects of linguistics into English language teaching at school would be a promising way of addressing these three problem areas. A definitive answer cannot always be given since the influence of end-weight is not a matter of black and white but rather of a gradient nature. Sentence a would definitely count as not in line with end-weight, since the subject of the whole sentence is long and, being a clause itself, syntactically complex. The same goes for sentences c and f, where the subject is a heavy noun phrase. Sentence k probably would also have to be counted: it is true, the difference between subject and subject complement is not that large (11 as opposed to 8 words), but in addition to being longer the subject also involves a to-infinitive clause, which makes it structurally more complex than the noun phrase that constitutes the subject complement. A sentence like h, in contrast, would maybe not be counted as violating end-weight, since the first constituent, though longer than the others, is still fairly short and structurally simple. 10.2 Helpful sentence structures Above I have already hinted at the fact that English is an analytic language. The term ‘analytic’ contrasts with ‘synthetic’ and describes the way in which a language ex‐ presses grammatical relations in a sentence. A good example of a synthetic language is Latin, but also German. Both languages have a fairly rich system of inflectional markers, which serves to express which syntactic function a word or a phrase has in the clause. We could say that the information with regard to the grammatical function is combined with the semantic information, both types of information are synthesised in one word form. If, for instance, we have a situation where a dog bites a man we have two options in German: der Hund beißt den Mann or den Mann beißt der Hund. Who bites and who gets bitten is expressed through morphological means in the respective noun phrases. Because of that it doesn’t matter where the noun phrases are position‐ ened in relation to one another. In English, an analytic language, the direct object does not have any case marking (apart from pronouns) that can help us to identify it as such. Whether the man bites or gets bitten depends on the position that the respective noun phrase has in the clause, i.e. the man bites the dog or the dog bites the man. The infor‐ mation about the grammatical function is not part of the word itself but it is expressed somewhere else, namely in the word order: the two aspects are apart from one another, they have been ‘analysed’, hence ‘analytic’. Because word order plays such an important role in English, it is fairly fixed. You may recall from your English lessons in school that the word order of English is SVO, subject - verb - object. This does not seem to leave us a lot of options if the subject is very long and complex and should be put at the end of the clause or if the object is the given element and should, therefore, be put in the beginning. Fortunately, the reality 198 10 Texts and the relevance of sentence structure <?page no="200"?> fronting is not quite as bleak since the English language provides a fair amount of so-called non-canonical and secondary clause patterns. 10.2.1 Non-canonical clause patterns Non-canonical clause patterns are called non-canonical since they are non-typical var‐ iants of the general clause patterns of the English language. For instance, a clause with a fronted direct object is still a clause of the type SVO, but it arranges these elements differently, namely OSV. We can distinguish two basic types of non-canonical clauses, ‘fronting’ and ‘full-verb inversion’. Fronting leaves the relative order of subject and verb untouched but moves a post-verbal constituent to the front of the clause. This is often done to achieve emphasis but can also be used to make sure that the three prin‐ ciples discussed above are followed. Any post-verbal constituent can be moved to the front, as shown in table 10.1. All of the examples are instances of authentic language use and are taken from the British component of the International Corpus of English (Aarts & Wallis 2006). Fronting of subject complement: SVCs → CsSV Oh that’d be great. Oh great that’d be. (s1a-042: 132) Fronting of indirect object: SVOiOd → OiSVOd I ask everybody that cooks how they make pastry. Everybody that cooks I ask how they make pastry. (s1a-057: 131 Fronting of direct object: SVOd → OdSV She did not consider her duty to her husband, fighting on foreign field for his and her freedom, at all. Her duty to her husband, fighting on foreign field for his and her freedom, she did not consider at all. (w2f-005: 027) Fronting of object complement: SVOdCo → CoSVOd I thought him distinctly brilliant. Distinctly brilliant I thought him. (s1a-094: 312) Fronting of obligatory adverbial: SVOdA → ASVOd You've got the axons of the ganglion cells on the innermost surface of the retina. On the innermost surface of the retina you've got the axons of the ganglion cells. (s1b-015: 114) Fronting of optional adverbial: SVOd(A) → (A)SVOd I’ve got a lot of rejections in my cupboard. In my cupboard I’ve got a lot of rejections. (s1a-034: 196) Table 10.1: Examples of fronting (cf. Kreyer 2010b: 123). 10.2 Helpful sentence structures 199 <?page no="201"?> full-verb inversion passive clause agent deletion Full-verb inversion is more restricted. It only occurs in clauses of the copular type. The name of this class of clauses derives from Latin copula, which among other things means ‘link’. The term makes sense, since in copular clauses a property or a location is linked to the subject. Kreyer is a professor links the property of being a professor to the subject, Kreyer. In the sentence He is in front of his computer, is links the location/ property in front of his computer to the subject of the clause. Since copular clauses only come in two forms, namely with a subject complement or obligatory adverbial, there are only two kinds of full-verb inversion. With both kinds, the subject swaps position with the post-verbal constituent. Full-verb inversion involving subject complement: SVCs → CsVS The continuous logging of the forests is equally important in terms of forest depletion. Equally important in terms of forest deple‐ tion is the continuous logging of the forests. (w1a-013: 071) Full-verb inversion involving obligatory adverbial: SVA → AVS Some delightful votive paintings are on the walls of the church. On the walls of the church are some delightful votive paintings. (s2b-027: 126) Table 10.2: Examples of full-verb inversion (cf. Kreyer 2010b: 123-24). 10.2.2 Secondary clause patterns In addition to these non-canonical clause patterns, the English language provides so-called secondary clause patterns. They are called secondary because they are gram‐ matical clause structures that the English language provides on top of the existing major clause patterns (that might be instantiated canonically or non-canonically). They involve a re-ordering of elements plus additional elements. Maybe the best-known secondary clause pattern is the passive clause. Passive clauses can be used on clauses that have objects. The direct or indirect object is put in initial position, the subject goes to final position as part of a by-phrase. This by-phrase can also be omitted, in which case we talk of ‘agent deletion’ (the subject usually being the doer of the action, i.e. the agent). The two kinds of passive transformation are shown in figure 10.2. Fig. 10.2: The two passive transformations. 200 10 Texts and the relevance of sentence structure <?page no="202"?> existen‐ tial-there sentence The passive is quite flexible and can fulfil various functions with regard to the three principles discussed above. Obviously, if the direct or indirect object contains a link to the previous discourse, the passive transformation can put it in initial position. If the subject is very heavy or if it deserves focus, it can be deferred to final position through the by-phrase. In addition to that, agent deletion can also be exploited to achieve various effects. Say, you do not want to talk about who is responsible for an action or you simply do not know: the passive with agent deletion will be helpful. The two examples below are taken from online press reports on the Uvalde school shooting in May 2022. (10) Nineteen children and two adults were killed in a shooting at Robb Elementary School in Uvalde County on May 24. (www.texastribune.org/ series/ uvalde-texas-school-shoo ting/ ) - Nineteen children and two teachers were killed in a massacre at an elementary school. (www.nytimes.com/ article/ uvalde-texas-school-shooting.html) As you can see, neither of the two headlines mentions the agent. We will talk more about agent deletion in section 13.1.2, where we’ll discuss manipulation through language. In our context, agent deletion can also serve as a means to put one of the objects in final position and, this way, give it end-focus. We have seen above that a sequence of given-before-new information is important when it comes to word order. What can we do if we are at the beginning of a text and there is no known information? Fortunately, the English language has a special con‐ struction for this case as well. We find it very frequently at the beginning of fairy tales, for instance in Once upon a time there was a king. The initial adverbial is so vague that it almost does not contain any real information. In addition, we see a so-called exis‐ tential-there sentence. These constructions only serve the purpose of establishing that something (in the case above, a king) exists. Since the existence of this something has to be established, this something has to be unknown to the reader. Following given-be‐ fore-new, the new information is put at the end of the clause. However, existential-there sentences are also particularly well-suited to introduce a new topic into the discourse, as is shown in the example below. (11) There are many more definitions of art besides Tolstoy's own and those which he denounced. For the purposes of this book, just two assertions will have to content us: first, some works have been intended by their makers to be seen as art; second, there is a consensus today that other works are to be described as art. (BNC-A04: 8-10) (11) is taken from a text on art criticism. It begins with a short discussion of Tolstoy’s definition of art and, then, wants to direct our attention to the fact that there are also other facets to art, of which two are then briefly discussed. The structure of existential-there sentences is shown in figure 10.3. 10.2 Helpful sentence structures 201 <?page no="203"?> it-extraposi‐ tion Fig. 10.3: The existential-there construction. The final secondary clause pattern we will be looking at is that of it-extraposition. In contrast to the other two, this pattern is primarily motivated by the principle of end-weight. It occurs in copular clauses where the subject of the clause is a clause itself and, hence, syntactically heavy and complex as in the following example. (12) It is possible that your local authority may ask you to audition for them after you have been offered a place. This seems an odd way of going about things, but the advising panels do carry a professional adviser so that talent is not merely being judged by local civil servants. (BNC-A06: 164-165) The name ‘it-extraposition’ derives from the fact that the subject of the clause is put outside of its usual position and that an it is added, as shown in figure 10.4. Fig. 10.4: It-extraposition. As you can see in example (12), it-extraposition will also satisfy end-focus needs but processability seems to be primary. Notice how clumsy and difficult to process the non-extraposed counterpart in figure 10.4 sounds. We have now introduced several ways to move elements to new positions in a clause. Let’s apply this knowledge to improve the text from the beginning of this chapter. 10.3 From text to good text - an example Before you read on … … please take (again) a look at the text from the beginning of this chapter (repeated below for convenience) and rewrite the text so that all of the sentences follow the principles described above. For each sentence that you rewrite make clear which principle is involved (sometimes it might be more than one) and name the 202 10 Texts and the relevance of sentence structure <?page no="204"?> non-canonical or secondary clause pattern that you used (if possible). Please note that the goal of this exercise is to apply what we have covered in this section in a more or less mechanical fashion in order to get a grip on the principles and sentence structures we have explored. There might be other ways to make the text sound better. But that is not our concern at the moment. a) That linguistics is a difficult subject for most first-year students becomes apparent fairly early in their first semester. b) There are three reasons for this. c) The fact that linguistics is usually not taught at school is the first reason. d) In this respect, literature or cultural studies have a more privileged status. e) Teachers at school teach these aspects and the syllabus lists them as important. f) The fact that students have to learn many new terms in linguistics is related to the first reason. g) Because students do not have any experience with linguistic terminology and ap‐ proaches they often feel a lot of pressure to get a grip on all the new concepts. h) The abstract nature of linguistics is another reason. i) Some students are reminded of maths by linguistics. j) Often they are scared off by this. k) To include aspects of linguistics into English language teaching at school would be a promising way of addressing these three problem areas. The suggestions I make below try as much as possible to do justice to all the principles that we have discussed above. As said before, this is maybe not the best version of the text in general, but the idea is to illustrate the use of sentence structures to move elements around so that a text becomes more readable. a) That linguistics is a difficult subject for most first-year students becomes apparent fairly early in their first semester. a’) It becomes apparent fairly early in their first semester that linguistics is a difficult subject for most first-year students. The revised version makes use of it-extraposition. This way, the long and complex subject clause is put to final position, where it is in line with the principle of end-weight. In addition, this ordering follows end-focus. The extra-posed subject is what the whole text is about and it makes sense to put it in a prominent position at the end of the clause to introduce it as a topic. b) There are three reasons for this. b’) This is due to three reasons. The rewriting above does not follow a particular non-canonical or secondary clause pattern. Instead it rewrites an existential-there sentence as a canonical SVC clause and slightly alters the wording. It follows the principle of given-before-new by preposing the demonstrative pronoun this, which links back to the previous sentence, and it is in 10.3 From text to good text - an example 203 <?page no="205"?> line with the principle of end-focus, since the three reasons are the topic for the rest of the text. c) The fact that linguistics is usually not taught at school is the first reason. c’) The first reason is the fact that linguistics is usually not taught at school. The new version, again, makes use of full-verb inversion to make the sentence more in line with the principle of end-weight. At the same time it follows given-before-new, since the previous clause introduced three reasons and the first reason can thus serve as a link to the previous discourse. As in the example of it-extraposition above, the new word order is also conducive to the principle of end-focus, because the extraposed subject will remain the theme for the following two sentences. d) In this respect, literature or cultural studies have a more privileged status. d’) In this respect, a more privileged status is attributed to literature or cultural studies. The original sentence is already in line with the given-before-new principle due to the adverbial in this respect. The use of the passive construction (plus a necessary new verb) might be considered a little too much, but for the sake of the exercise we could argue that it marks literature or cultural studies as the focus, which makes sense as the phrase is picked up again in the following sentence. e) Teachers at school teach these aspects and the syllabus lists them as important. e’) These aspects are taught by teachers at school and the syllabus lists them as important. The passive clause used above puts These aspects in initial position where it can serve as a link to the previously mentioned literature or cultural studies, i.e. the re-writing is in line with the given-before-new principle. f) The fact that students have to learn many new terms in linguistics is related to the first reason. f ’) Related to the first reason is the fact that students have to learn many new terms in linguistics. Here, we make use of a full-verb inversion. This change serves several purposes. Firstly, the fronted subject complement can now serve as a link to the idea of the three reasons that have been introduced in the second sentence of this text. It thereby guarantees that the principle of given-before-new is heeded. Secondly, the many new terms in linguistics are elaborated upon in the next sentence. So, the re-written version is more in line with end-focus. Finally, the fact that students have to learn many new terms in linguistics is a heavy noun phrase, which according to the principle of end-weight, should occur at the end of the clause. g) Because students do not have any experience with linguistic terminology and approaches they often feel a lot of pressure to get a grip on all the new concepts. g’) Students often feel a lot of pressure to get a grip on all the new concepts, because they do not have any experience with linguistic terminology and approaches. 204 10 Texts and the relevance of sentence structure <?page no="206"?> g’’) To get a grip on all the new concepts often puts students under a lot of pressure, because they do not have any experience with linguistic terminology and approaches. The adverbial Because students … and approaches in the sentence above is optional and can be moved fairly freely. If we move it to the end of the superordinate clause the noun phrase all the new concepts moves closer to many new terms in linguistics in the previous sentence, to which it links up. The new version, thus, is more in line with given-before-new. This effect is enhanced in the second alternative version, where the relevant noun phrase is moved even further to the beginning of the clause. h) The abstract nature of linguistics is another reason. h’) Another reason is the abstract nature of linguistics. The above sentence serves a pivotal function as, by introducing another reason, it shifts the topic from the many new terms of linguistics to the abstract nature of linguistics. Through the full-verb inversion the subject complement is fronted and can serve as a link to the previous discourse, particularly to sentence b, where three reasons are introduced. At the same time, this ordering serves the end-focus principle as the abstract nature of linguistics, which is the topic for the next two sentences, is put into the prominent clause-final position. i) Some students are reminded of maths by linguistics. i’) Linguistics reminds some students of maths. By choosing the active instead of the passive version of the clause, the subject, Linguistics, occurs at the beginning of the sentence, where it can provide a useful link to the abstract nature of linguistics from the previous clause. In addition, this similarity to maths, now, is in final position and closer to the next clause, which discusses this aspect a little more. j) Often they are scared off by this. j’) Often, this scares them off. Similar to the previous sentence, this re-written active version puts the linking subject into initial position. k) To include aspects of linguistics into English language teaching at school would be a promising way of addressing these three problem areas. k’) A promising way of addressing these three problem areas would be to include aspects of linguistics into English language teaching at school. k’’) All of the three problem areas could be addressed if we included aspects of linguistics into English language teaching at school. Finally, the last sentence can be rewritten by making use of full-verb inversion, which leads to two effects. The noun phrase these three problem areas is what the previous sentences were about. The re-writing puts this noun phrase closer to the beginning of the sentence, and, this way, is more in line with the given-before-new principle, which 10.3 From text to good text - an example 205 <?page no="207"?> makes it easier for the reader to know what to do with the information provided in the whole sentence. In addition, and more importantly maybe, the new version gives end-focus to a possible solution to these three problems. The second alternative is even more in line with given-before-new, but it could be argued that it does not do as well as a concluding sentence since the main point, the solution to the whole problem, now occurs in a subordinated if-clause. Example 13 juxtaposes the original version and the new version. As far as the processing principles are concerned the text has improved considerably. (13) That linguistics is a difficult subject for most first-year students becomes apparent fairly early in their first semester. There are three reasons for this. The fact that linguistics is usually not taught at school is the first reason. In this respect, literature or cultural studies have a more privileged status. Teachers at school teach these aspects and the syllabus lists them as important. The fact that students have to learn many new terms in linguistics is related to the first reason. Because students do not have any experience with linguistic terminology and approaches they often feel a lot of pressure to get a grip on all the new concepts. The abstract nature of linguistics is another reason. Some students are reminded of maths by linguistics. Often they are scared off by this. To include aspects of linguistics into English language teaching at school would be a promising way of addressing these three problem areas. (13’) It becomes apparent fairly early in their first semester that linguistics is a difficult subject for most first-year students. This is due to three reasons. The first reason is the fact that linguistics is usually not taught at school. In this respect, a more privileged status is attributed to literature or cultural studies. These aspects are taught by teachers at school and the syllabus lists them as important. Related to the first reason is the fact that students have to learn many new terms in linguistics. To get a grip on all the new concepts often puts students under a lot of pressure, because they do not have any experience with linguistic terminology and approaches. Another reason is the abstract nature of linguistics. Linguistics reminds some students of maths. Often, this scares them off. All of the three problem areas could be addressed if we included aspects of linguistics into English language teaching at school. New tools in your toolkit Given-before-new principle ▶ You have learned that sentences are generally more easily integrated into a text if they start with given/ known/ old information and proceed to unknown/ new information. Texts that follow this principle will usually be easier to read. Principle of end-focus ▶ You have learned that readers would usually expect the most important information of the sentence to occur at its end. You can exploit this to make sure you get your message across and you can also use it to make the text structure more accessible to your reader. 206 10 Texts and the relevance of sentence structure <?page no="208"?> Principle of end-weight ▶ You have learned that long and complex clause constituents would usually occur towards the end of the clause, making it easier for your reader to process the clause. Sentence structures ▶ Your awareness of non-canonical and secondary clause patterns provides you with options to move constituents of a clause to initial or final position so that your texts follow the three principles above as much as possible and, as a consequence, can be read more easily. If you want to know more C A L L I E S , Marcus (2009): Information Highlighting in Advanced Learner English. The syntax-prag‐ matics interface in second language acquisition. Amsterdam: John Benjamins. ▶ Provides interesting insights into the use of non-canonical word order by advanced German learners of English as opposed to native speakers and discusses pedagogical implications. K R E Y E R , Rolf (2010b): Introduction to English Syntax. Frankfurt a.M.: Peter Lang. ▶ Useful if you want to read up on clause and sentence structure in general. Parts of this chapter are based on this textbook, which provides a more detailed discussion of syntactic variation and text structure. W A R D , Gregory and Betty B I R N E R (2006): “Information structure and non-canonical syntax". In: Laurence R. Horn & Gregory Ward (eds.): The Handbook of Pragmatics. Malden, MA: Blackwell. 152-174. ▶ This paper provides a comprehensive overview of the different non-canonical and secon‐ dary clause patterns that serve to create a general given-before-new information structure in texts. Additional online content. 10.3 From text to good text - an example 207 <?page no="210"?> 11 Texts and the relevance of the situation I MU M8. LTNS. RUOK? ( A T E X T M E S S A G E ) What this chapter is about It is good to know about text structure, about principles of processing and about sentence structures that help us to create texts that are (more) easily understood. However, texts do not occur in a vacuum. They are always part of a communicative situation. A text may have an excellent structure that can be processed with almost no effort, however, if it is not suitable for the situation it occurs in, the text still is no good. This chapter will look at two particularly important aspects of the communicative situation, the question whether we are dealing with spoken and written language and consequences that derive from the relationship between producer and recipient. The last two chapters have explored what it is that makes a text a (good) text. We have taken a look at standards of textuality and at ways of describing text structure in chapter 9. Chapter 10 has focused on questions of text processing and it has explored sentence-level resources that can contribute to texts that are easy to process. One important aspect has been ignored so far, though. Texts do not exist in isolation but they are embedded in a given communicative situation. This situation would usually have an impact on the text itself, as DeBeaugrande and Dressler (1981: 9) make clear in their discussion of the situationality of texts. Fig. 11.1: A traffic sign. <?page no="211"?> The text on the left of figure 11.1 is shaped significantly by the situation of use. This becomes particularly apparent when we compare it to a more explicit version on the right (DeBegraunde & Dressler 1981: 10): (1) Motorists should proceed slowly, because children are playing in the vicinity and might run out into the street. Vehicles can stop more readily if they are moving slowly. Even though the alternative under (1) is a lot more specific than the actual traffic sign in figure 11.1, it would not be appropriate for the given situation since the text is too long and drivers would usually not be able to read all of it. Even if they were able to do that, it would divert their attention from the road in front of them, which would make the sign especially problematic. Another case in point is early text messaging. Due to the fact that texts were not for free and limited to 140 characters, as well as the fact that screens on early mobile phones were very small and typing messages was usually quite laborious (no swyping or auto-complete), language use was adapted. For instance, so-called ‘SMS language’ or ‘textese’ would show abbreviations like those shown at the beginning of this chapter and in (2) below. (2) thx thanks - m8 mate - lol laughing out loud - ltns long time no see - b4 before - mu miss you - ruok are you okay Obviously, abbreviations of this kind are not restricted to text messaging but show in all texts that result from situations where time or space is of the essence: afk (away from keyboard), brb (be right back) and otw (on the way) are just a few examples from online-gaming communication. The language that we use also changes with regard to the person we are communi‐ cating with. Take a look at the following examples. (3) Hi Y, count me in! X (4) Student: I need your signature. [throws document on my desk] Example (3) is the full text from an email that a colleague of mine received from a student who wanted to register for one of his classes. Example (4) shows how a student of mine once asked me to sign a document for him. Both instances of language use would (maybe) be appropriate among friends but not in an interaction of students with their professors. The two alternatives shown below seem to be more suitable. (3’) Dear Prof Y, - I would like to take part in your seminar on XYZ. I’d be grateful if you could let me know whether I could participate. - Thank you very much and best wishes, - X 210 11 Texts and the relevance of the situation <?page no="212"?> register (4’) Could you please sign this for me? [hands me the document] Variation in language use of the kind exemplified above, i.e. variation with regard to the situational context, is referred to as ‘register’ (e.g. Biber & Conrad, 2019: 6). To some extent questions of register do play a role in language teaching at school already. Questions of degree of formality, for instance, were relevant when your teacher asked you to write two texts about the same incident, one being a formal complaint the other being a letter or an email to a friend. Other aspects of register, however, do not seem to figure prominently in school. The examples above illustrate two of the most relevant dimensions when it comes to fitting texts to the situation they are used in. The first has to do with the mode and the medium of communication. A fundamental distinction in this respect is that between spoken and written language, which will be discussed in the following section. The second dimension relates to the relation between participants, e.g. one friend to another friend or an employee to their employer. This aspect will be explored with a focus on politeness in section 11.2. 11.1 Spoken and written English Before you read on… … try and recall whether you have ever had the same experience? You are listening to a presentation by one of your fellow students. They explain things well and it is relatively easy to follow what they have to say. Suddenly, things change: they still speak in English and you understand the words they use but somehow you find it really difficult to follow their train of thought. Then, just as suddenly, things change again and it is easy for you to understand everything they say. What’s going on? I have had the above experience many many times as a student and also as a professor. It took me a while to understand what was going on. The problem that we encounter in situations like these is that the presenters use written language but communicate it to us through the auditory channel. Written and spoken English, of course, are the same language but they are sufficiently different to cause problems for us, particularly when we use written structures where we should use spoken ones. What does that mean? Before you read on … … take a look at the following two examples. Which would you say is written and which of them is spoken? Or, which of the two do you think you could understand more easily if read out aloud to you? Can you give reasons for your decision? Which differences between these two examples can you identify? 11.1 Spoken and written English 211 <?page no="213"?> process and product a) If you restrict the number of calories drastically, you will lose weight very quickly. But usually this will not be permanent. b) A drastic restriction of the number of calories will result in very quick but usually non-permanent weight loss. Example a) represents spoken language, example b) is written language, at least this is what I had in mind when I came up with the examples. When it comes to differences between the two, note that both examples convey exactly the same information. The only difference, and a very important one at that, is how they present that information. Halliday (1989) emphasises the important distinction between process and product when it comes to spoken and written language. There are two interpretations of the dichotomy. The one that comes to mind first is the obvious one: the speaking of a language in and of itself is a process and the recipient witnesses this process. Writing a text, of course, is a process as well but the addressees only see the finished product. For example, when you read these lines you have no idea of how often I rephrased and reformulated them. You have no idea how often I started and restarted the sentences and how often I deleted passages. It is important to note, though, that it is absolutely normal to rephrase and reformulate during speaking. Another feature that we can witness are filled and un-filled pauses. Former US President Barack Obama is an ex‐ cellent example. Here is the answer he gave to a question at a town hall as to why he wants to control, restrict and limit gun use and manufacturing. Filled pauses are rep‐ resented by er, uhm, and erm, silent pauses by […]. (5) Alright, well, let me er uhm erm - it’s a multi-part question. So, so let me just erm say a couple of things. First of all […] er the notion that […] I […] or Hillary of Democrats or whoever you want to choose are […] hell bent on taking away folks’ guns […] is just not true. And and I I I don’t care how many times the NRA says it. (https: / / www.youtu be.com/ watch? v=6imFvSua3Kg; 0: 48 - 1: 19) As you can see, when speaking spontaneously, even the former President of the United States, who is an excellent speaker and rhetorician shows many filled and unfilled pauses and repetitions in his language. And this is no problem at all. On the contrary, pauses, repetition and other signs of hesitation and planning are actually useful for the speaker as well as the recipients. It gives the speaker time to plan their utterances while at the same time it makes it easier for the recipients to process what has just been said. However, with regard to the task above the second understanding of the contrast between process and product is more important. Halliday (1989: 81) argues that “written language represents phenomena as products. Spoken language represents phenomena as processes.” What does that mean? 212 11 Texts and the relevance of the situation <?page no="214"?> Clauses and their elements In section 9.2.2 we have looked at the distinction of clauses and sentences and we have reviewed the distinction of obligatory and optional elements. It might be useful now to have a look at the different obligatory clause elements that the verb can demand. Take a look at the following list. Verb type Clause type Obligatory clause elements Example intransitive SV subject [The dog] S barks copular SVC subject, subject comple‐ ment [The dog] S is [fierce] Cs - SVA subject, obligatory ad‐ verbial [The dog] S is [in the shed] As monotransi‐ tive SVO subject, direct object [The dog] S bit [the man] Od ditransitive SVOO subject, direct object, indirect object [The dog] S gave [the man] Oi [a start] Od complex transitive SVOCo subject, direct object, object complement [The man] S considers [the dog] Od [a threat] Co - SVOAo subject, direct object, object complement [The man] S put [the dog] Od [in the shed] Ao Table 11.1: verbs, their clause types and their obligatory clause elements. Note that in addition to the obligatory clause elements you can also have additional optional adverbials, e.g. [The dog] S [was barking] V [loudly] (A) [in the shed] (A) [yesterday] (A) . Halliday claims that spoken language presents events as processes. A process is something that is described by a verb. A verb, together with its obligatory and optional clause elements, creates a clause. Let’s take a look at our example a) from the task above, here repeated as (6) for convenience. (6) If you restrict the number of calories drastically, you will lose weight very quickly. But usually this will not be permanent. 11.1 Spoken and written English 213 <?page no="215"?> We have two sentences, the first being a complex sentence with an initial subordinate clause and the main clause, the second being a simple sentence, consisting of one clause only. All in all, we have three clauses with three main verbs, namely to restrict, to lose and to be. The first of these verbs is a monotransitive verb, i.e. it demands two constituents, a subject (you) and a direct object (the number of calories). In addition, the first clause has an optional adverbial, drastically. Before you read on … … provide a similar analysis for the remaining two clauses. In addition, what is the function (in terms of clause constituents) of the first clause in the first sentence? The second clause, you will lose weight very quickly, has the same functional constitu‐ ents as the first. The verb to lose needs a subject (you) and a direct object (weight). Again, the adverbial (very quickly) is optional. The third clause, But usually this will not be permanent, is an example of a copular clause. The verb to be links (lat. copulare means ‘to link’ or ‘to join’) the subject (this) to an attribute (permanent). This clause, too, has an optional adverbial, namely usually. As regards the function of the first clause in the first sentence, we see that the subordinate clause as a whole functions as an optional adverbial to the main clause of the first sentence: when will you lose weight very quickly? If you restrict the number of calories drastically. These processes (that according to Halliday are typical of speaking) are presented as products in writing. A product is a thing. Things are described by nouns and noun phrases. By the way: The noun phrase The noun phrase is a so-called endocentric phrase, which literally translates into ‘a phrase that has its centre within’. This means that a noun phrase can be reduced to one of its elements. If you had to pick the most important word of the noun phrase the ferocious dog next door, for instance, this would be dog. The technical term for the most important word of a noun phrase (and for the most important word in any endocentric phrase, for that matter) is called ‘the head’. Once we have identified the head, it is very easy to describe the structure of the noun phrase: (determiner) + (premodification) + head + (postmodification). As you can see, the head is the only obligatory element in the noun phrase. Everything after the head is postmodification. Before the head we find determin‐ ers and premodification. The function of the determiner is to make clear (to determine) which part of the world the noun phrase refers to, e.g. this cold glass of water as opposed to that cold glass of water. Preand postmodification serve 214 11 Texts and the relevance of the situation <?page no="216"?> to describe the head in more detail, e.g. this cold glass of coke as opposed to this hot glass of water. Figure 11.2 provides a comparison of the spoken and the written version. Fig. 11.2: A comparison of spoken sentences in (6) and their written counterpart in (7). We have already seen that both versions convey exactly the same information. However, the three clauses in (6) have become one clause in (7). The main verb of this clause is to result in. It expresses the conditional link of the first and second clause in (6): if you do A, this will result in B. Note how two of the main verbs in (6) have become heads of noun phrases in (7): restrict is now restriction, lose is now loss. With the exception of the subjects, all the other bits of information in (6) become preand postmodifying items of these two head nouns in (7). The optional adverbial drastically becomes the premodifying adjective drastic and the direct object the number of calories becomes the postmodifying prepositional phrase of the number of calories. The direct object weight and the optional adverbial very quickly in the spoken version are premodifying items of the head noun loss in the written version. The pronoun this in the spoken version refers back to the losing of weight described in the previous sentence. Accordingly, the subject complement and the optional adverbial in the third clause of (6), usually and (not) permanent, also become premodifying material of the head noun loss in (7). All in all, we can see how the processes represented by verbs and clauses in (6) become products represented in noun phrases in (7). Before you read on … … take a look at the following sentence. It is supposed to be representative of spoken English. Please, reformulate this sentence, trying to pack as much of the information as possible in noun phrases. After the teacher forcefully denied all accusations that he has treated students unfairly, students protested all over the university. The example above consists of two clauses. The first is a temporal subordinate clause and functions as an optional adverbial to the second clause, which is the main clause (note that the clause he has treated students unfairly is what we call a that-complement 11.1 Spoken and written English 215 <?page no="217"?> integration vs. fragmen‐ tation clause, which serves to postmodify accusations). We can assume that the teacher’s denial is the reason for the student protests. This relation should become the main verb of our written version. It could be expressed through caused, has resulted in, has led to or something similar. The written sentence, thus, would be of the form X has led to Y. X, in this case, would be a noun phrase that contains the information in the first clause After the teacher … unfairly, and Y would be a noun phrase representing the second clause students protested … university. As should have become clear from above, main verbs in a clause become the head nouns of noun phrases, when we transform a spoken sentence into a written one. The two main verbs, denied and protested, become the nouns denial and protests in our written version. All the other clause constituents in the spoken version become pre-and postmodifying items of the noun phrases. Let’s start with the second clause, since it is less complex. The subject of the clause, students, can become a noun premodifier of the head protests, resulting in student protests. The content of the optional adverbial all over the university can be captured by another premodification, leading to university-wide student protests. The first clause is a little more complex. The subject, the teacher, becomes the determiner of the noun phrase and the optional adverbial forcefully becomes a premodifying adjective, forceful, resulting in the noun phrase the teacher’s forceful denial. We could simply add the direct object of the clause as part of a prepositional phrase and stop here: the teacher’s forceful denial of all accusations that he has treated students unfairly. However, for the sake of the exercise, let us make the noun phrase as packed with nouns as possible. In particular, this means to turn the that-complement clause he has treated students unfairly into a noun phrase. Again, we need to identify the main verb and make it the head noun of a noun phrase i.e. treated becomes treatment. The remaining clause constituents can be expressed through preand postmodifying items, which results in unfair treatment of students. An alternative would be to only use premodification: unfair student treatment. Below you can compare the original spoken version with the written counterparts. (8) After the teacher had forcefully denied all accusations that he has treated students unfairly, students protested all over the university. (8a) The teacher’s forceful denial of accusations that he has treated students unfairly has led to university-wide student protests. (8b) The teacher’s forceful denial of accusations of unfair treatment of students has led to university-wide student protests. (8c) The teacher’s forceful denial of accusations of unfair student treatment has led to university-wide student protests. What is the relevance of this for you as a future teacher? The relevance lies in the fact that speaking and writing are two fundamentally different modes of language use. When speaking happens in real time, time constraints become very important, whereas they are more or less irrelevant when writing. Wallace Chafe captures this difference in the dichotomy of integration versus fragmentation. 216 11 Texts and the relevance of the situation <?page no="218"?> As we write down one idea, our thoughts have plenty of time to move ahead to others. […] In writing we have time to mould a succession of ideas into a more complex, coherent, integrated whole, making use of devices we seldom use in speaking. (Chafe 1982: 37) Noun phrases of the type that we have seen in examples (7) and (8) above illustrate such a moulding of a series of ideas into one “complex, coherent, integrated whole”. In contrast, the time constraints of speaking do not allow for such a moulding of ideas. Instead, [w]e normally produce one idea unit at a time. That is apparently all we have the capacity to pay attention to, and if we try to think about much more than that we are likely to get into trouble. (Chafe 1982: 37) So, spoken language is more fragemented whereas written language is rather integra‐ ted. Carter and McCarthy (2006: 170) describe the difference in a similar fashion. Spoken grammar, in their view, is “more like the strung-together coaches of a train or links of a chain”, while in writing we find “a carefully constructed hierarchy of embedded structure, one inside the other.” Even though the structural idiosyncrasies of spoken and written language described above can best be explained as a result of different circumstances of production, they are also highly relevant for reception and processing. If our ideas are presented one idea unit at a time, if they are presented like strung-together coaches of a train, they are much easier to process for our addressees than if we presented them in large embedded structures. This is also what you can find in the example from Barack Obama in (5), here repeated as (9): (9) Alright, well, let me er uhm erm - it’s a multi-part question. So, so let me just erm say a couple of things. First of all […] er the notion that […] I […] or Hillary or Democrats or whoever you want to choose are […] hell bent on taking away folks’ guns […] is just not true. And and I I I don’t care how many times the NRA says it. While Obama plans and produces one idea after another, the audience has enough time to process each of the individual ideas. Notice the difference if we rendered (9) into a more written version (similar to what we did in the last exercise). (9a) My first answer to your multi-part question is to stress the falsehood of the frequently repeated NRA claim that I, or Hillary, or Democrats or whoever you want to choose have made the decrease of the number of privately owned firearms a priority of our politics. While this might sound somewhat more learned or academic, it is not suitable for spoken language. The reason for this lies in the fact that it is costly for humans to keep structures open in their mind for too long. In section 10.1.3 we discussed Hawkins’ (1994) Principle of Early Immediate Constituents, which stresses the importance of rapid processing of syntactic structures. Obviously, long, complex and intricately embedded noun phrases take a lot of time to process. And, while we are trying to do 11.1 Spoken and written English 217 <?page no="219"?> that, we have to keep a lot of information active in our minds. Take, for instance, the noun phrase in (9a) which starts in the falsehood of the …. This noun phrase is 36 words long, all of which have to be kept active until, finally, we have reached the end of the noun phrase … of our politics. This can be done while reading, where our eyes can move back and forth freely, whereas we cannot do it when someone talks to us. Now, let’s get back to the task from the beginning of this section and the experience described there: You are listening to a presentation by one of your fellow students. They explain things well and it is relatively easy to follow what they have to say. Suddenly, you cannot follow them any more, even though they still speak English and you know the words. Then, just as suddenly, you can understand everything they say easily again. My explanation for this phenomenon is that in the first part of the presentation the presenters speak about something that they have fully understood. As a consequence, they feel confident enough to explain it in their own words and to create their sentences spontaneously. This means they will use spoken language. Then there is a part in the paper they want to present that they have not fully understood but they do have a feeling that this part is essential, so they do not want to leave it out. But what they do is to more or less copy this part into their presentation. Of course, the copied part is written language and it comes with the structures typical of written language, particularly written academic language, namely long and complex noun phrases. The rest of the paper, again, is easier for them to understand and they feel confident to use their own words and to create sentences and structures as they speak, resulting in spoken language with its typical progression from one idea unit to the next. Spoken and written language are two fundamentally different modes of language use. More specifically, the intricate hierarchically structured NPs of written language are not suitable for spoken language. In addition, authentic spoken language shows a large variety of different markers of disfluency, like false starts, filled and unfilled pauses and repetitions. We have seen in the Obama example above that disfluency is completely natural in spoken discourse. Here is another example from the spoken component of the British National Corpus. (10) Jean: Do you do you think they get a false idea from the television or something, wha-what is it? - Brenda: Well I dunno I thi-- I think the teachers to be honest, Jean, no you - Jean: Ah! - Brenda: know <pause> th-the teachers in this <pause> I mean <pause> you know I don't know, I mean <pause> so-some of the teachers have got <pause> got really weird ideas! (BNC-KBF: 1052-55) In speaking in general and in conversation in particular disfluency is the norm and it is perfectly acceptable. Götz (2013), for example, in a study on fluency in native and non-native speakers finds that “unfilled pauses are a natural result of speech production” (20). Her comparison also reveals that discourse markers and small words such as you know, like, or sort of as well as repeats are significantly underused by 218 11 Texts and the relevance of the situation <?page no="220"?> face learners of English (127). The latter features are particularly prominent in Brenda’s answer in (10) above. Obama does it, native speakers of English do it, you know that you do it when you speak in your mother tongue. Why should it be wrong, to do it when you speak English, and, more importantly, why should it be wrong if your future students do it? 11.2 The interpersonal dimension - politeness Before you read on … … imagine you want to write something down quickly but you have forgotten your pen. You want to ask the person next to you to lend you their pen. Now imagine that this person either is a close personal friend of yours or your professor. Will this have an impact on how you ask for the pen? If so, what could the two versions look like? I assume that your request will differ depending on who you are asking. It would make a difference to me. If I asked a friend, a request like “Give me your pen, please! ” or “Can I borrow your pen? ” might be suitable. If the person I asked for a favour was not well-known to me or if I considered them my superior, I would phrase the request more carefully, i.e. “Would you mind lending me your pen for just a minute? ” or “May I borrow your pen, please? ” seem to be more appropriate than the more direct versions above. These examples show that the relationship between interactants is a very important situational dimension. For example, you would usually greet close friends of yours differently than you greet your teachers or your dentist. Above we have seen that asking for a favour will most likely differ if you are asking a favour of a friend of yours or of someone you consider superior. The same holds true for denying a favour. Obviously, this list of possible scenarios could be extended further. At the core of these and countless other examples is the central concept of politeness. When I recall my own school days, politeness seemed to be somehow important but the concept itself remained rather elusive. What I was being told was that the English are more polite than the Germans and that politeness was very important. All of this is true, however, I wasn’t told what it meant to be polite in English, other than a few phrases like “Would you mind doing X? ” or “May I Y? ”. Linguistics helps us to get a clearer understanding of the nature of politeness and it describes how politeness shows in language use. This is what this section is about. Maybe the most influential contribution to our understanding of politeness was made by Brown and Levinson (1987). They interpreted politeness with regard to the concept of face. Their use of ‘face’ is reminiscent of the use in the idioms to lose face or to save face, meaning ‘to become less respected’ or ‘to avoid becoming less respected’. Accordingly, they define face as “the public self-image that every member wants to 11.2 The interpersonal dimension - politeness 219 <?page no="221"?> claim for himself ” (1987: 61). In other words, we can say that face represents the way we want others to regard and treat us. Two aspects of face can be distinguished: negative face: the basic claim to territories, personal preserves, rights to non-distraction - i.e. to freedom of action and freedom from imposition […] positive face: the positive consistent self-image or ‘personality’ (crucially including the desire that this self-image be appreciated and approved of) claimed by interactants. (61) By the way: The negative in negative face From my experience, students sometimes have problems with the use of the term ‘negative face’ (and the associated concept ‘negative politeness’, which we will discuss a little further down below). It might help to focus on the second part of the definition of the term, i.e. negative face is the basic claim to freedom from imposition. This way, negative face is defined as the absence of something, as the negation of something, hence the term. The term ‘positive face’, in contrast, is usually easy to make sense of, since it reminds us of saying something positive about a person (which is part of it). However, it is important to note that ‘negative face’ is not about how we say negative things about or to a person. This, too, belongs to positive face: when I say something negative about/ to you, I challenge the positive self-image that you claim for yourself. Negative face is solely about a person’s right to do whatever they want. By the way, the linguistic concept of negative face is validated by psychological research into a phenomenon called ‘reactance’. Reactance Theory claims that “[in] general, people are convinced that they possess certain freedoms to engage in so-called free behaviors. Yet there are times when they cannot, or at least feel that they cannot, do so […], and this is where reactance comes into play. Reactance is an unpleasant motivational arousal that emerges when people experience a threat to or loss of their free behaviors” (Steindl et al. 2015: 205). Brown and Levinson’s idea of face becomes relevant for politeness when we take into consideration how communication relates to and interacts with our face needs. Before you read on … … imagine someone said the following things to you. a) You look great! b) Would you mind closing the window? c) Stop talking! d) Have another biscuit! e) You look a little tired. f) I disagree! 220 11 Texts and the relevance of the situation <?page no="222"?> face-threat‐ ening act g) Have you ever thought about giving up smoking? h) It’s cold in here. Which of these would have an impact on your negative face and which would have an impact on your positive face? Is it possible to impact both with one utterance? Example a) is a typical example of an utterance that has consequences for your positive consistent self-image, i.e. your positive face. A comment of this kind would usually support and strengthen this positive self-image. However, it is important to note that ‘related to positive face’ does not mean the utterance has to be positive. Telling someone that they look tired as in example e) is usually not positive but it still concerns your positive face since it challenges your positive consistent self-image. The sentence I disagree! (f) is similar in that respect because the speaker’s disagreement signals that they do not approve of the way the hearer sees something. Example b) is an obvious case of an utterance that imposes on my freedom of action. You might be sitting in a seminar room, happily doing nothing and enjoying it. If I made that request it has consequences for your negative face, because in a way you cannot do exactly what you want to do anymore (assuming that you didn’t want to close the window anyway). Example g) is similar in that it suggests a change in your behaviour, i.e. it challenges your freedom of action. The same holds true for c). It challenges your freedom of action because, obviously, talking (to the person next to you in my seminar) is exactly what you want to do at this moment and I demand of you that you stop doing just that. So, I deny you your freedom of action - a clear infringement on your negative face. In contrast to b), however, g) and c) at the same time threaten your positive face. Example g) might indicate some disapproval with your habit, which is a threat to your positive face. Example c) threatens your positive face, since I act as someone that has the authority or power to tell you what to do, a behaviour that is not particularly appreciative of your self-image. d) is interesting since it is similar to c) in that it kind of tells you what to do, thereby restricting your freedom of action. At the same time, however, when saying that, it seems as if I want you to feel comfortable and well. This, in contrast to c), would have a positive effect on your positive face. Finally, example h) is special in that on the surface it makes a comment about the temperature in the room. However, if you were sitting in my seminar next to an open window, you might take it as a hint to close the window. The effect of h) would be similar to that of b), it is an imposition on your freedom of action, but it leaves you more of a way out, which is important when it comes to politeness. As these examples show, there are many ways in which utterances can have an effect on your face. If this effect is negative, Brown and Levinson speak of ‘face-threatening acts’ (FTAs), i.e. any act that threatens the adressee’s positive or negative face. Note that this act can involve language but it doesn’t have to. If you asked me after the seminar “Do you have a minute? ” I would threaten your positive face by saying “No, I don’t have time for you! ” but it would be just as face-threatening if I just turned my back on you. 11.2 The interpersonal dimension - politeness 221 <?page no="223"?> positive and negative politeness off record on record Having clarified the basic notions of face and face-threatening acts, we are now in the position to explain what politeness is about. Put simply, politeness is face work. Brown and Levinson focus primarily on politeness as a reduction or mitigation of face-threatening acts. ‘Positive politeness’, in this context, refers to attempts on the part of the speaker to mitigate threats to positive face, ‘negative politeness’, conversely, refers to attempts to mitigate threats to negative face. For instance, saying You look a little tired! instead of saying Wow, you look like s***! is a form of positive politeness. In both cases, I say something negative about the addressee, i.e. I will threaten their pos‐ itive face, however, phrasing it as a little tired will be a lot less threatening to their self-image than phrasing it as like s***. By the way: Positive politeness is not only about saying positive things The concept of positive politeness can easily be misunderstood. Students often think that positive politeness is just about saying nice things to people, e.g. paying someone a compliment or encouraging someone. This is, strictly speaking, not true! At the heart of Brown and Levinson’s idea of politeness is the face threatening act. Positive politeness is about mitigating threats to the hearer’s positive face. For example, if you ask me “Have I passed the exam? ” and I have to tell you that you haven’t, my answer is a threat to your positive face. If I answer “I’m sorry, I am afraid you haven’t” instead of “No, you failed! ”, I try to somewhat soften the blow, I am being positively polite. However, Brown and Levinson (1987: 101) make clear that positive politeness strategies can be of a very general kind. For example, one way of softening the blow to a hearer’s positive face would be to make clear that I still consider him or her as a member of the in-group, as a friend or as a person who I like in general. So, if I answered “Sorry, mate, you failed! ”, the address form mate indicates that I still like and appreciate you. Of course, there are approaches to politeness that would regard signals of appreciation as polite even if no face threat was involved. Leech’s (1983: 135) approbation maxim, for example, states that maximizing praise of the hearer is one way of acting politely. From a more general perspective Brown and Levinson suggest the following strategies when it comes to face-threatening acts (cf. figure 11.3). If I know that my next utterance will be a threat to your face, the first decision I have to make is whether I want to do the FTA or not. I could, for instance, decide not to ask you to close the window and close it myself. If I choose to go with the FTA, I can either go on record or off record. In the latter case I phrase my utterance in such a way that it could also be interpreted as being fully irrelevant with regard to your face needs. Utterance h) from the task above, “It’s cold in here.”, is a good example, since it might just be me making a statement about the temperature in the room. Going off record can be interpreted as the strongest form of politeness, since one possible interpretation of the utterance is one that does not threaten your face in any way. Going on 222 11 Texts and the relevance of the situation <?page no="224"?> redressive action conventional‐ ised indirect‐ ness record, in contrast, always involves a face-threat, there is no other way to interpret my utterance. Doing this without redressive action, going on record baldly, means that I phrase my utterance as directly as possible, e.g. “Close the window, please! ”. When I go on record with redressive action I am being polite in the strict sense. The term ‘redressive action’, in my view, is chosen particularly well, since the verb to redress means ‘to remedy, to set some‐ thing right’. In cases of positive or negative politeness this is exactly what I do: I am setting right my threat to your positive or negative face. Fig. 11.3: Strategies for doing FTAs (after Brown and Levinson 1987: 61). A final comment is in order here concerning off-record strategies. Consider the examples below. (11a) Close the window, please! (11b) Would you mind closing the window? (11c) It’s cold in here. (11a), obviously, is on record, (11c), obviously, is off record. What about (11b)? Strictly speaking, we could argue that it is off record, since I’m just asking a question whether the addressee would mind closing the window. However, this and similar indirect requests like “Can you pass me the salt? ” or “Can you tell me the time? ” have become conventionalised, i.e. the speakers of the English language have reached an agreement that questions of this kind are not to be understood as questions in the real sense but that they have to be inter‐ preted as a special kind of request. In that case Brown and Levinson speak of ‘conventional‐ ised indirectness’ and regard it as a form to go on record with redressive action. Before you read on … … imagine the following situation. You are in a restaurant with one of your friends. When the bill comes your friend realises that s/ he didn’t bring any money. Let us assume s/ he wants you to pay. For each of the four alternatives below name the relevant strategy for FTAs. a) Pay for me! b) Can you pay for me, too? 11.2 The interpersonal dimension - politeness 223 <?page no="225"?> c) Oh, I forgot my wallet. d) I’m sorry, but could you help me with a little cash? Option a) is an example of the baldly on-record strategy. The utterance can only be interpreted as an order or a request and there is no redressive action involved. Both, examples b) and c) are off record at first sight: b) asks a question about the ability of the hearer to do something, and c) makes a statement about the speaker. However, asking whether someone is able to do something for someone (Can you pay for me? ) has become a conventionalised form of request. Being a case of conventionalised indirectness, utterance b), therefore, needs to be analysed as an instance of negative politeness. c) is an example of the off-record strategy. The utterance can be interpreted as a request, but it could also be interpreted as an expression of surprise on the part of the speaker. This becomes evident if we imagine that immediately afterwards your friend said “I guess I’ll have to pay with my mobile, then.” Example d) shows three different ways of implementing negative politeness. Firstly, the speaker apologizes to the hearer (I’m sorry …). In addition, the use of the modal verb could (instead of can) signals that the speaker does not think that it is likely that the hearer can help him or her out. This is polite since the request becomes less coercive. As a consequence it becomes easier for the hearer not to fulfil the request. Finally, the phrase a little cash minimizes the degree of the imposition itself, thereby minimizing the face threat. After this discussion of the nature of politeness and the way it impacts language use we are now in a better position to really understand why some formulations are polite while others are not and we are able to explain different degrees of politeness. As we have seen, politeness is about face. More specifically, positive face “includes the desire to be ratified, understood, approved of, liked or admired”, and negative face is “the want of every ‘competent adult member’ that his actions be unimpeded by others.” (Brown and Levinson 1987: 62) Of course, section 11.2 can only scratch the surface of Brown and Levinson’s seminal work on politeness (after all, they list a total of 15 positive politeness and 10 negative politeness strategies). However, with regard to the EFL classroom I think it makes sense to focus on two aspects when it comes to politeness: Firstly, politeness is about signalling to the addressee that I am aware of his or her face needs, even though what I am about to say will threaten his or her face. Secondly, politeness is about mitigating the threat itself. I think it can be helpful to keep these two aspects apart, because they are taken care of by different parts of a message. Consider the following examples relating to positive politeness. (12a) You failed! (12b) I’m very sorry, but you failed! (12c) I’m afraid you failed. (12d) I’m terribly sorry, but unfortunately I have to tell you that you did not quite get the points necessary. (12e) No, you achieved a little less than the necessary 50-%. 224 11 Texts and the relevance of the situation <?page no="226"?> (12a) is bald on record, which in terms of the above means the speaker neither signals any awareness of face needs nor does the speaker mitigate the threat to positive face. (12b) and (12c), in comparison, do not soften the blow either, but we see an expression of regret on the part of the speaker that they have to give the addressee the bad news. Through I’m very sorry … or I’m afraid …, the speaker signals that s/ he is aware that the news is bad and that it goes against the addressee’s desire to be approved of or admired. In (12d) the speaker stresses his/ her awareness of the addressee’s face needs even more strongly: I’m terribly sorry, but unfortunately I have to tell you … In addition, (12d) mitigates the threat to the addressee’s positive face by stating that he or she failed only narrowly. (12e) does so, too, but it does not signal the speaker’s awareness of the addressee’s face needs. The same holds true for negative politeness. The examples illustrate that the two aspects of politeness, signalling awareness of face needs and threat mitigation, can be realised to various degrees. (13a) Take me to the airport. (13b) Can you give me a lift to the airport? (13c) Could you give me a lift to the airport? (13d) Sorry to bother you, but you have to take me to the airport. (13e) I hate to impose, but could you maybe give me a lift to the airport? It’s only ten minutes from here. The first of the examples above is bald on record since, like (12a) above, it neither signals awareness of negative face needs nor does it mitigate the degree of imposition. (13b) and (13c) are interesting in that the only difference between the two is the modal verb, can as opposed to could (as we have seen in the task above). You might have already learned in school that the second is more polite than the first, similar to the pairs may/ might or will/ would. On the surface both versions ask the addressee about their ability to take the speaker to the airport. However, we have seen above that the past tense form could suggests a lower degree of likelihood than the present tense form can. By expressing a lower likelihood the speaker emphasises more strongly that s/ he is aware of the imposition and, accordingly, of the face needs of the addressee. Sorry to bother you … in (13d) signals a general awareness of face needs, however, the speaker does not mitigate the threat to the addressee’s negative face, the request itself is rather blunt (even though it is phrased as a statement). In (13e) the awareness of face needs is signalled twice. I hate to impose … makes clear that the speaker is aware the s/ he is imposing and at the same time could you maybe … expresses a very low degree of likelihood that the addressee will be able to fulfil the request (indicating how outrageous the request is in the first place). At the same time (13e) reduces the degree of imposition by stating that it only takes a little time to take the speaker to the airport. By way of conclusion, take a look at example (14), which shows politeness on many levels. 11.2 The interpersonal dimension - politeness 225 <?page no="227"?> (14) I apologize for sending you this intrusive request for review as I know that academics are always asked to do so much for so little! Nevertheless, I was just wondering if you happen to have some time in the coming 6 weeks for checking this manuscript. […] Many many thanks for your consideration. The example is taken from an email I received. The writer signals a very strong awareness of my negative-face needs. He apologizes, he calls his request “intrusive” and he acknowledges that “academics are always asked to do so much for so little”. The use of many many thanks at the end of his e-mail emphasises his awareness of the imposition even more. The threat to my negative face is reduced through the phrase if you happen to have some time. The review does not appear to need too much time and happen to seems to suggest that I should only agree to do the review if by any chance there might be some time left in my budget, i.e. the writer gives me an easy option to deny his request. I hope to have shown that there is more to politeness than “the British are more polite than the Germans”. Politeness is not some vague idea but can be described with regard to distinct and clearly definable dimensions. As a consequence, expressions of politeness do not have to be learned by heart. You can help your students understand why and how politeness works, making them more competent in an important area of language use. New tools in your toolkit Spoken and written mode ▶ You have raised your awareness that spoken and written language are different and show different structures. In particular you have learned that filled and unfilled pauses, false starts and hesitation markers are completely natural phenomena in spoken language. Process and product ▶ You have learned that spoken discourse shows a tendency to recount events as a process, i.e. through clauses, whereas written language, in particular written academic discourse, represents events as products, typically as complex noun phrases. Fragmentation ▶ You have learned that spoken language would usually present only one idea at a time to the recipient. This makes it easier for the speaker to plan their utterance and makes it easier for the listener to process the speaker’s utterance. Integration ▶ You have learned that this is a feature primarily found in written discourse, often resulting in information-packed noun phrases. You have also learned that the structures are too complex for spoken discourse. Politeness ▶ You have learned that politeness is a concept that can be described quite precisely with regard to positive and negative face. Politeness as face work falls into two categories with regard to the speaker: 1) signalling an awareness of the addressee’s face needs and 2) attempts to mitigate face threats or to enhance the addressee’s face. In addition, you are more aware now of how language formulae serve these purposes. 226 11 Texts and the relevance of the situation <?page no="228"?> Positive politeness ▶ You have understood that positive politeness is primarily about finding the right way to say something that is inherently hurtful to the addressee. Negative politeness ▶ You have understood that negative politeness is primarily about the right way to ask a person to do something even though it restricts their freedom of action. If you want to know more B R O W N , Penelope and Stephen C. L E V I N S O N . (1987): Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. ▶ A classic on politeness. H A L L I D A Y , Michael. A.K. (1989): Spoken and Written Language. Oxford: OUP. ▶ A good introduction into the main differences of spoken and written English from a systemic-functional perspective. K R E Y E R , Rolf (2010b): Introduction to English Syntax. Frankfurt a.M.: Peter Lang. ▶ Chapter 8 provides a more detailed discussion of spoken and written syntactic structures. L E E C H , Geoffrey. (2014): The Pragmatics of Politeness. New York: Oxford University Press. ▶ Suggests a model of politeness that differs from Brown and Levinson’s face-based approach and applies this model to the analysis of (im-)politeness in English. Additional online content. 11.2 The interpersonal dimension - politeness 227 <?page no="230"?> Abstract Contents 12 The recipient 12.1 Humans - the not so rational animals 12.2 Communication and the unpacking of information 12.2.1 Exploiting the Cooperative Principle 12.2.2 Background knowledge 12.2.3 Don’t think of an elephant! 13 The producer 13.1 Packaging information 13.1.1 The right word at the right time 13.1.2 Beyond the word level 13.1.3 Of swarms and floods - the power of metaphor 13.1.4 Framing 13.1.5 (Faulty) logic and argumentation 13.2 Walking uphill The final part of this book deals with the power of language from the perspective of the recipient (i.e. what are features of communication and mental processes, that make us vulnerable to manipulation through language? ) and that of the producer (i.e. what are strategies that they can use to manipulate? , how can they exploit the recipient’s vulnerability? ). The topic will not necessarily figure in the EFL class‐ room. However, I’ve included this part for various reasons. First, an awareness of how language can be used for manipulation is becoming more and more important in times of social media, fake news and so on. This is why I think you as a teacher need to be made aware of the power of language and the problems it brings with it. If you are aware, you can make your students aware of them, too. Such an awareness is needed - now more than ever. The power of language Language is a weapon, keep it honed! (K U R T T U C H O L S K Y ) <?page no="232"?> 1 This section is largely based on Kreyer (in press). 12 The recipient But wait! Call now and we’ll also send you the mandolin slicer set free … and there is even more. Order right now and you’ll get this grater attachment … (N I C E R D I C E R P L U S C O M M E R C IA L ) What this chapter is about As humans we like to think of ourselves as rational beings or, at least, as capable of rational thought. While this is generally true, we are often a lot less rational than we think. In addition, there are features of communication, interpretation and processing that contribute to the efficiency of human language as a communication system but at the same time make us vulnerable to manipulation through language. The aim of this chapter is to explore this vulnerability. Manipulation through language would be a lot more difficult if we were as rational as we like to think we are. Unfortunately, this is not the case as the next section shows. 12.1 Humans - the not so rational animals 1 The idea that humans are special because we have reason was pointed out to us by Aristotle roughly 2400 years ago: Now, other animals live chiefly a life of nature; and in very few things according to custom; but man lives according to reason also, which he alone is endowed with; wherefore he ought to make all these accord with each other; for if men followed reason, and were persuaded that it was best to obey her, they would act in many respects contrary to nature and custom. (Aristotle 1332b) It is our faculty to use reason that distinguishes us from other animals. Cognitive linguist George Lakoff disagrees when he says: <?page no="233"?> That what makes us people is we’re all rational animals, and therefore we have the same reason, because we’re all human beings. So it follows from that: If you tell people the facts, that will lead them to the right conclusion. And, it doesn't work. The facts mean nothing […]. (Rosenberg 2014: n. pag.) “The facts mean nothing” is definitely a very strong claim, but in light of fake news, Pizzagate, QAnon and the often devastating effects of ‘alternative facts’ spread through social media all of us may agree that facts have come to be less important over the last few years. However, the quote above is from the year 2014. Whatever George Lakoff may have in mind must go deeper than Pizzagate, QAnon and bogus claims of a stolen election. There is a huge amount of research which shows how vulnerable our reason is when it comes to language use. In one strain of research Elisabeth Loftus (1975) explored the power of leading questions. Two groups of informants were asked about the frequency of headaches, as shown below. (1) a) Do you get headaches frequently, and, if so, how often? - b) Do you get headaches occasionally, and, if so, how often? As you can see, the questions only differed in the adverbial, i.e. frequently as opposed to occasionally. This difference, though arguably small, had a huge effect on the frequency with which headaches were remembered to occur. While the second group reported an average of 0.7 headaches per week, the first group reported over three times as many, namely 2.2 headaches per week. In another experiment, Loftus (1975) showed her informants videos of car crashes and traffic accidents. Again, she asked two groups two slightly different questions. (2) a) Did you see the broken headlight? - b) Did you see a broken headlight? Group A reported to have seen the broken headlight in 15 % of all cases, while group B answered in the affirmative in only 7 % of all cases. Again, we see how strong the influence of the definite article in comparison to the indefinite article is. This is even more surprising if we take into consideration that there was no broken headlight at all. Loftus and Palmer (1974) asked their informants to estimate the speed of cars involved in a traffic accident. This time differences in wording concerned the verb used: “About how fast were the cars going when they hit/ smashed into each other? ” While the estimated speed with the more neutral verb hit was 34.0 mph, the less neutral smash led to a considerably higher average estimate of 40.8 mph. “Okay,” I hear you say, “that’s interesting and impressive, but it has nothing to do with reasoning, only with memory” and you would be right, of course. However, it has also been shown in a number of experiments how the choice of words and syntactic constructions can influence our decision-making processes. 232 12 The recipient <?page no="234"?> Before you read on… … take a look at the following experiment. Tversky and Kahnemann (1981: 453) describe an outbreak scenario as shown below and divided the informants into two groups, which were given the respective information on possible interventions that are shown under a) and b). Imagine that the U.S. is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease, A and B, have been proposed. a) If Program A is adopted, 200 people will be saved. If Program B is adopted, there is 1/ 3 probability that 600 people will be saved, and 2/ 3 probability that no people will be saved. b) If Program A is adopted, 400 people will die. If Program B is adopted, there is 1/ 3 probability that no people will die, and 2/ 3 probability that 600 will die. To what extent do Programs A and B differ with regard to the number of survivals and deaths in cases a) and b)? If you have wondered whether I have made a mistake or whether there was a typo, you did well. With regard to the facts presented there is no difference between the number of survivals and deaths in cases a) and b). The only difference is in the wording: in a) 200 out of 600 people will be saved with Program A, which is the same as 400 out of 600 people dying in b). With Program B there is a likelihood of 1/ 3 of 600 people being saved a), which is the same as none dying b) and a likelihood of 2/ 3 of no people being saved a) or 600 dying b). While the facts are identical in both cases the differences in wording still had a huge effect on the decision-making process. Those informants that were given the description shown in a) opted for Program A in 72 % of all cases and for Program B in 28 %. Those given the description in b) decided for Program A in 22 % of the cases and chose B in 78-%. The wording alone makes a huge difference. The next experiment makes clear that linguistic structures can have an impact, too. An experiment by Langer et al. (1978) involved a queue in front of a copy machine (or: Xerox machine) and a researcher who asked for permission to jump the queue. The experimental conditions involved two dimensions. The first was the extent of the favour: the researcher claimed they wanted to do either 5 pages or 20 pages. The second dimension concerns the absence or presence of an excuse and whether the excuse contains only placebic information or real/ sufficient information. This leads to a total of six possible requests (3a-f) on the part of the researcher who wants to jump the queue: 12.1 Humans - the not so rational animals 233 <?page no="235"?> (3) a) small favour, no reason: Excuse me, I have 5 pages. May I use the Xerox machine? - b) large favour, no reason: Excuse me, I have 20 pages. May I use the Xerox machine? - c) small favour, placebic information: Excuse me, I have 5 pages. May I use the Xerox machine, because I have to make copies? - d) large favour, placebic information: Excuse me, I have 20 pages. May I use the Xerox machine, because I have to make copies? - e) small favour, sufficient information: Excuse me, I have 5 pages. May I use the Xerox machine, because I’m in a rush? - f) large favour, sufficient information: Excuse me, I have 20 pages. May I use the Xerox machine, because I’m in a rush? Table 12.1 below shows the results of this experiment in terms of the proportion of successful requests. Favour No information Placebic information Sufficient information small: 5 pages 60% 93% 94% large: 20 pages 24% 24% 42% Table 12.1: Proportion of successful requests in dependence on extent of favour and quality of information (Langer et al. 1978). Interestingly, if the favour is small, it does not seem to make any difference whether the researcher provided placebic or sufficient information. In both cases almost all subjects agreed to let the researcher jump the queue. The situation is different if the favour is large: only if the researcher provided a real reason did the subjects feel more inclined to let him or her jump the queue. What can we make of this? It seems that with decisions that are not particularly costly, as in the case of the small favour, the mere presence of a causal subordinate clause lets the test subjects assume that the researcher has a reason for the request. The causal structure seems to be processed more or less automatically without checking whether the request is reasonable or not. In case of costly decisions, test subjects are more prone to check whether the request does actually make sense. To sum up, the experiments above have shown that different subsystems of the language system seem to have an impact on our memory and reasoning. Example (2) showed grammatical meaning (definite vs. indefinite article) to have an impact, examples (1) and those in the task above demonstrated the relevance of lexical meaning, while example (3) showed the influence that mere grammatical structures can have. All in all, we are not as rational as we like to think we are. This makes us vulnerable to manipulation through language. In the following we will explore aspects of the 234 12 The recipient <?page no="236"?> conduit metaphor communication process that add further to the recipient’s vulnerability. The next chapter, then, will describe ways in which this can be exploited by the producer. 12.2 Communication and the unpacking of information The layperson’s understanding of communication has been described aptly in Michael Reddy’s (1979; also see Lakoff & Johnson 1980: 10-12) ‘conduit metaphor’. According to this metaphor communication involves three major steps. First, a sender has some thoughts in their mind. These thoughts are put into a package, i.e. they are expressed in a text. Second, the package is sent through a channel (or conduit) to a receiver. This channel is any kind of connection that allows the transferral of this text, e.g. sound‐ waves transmitted through the air or black marks on white paper or on a screen etc. Third, when the text package reaches the receiver, they open the package and the information from the sender’s mind now also is stored in the receiver’s mind. For the purposes of this chapter we will focus on this third stage. The conduit metaphor leads to a common misconception when it comes to commu‐ nication. If a person puts an object into a package and sends this package to us we will be in possession of this very same object once we open the package. Obviously, language does not work that way. Let us assume a friend of yours sends you the message “Having a great time at the beach! ” This message will lead to the creation of a certain image in your mind. Let us assume that for you that would mean lying in the sun, reading a good book, sleeping and going for an occasional swim. These would, then, be the images that are conjured up by the message and you start to envy your friend. A few minutes later your friend sends you pictures of them in a large crowd, lots of booze and big loudspeakers in the background - and suddenly you do not envy them that much anymore. A text, obviously, never is the thing itself or a fully objective representation of the thing. It cannot be. A text is only a stimulus that leads the recipient to create images in their mind. The conduit metaphor suggests that the recipient is fairly passive when it comes to unpacking information, when it comes to creating mental images from text. In actual fact the recipient is very active when doing that. The unpacking of the text package is influenced by a large range of assumptions about the communication process itself and about how the world works. These assumptions, as we will see in what follows, can be exploited and make us susceptible to manipulation. 12.2.1 Exploiting the Cooperative Principle Before you read on … … take a minute to think about this: When I was a PhD-student I stumbled across a homeshopping channel where they were selling some product. I do not know 12.2 Communication and the unpacking of information 235 <?page no="237"?> quality exactly which anymore, let us call it ‘the Knife King’. Being desperate for a break from analysing linguistic data, I watched the commercial till the end where I was told the following: “If you order tonight, you will get a second Knife King for free! ” Being desperate for a break from analysing linguistic data, I immediately ordered the product. Later that month and, again, being desperate for a break from analysing linguistic data, I was watching the same homeshopping channel again and the same commercial was on. To my surprise I heard “If you order tonight, you will get a second Knife King for free! ” Being young and naïve and not the linguist that I am now, I felt that I had been lied to. Was I right to feel betrayed? The answer is a complex one. In a very strict sense I had not been lied to. I was told that if I ordered at a particular point in time I would get a second product for free, and that is exactly what would have happened. On the other hand, I was somehow led to believe that only if I ordered at this particular point in time would I get the second Knife King. Obviously, this was incorrect, and, a little less obviously, the advertisers never actually said this. However, they counted on me interpreting their message this way. At the heart of faulty interpretations like the one described above is our assumption that humans generally cooperate when they communicate. Why should the advertisers say “If you order tonight …” if it was not only tonight that they would send me a second product for free? There is no point in adding this condition if it is no real condition at all. At least this is what I assume if I take it for granted that the advertisers behave in a cooperative fashion (as I said above: I was young and naïve). Actually, my assumption was not that naïve. When communicating, interactants would usually assume that everyone behaves cooperatively. This assumption makes a lot of sense because it makes communication easy: just imagine the amount of effort it would take to challenge and question everything that others say to you. In a way we have to work under the assumption of cooperation because otherwise communication would break down. What is more, we usually experience that our assumption is correct. People usually do interact cooperatively in communication. But what does that mean? An interesting theory on the nature of co-operative communication was suggested in 1975 by Herbert Paul Grice’s Cooperative Principle. He writes: Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged. (1975: 45) More specifically, he suggests that communication can be regarded as cooperative if the interlocuters follow guidelines (referred to as the ‘Gricean Maximes’) in four areas, namely quality, quantity, relation and manner. Quality has to do with the truth of what we say. Do not say anything that you know to be false and do not make any claims for which you lack evidence. For example, let us assume you want to know whether you have passed the final exam in one of my courses. I know that you have failed. If I 236 12 The recipient <?page no="238"?> quantity relation manner followed the Maxim of Quality, I would tell you the truth: “I am very sorry, but you failed”. To lie and say “Congratulations, you’ve passed! ” would be violating the maxim. Quantity has to do with the amount of information that is being transferred: do not say more than is required and do not say less than is required. An answer in line with quantity would be a simple “Yes” or “No”, or “Yes, you passed” or “I am very sorry but you didn’t”. If I said something like “You did really good on the first half ”, I would not give you the amount of information that you are looking for, I would say less than is required. If I told you that you failed and also told you the points you got for each task and discussed your mistakes with you, that would be too much information for you at that point, at least more information than you have asked for. I would say more than is required. It is important to note that quantity is independent of quality. Sometimes students of mine argue that if you lie you do not provide the necessary amount of correct information. While I do understand this line of argumentation we should still keep the two aspects of cooperation apart: if my lie was successful you would not think that I had not given you enough information. However, if I only told you that you got the maximum points in the first three questions or so, you would maybe ask me “And, have I passed? ” So, we can clearly distinguish quality from quantity and should do so. In addition, our contributions should be relevant, the respective maxim is called Maxim of Relation. To follow the maxim means to make a contribution that is related to the issue we are currently discussing. For example, even if I lied to you and said “Con‐ gratulations, you’ve passed! ”, I would still be following the Maxim of Relation, because what I say does relate to your question, even though it is not factually correct. The same is true for my maybe somewhat evasive answer “You did really good on the first half ”. Even though it does not provide you with all the information you are looking for, it does provide you with relevant information. An irrelevant answer would be something like “I think term papers are better ways to assess somebody’s expertise”. This is not in any way related to the original question and, hence, not relevant. If I gave you that answer you might respond “That’s all very interesting, but could you please tell me whether I’ve passed or not? ” If you think that I have already told you through the fact that my answer was evasive, you are right. For the moment, however, it is important to distinguish Relation from Quality and Quantity. A contribution is relevant if it relates to the issue that we are currently talking about, even if it violates Quality and Quantity. The final maxim is that of Manner. It demands of us that in our contribution we should avoid overly complex vocabulary and syntax. All of the answers above are in line with that maxim: irrespective of whether they follow the other maxims or not, all of them are clear and concise. An example of a violation of Manner is the following answer: “A comparison of the minimum number of points necessary for passing the exam with the number of points you have achieved leads me to conclude that I’d be lying if I said ‘Yes’.” It follows all of the three previous maxims: it is true, it is the right amount of information and it is relevant, the answer is just unnecessarily complex and complicated. It is often not easy to keep manner apart from quantity, particularly in 12.2 Communication and the unpacking of information 237 <?page no="239"?> those cases as the one above, when the answer is really, really long. Often this leads students to conclude that too much information is given. However, this is not the case. My answer does not actually give you a lot of information. It merely tells you that you have not achieved the points necessary for passing and that you have not passed. From a logical perspective this is equivalent to the shortest possible answer to your question: “No! ” My answer just used a lot of words to provide you with the same amount of information. Grice himself has already pointed out that the Cooperative Principle is not limited to verbal interaction but can be related to any kind of human interaction. The following task (based on Grice) should make that clear. Before you read on … … let us assume you want to bake a cake and ask me for 200 g of sugar. Each of the actions below violate one of the maxims. Can you identify which one? 1. I give you 200-g of sugar, but not as one cup but in 20 teaspoons. 2. I give you 200-g of salt. 3. I give you 170-g of sugar. 4. I pre-heat the oven for you. Again, it is very important to keep the individual maxims apart. What is needed at this moment in the baking process is 200 g of sugar. The first action gives you exactly that amount of sugar but it makes the whole procedure unnecessarily long and complicated, i.e. we are talking about a violation of Manner. If I give you 200 g of salt, I violate Quality. You cannot tell salt apart from sugar if you do not taste it. You think I am giving you what you ask for, but I actually give you something else and it is not obvious to you. This is similar to lying in a talk exchange. If I give you only 170 g of sugar I give you what you are asking for but I do not give you enough of it. I violate the Maxim of Quantity. Finally, pre-heating the oven might be relevant at some other stage in the baking process but it is not relevant at this stage when you ask me to give you sugar. The Maxim of Relation is violated. By the way: Saying more than you are actually saying - flouting The cooperative principle can also account for the fact that sometimes we understand more than was actually said. For example, if you ask me “Have I passed the exam? ” and I answer “I think term papers are better ways to assess somebody’s expertise,” you will maybe interpret that as a hint that you have not passed. Grice suggests the concept of ‘flouting’ to explain this phenomenon. When flouting a maxim, speakers do not follow a given maxim but they make it very obvious so that their addressee is aware that they do not follow the maxim. The addressee, however, would still work under the assumption that the speaker co-operates and 238 12 The recipient <?page no="240"?> exploiting the coopera‐ tive princi‐ ple exploiting quantity scalar impli‐ cature they will search for a ‘hidden’ meaning. So, you might think “Why would s/ he tell me that? ”, which would lead you to the conclusion that you have not passed. Grice calls this conclusion a ‘conversational implicature’. As long as we do not have any reason to think otherwise, we assume that others behave cooperatively if they communicate with us. Often this assumption of cooperation is exploited by other people. This is where the Cooperative Principle becomes relevant with regard to the power of language and the use of language for manipulation. We expect that the people we talk to follow the four maxims described in this section. So, when I tell you that two weeks ago (from the time of writing this passage) we got our dog back from the vet, you have no reason to assume that that would not be true, because usually people act cooperatively and, accordingly, would follow the Maxim of Quality. This good-will and our assumption that people act cooperatively can be ex‐ ploited by others. The most obvious case is that of lying. We have seen a lot of this during the Trump presidency. A famous (or rather: infamous) example is the lie of the stolen election of 2020. Even though no-one has ever been able to provide any evidence that backs up Trump’s claim, a poll by Quinniac University after the 2020 election showed that “77% [… of Republicans] believe[d] there was wide-spread voter fraud” (https: / / poll.qu.edu/ Poll-Release? releaseid=3734). As for the exploitation of the Maxim of Quantity consider the following example. Let us assume you took a 20 Euro bill from your friend’s wallet without telling them. Later your friend finds out that there is less money in their wallet and asks you whether you took 10 Euros without telling them. If you answer “Yes” you have exploited the Maxim of Quantity. You know that you have taken more money from your friend and you know that your friend would be interested in the exact amount. Your friend as‐ sumes that you would tell them if it was more than 10 Euros that you took. This as‐ sumption is an example of what is called ‘scalar implicature’. Whenever we talk about things that can be quantified or mapped onto a scale we assume that our conversational partner makes the strongest possible statement. The advantage with exploiting the Maxim of Quantity is that it is not a lie in the strict sense, which makes it a useful tool for politicians. They may report or admit to something but will not disclose the full extent of what is actually going on. The public, however, assumes that the politician adheres to Quantity and, as a consequence, the public believes that there is not more to tell. At the time of writing President Joe Biden is accused of doing just that. Classified documents were found in one of Joe Biden’s offices in Washington and, later, in his home in Delaware. When Biden first spoke to the public about the classified documents in his Washington office he already was aware of the second batch of documents but did not mention these. As a consequence of scalar implicature, the public was led to believe that there were no other classified documents in places where they were not supposed to be. Again, the advantage here is that Biden can say he never lied about the documents. 12.2 Communication and the unpacking of information 239 <?page no="241"?> exploiting relevance whatabout‐ ism The Maxim of Relevance is frequently exploited in advertisements. We have already discussed the example of the Knife King and the statement “If you order tonight, you will get a second Knife King for free! ” Under the assumption of cooperativeness you would assume that this bit of information is relevant. This leads you to conclude that it is only if you order tonight that you will get the second Knife King. If you order later, this will not happen. Otherwise they could just have said “buy one get one free! ”, right? So you are lead to believe that this really is a special offer that is only valid tonight, which it obviously was not. By the way: Faulty Reasoning and logical fallacies From a logical perspective, there is something else going on as well, namely faulty reasoning. A valid form of logical conclusion is that of modus ponens. It is a pattern of reasoning that involves two premises, 1) an implication of the form ‘if A is true, then B is also true” and 2) “A is true”. On the basis of these premises we reach the logical conclusion “B is true”. For example, “if you cheat, I will fail you.” In this case A is “you cheat” and B is “I will fail you”. If from A follows B the following is also true: from not B follows not A, i.e. “if I do not fail you, you have not cheated (or at least you have not been caught)”. On the basis of the additional premise “B is not true” we can then conclude “A is not true”. This pattern is referred to as modus tollens. While modus ponens and modus tollens lead to correct deductions, people often come to incorrect conclusions. For example they might think “if I do not cheat, he will not fail me.” This, of course, is not true since I also fail people if they do not get the necessary points to pass the exam. It is this faulty reasoning that lines like “If you order tonight, you will get a second Knife King for free! ” exploit. People will often assume that if they do not order tonight, they will not get the second Knife King for free. This assumption is logically incorrect. We will explore more examples of this kind in section 13.1.5. Another example of the exploitation of Relevance is the following. I was once going through a shopping mall and I saw a poster that advertised a bicycle helmet. The poster showed a very reasonable price and, which was more interesting to me, it advertised that the helmet was Testsieger 2019. Assuming that the advertisers were following the Maxim of Relevance, I thought that this would be a really good purchase as the helmet was cheap and, obviously, really good (i.e. safe) since it was Testsieger. However, when I got closer to the poster, the small print told me that the helmet won the award of the journal Ökotest. The advertiser exploited Relevance. They counted on people assuming that the helmet won an award for its protectiveness, the feature that is most relevant when it comes to bicycle helmets. I am certain I was not the only one disappointed. Politicians will frequently violate the Maxim of Relevance. One particularly effective form is the so-called ‘whataboutism’. Instead of giving, say, an interviewer an answer 240 12 The recipient <?page no="242"?> about X the addressee will counter with a question of the kind “Well, what about Y? ”, thereby dodging an uncomfortable situation and redirecting the focus of the inter‐ viewer and the viewers or listeners at home. A number of illustrative examples of whataboutism are given by John Oliver’s segment on whataboutism in his show Last Week Tonight. Take as an example the way Trump reacted to questions in the context of the Charlottesville incident. On the 12 th of August 2017 a group of white nationalists were marching through Charlottesville in Virginia, which led to counter protests. One right-wing extremist purposely drove his car into a group of counter protesters, killing one woman and injuring at least 19 other people. Asked as to why he had not con‐ demned right-wing extremism more strongly in response to Charlottesville, Trump said the following: (1) What about the alt left? They came charging at the, as you say, alt right. Do they have any semblance of guilt? […] What about the fact that they came charging, that they came charging with clubs in their hands, swinging clubs, do they have any problem? (https: / / youtu.be/ RS82JNd0YzQ - 0: 25) Note that “Trump’s favorite dodge to a tough question” (https: / / time.com/ 4941771/ do nald-trump-whataboutism-rhetoric/ ) is also frequently used by others. The following examples (also presented by John Oliver) are taken from FoxNews: (2) The mainstream media focused on the Trump campaign and allegations of collusion with the Russians. But what about the Democrat’s possible ties to Moscow? (https: / / youtu.be/ RS82JNd0YzQ - 1: 28) (3) Former Security Advisor General Michael Flynn investigated for his private meeting with Russia, but what about Hillary Clinton? (https: / / youtu.be/ RS82JNd0YzQ - 1: 36) (4) The media wants to call into question the credibility and trustworthiness of this administration, but what about Benghazi? What about the blatant lies that the Obama administration told us? What about the fact that Ben Rhodes bragged about lying to the media and the public about the Iran deal? What about the fact that Jonathan Gruber basically said the American people was stupid? (https: / / youtu.be/ RS82JNd0YzQ - 1: 43) Oliver rightly observes that whataboutism “implies that all actions regardless of context share a moral equivalency” (https: / / youtu.be/ RS82JNd0YzQ - 1: 10). Whether you purposely drive your car into a goup of people or whether you charge at a group of people with clubs is basically more or less the same. While both actions are to be condemned forcefully, it might be argued that there is a difference between the two. Similarly, Trump seems to have been a much more prolific liar than Obama. The New York Times, for instance, reports that in his first 100 days of presidency, Trump had already lied six times more frequently than Obama in his whole eight years as President (https: / / www.nytimes.com/ interactive/ 2017/ 12/ 14/ opinion/ sunday/ trump-li es-obama-who-is-worse.html). With regard to this seeming moral equivalency Oliver continues: “since nobody is perfect all criticism is hypocritical and everybody should do whatever they want. It is a depressingly effective tool.” (https: / / youtu.be/ RS82JNd0 YzQ - 1: 08) 12.2 Communication and the unpacking of information 241 <?page no="243"?> The Maxim of Manner can be exploited to signal to the addressee that he or she would not understand what the speaker is talking about anyway. Let us assume the following (of course, 100 % hypothetical) scenario. You are in one of my seminars and ask me a question that I cannot answer. I do not want to admit that I do not know the answer but at the same time I have to say something. One possible strategy for me is to talk about something in such a complex and technical way that it intimidates you. You do not even try to find out whether what I said makes any sense and you accept my answer even though it did not help you at all. In this case, I violated the Maxim of Manner and I did it to hide my own ignorance. The German ‘comedian’ Georg Schramm provides another interesting example: (5) Wenn Sie einen Politiker fragen „Wie ist das mit dem Abzug deutscher Soldaten [aus Afghanistan]? “, dann können Sie - ich geben Ihnen aus dem Gedächtnis eine Kostprobe - folgende Antwort kriegen, zum Beispiel: "Die Dislozierung unserer Streitkräfte muss unter der Prämisse einer nachhaltigen Restrukturierung betrachtet werden, wobei eine Übergabedividende unter Berücksichtigung einer Reinvestierung in noch nicht befriedete Regionen betrachtet werden muss.“ Wenn Sie den Satz verstehen, hat er seinen Zweck verfehlt. […] Sehen Sie, dieser Satz - den sollen Sie nicht verstehen. Herrschaftssprache soll bewirken, dass Sie verzichten drauf. […] In Originaltexten würde es einfach heißen: „Wir zieh’n erstmal keine Soldaten ab. Wenn wir welche überhaben, kommen sie dahin, wo wir noch mehr brauchen.“ Das würden Sie verstehen, da würden Sie auch mitdiskutieren wollen. (https: / / youtu.be/ wiiqNx-3Goo; 4: 20) So, if you hear your teachers or politicians talk in an extra complex way with lots of technical terms, it might be worth taking a closer look. 12.2.2 Background knowledge In the previous section we have explored the relevance of the communication process itself when it comes to questions of manipulation through language. We will now take a closer look at the influence of background knowledge. Before you read on … … please take a look at the following little story. A father and son are in a horrible car crash that kills the dad. The son is rushed to the hospital; just as he’s about to go under the knife, the surgeon says, “I can’t operate - that boy is my son! ” What’s going on? If you find it hard to solve this riddle, don’t worry, you are not the only one. I had been working on it for over ten minutes when I decided to ask my wife. She found the answer, but at least she also had to think about it for two minutes. This riddle has been 242 12 The recipient <?page no="244"?> background knowledge inference used in a number of studies and the picture is always more or less the same: the ‘correct’ answer, i.e. the answer that researchers were looking for, would usually only be given by a small fraction of the informants. Many subjects, for instance, would suggest that we are dealing with a same-sex couple, so that the other father is the surgeon. Other responses were that the surgeon is just confused, that all of it was just a dream or that the father in the story was a catholic priest (Barlow 2014: n. pag.). The answer that the researchers were looking for is that the surgeon is the boy’s mother. Apparently, this does not really fit the experience and the worldview of the informants. I know that it did not fit mine, when I failed to come up with that explanation. To make things worse, when I first encountered this riddle, I was already interested in the field of language and society and, more specifically, in the influence that language can have on our representation of gender. What is more, I read the riddle as part of a thesis that explored the relation of language and gender. Still, I was not able to find the solution. This came quite as a shock to me: after all, I consider myself as all for gender equality and as more aware than average when it comes to language and the representation of gender. Relief came when I learned that in one study only 15 % of children between 7 and 17 suggested the surgeon might be the mother and that only 14 % of students gave this answer. Even among self-described feminists only 22 % came up with this explanation. The numbers are similar for an alternative version of the story, where a mother and a daughter are in the car crash and a nurse (in this case the father) cannot attend to his daughter (Barlow 2014: n. pag.). What we can learn from this riddle is to what exent our world knowledge, our understanding of how things are influences the processing of a message. Whenever we hear or read something, we interpret the message on the background of how we think the world works. In the previous section we have seen how our assumptions about communicative behaviour can be exploited for manipulation. We will now turn to the role that background knowledge plays when it comes to interpreting a message. The conduit metaphor suggests that the receiver is fairly passive when it comes to the decoding of the message: it assumes that exactly the same information that the sender has put into words is retrieved from these words as the reader processes them. This is not true, since we would usually interpret any message on the basis of our background knowledge. This is why I had so much difficulty concluding that the sur‐ geon is the boy’s mother. However, background knowledge not only helps (or hinders) us in understanding a message, it also enriches the message with regard to details that have not been explicitly mentioned. For example, above I told you that we took our dog back from the vet. This changes whatever image you have of me in that it becomes part of your conceptualisation of me that I own a dog. Note that this is not something that I have stated explicitly, as for instance in I have a dog. Two weeks ago I took it back from the vet. Still, the effect is more or less the same. This bit of information becomes part of your representation of me. This is a natural process. It is called ‘inference’ and denotes the process of deriving information from a message that is not explicitly stated in the message itself. 12.2 Communication and the unpacking of information 243 <?page no="245"?> The effect of such inferences has been documented in many psycholinguistic experiments. Stanfield and Zwaan (2001), for instance, were able to show that objects in a picture were recognised faster if the orientation of the object matched the orientation of the object in a sentence test subjects had previously read. For example, they read sentences like those shown in (6). (6) a) Susan drove a nail into the wall. - b) Susan drove a nail into the floor. After reading the sentence, they were shown a picture of an object and they had to decide whether the object was named in the sentence or not. In case of the sentences above subjects were faster in recognising the nail if the orientation of the nail in the picture was in line with the situation described in the sentence. Those that had read (6a) were faster if the nail was in horizontal orientation. If the nail was pointing downwards, those subjects were fastest that had read sentence (6b). Neither of the sentences under (6) make any reference to the orientation of the nail, however, our own experience tells us how we hold a nail if we want to drive it into a wall or the floor. It seems as if the verbal stimulus leads us to imagine the whole situation, which includes a particular orientation of the nail. By the way: The effects of inferences on memory Almost 30 years before Johnson et al. (1973) had shown that our world knowledge influences our memory of stories that we have heard. Two groups of test subjects were presented with twenty stories similar to those shown in shown in a) and b). a) John was trying to fix the birdhouse. He was pounding the nail when his father came out to watch him and to help him do the work. b) John was trying to fix the birdhouse. He was looking for the nail when his father came out to watch him and to help him do the work. After having listened to the story, the informants had to decide whether sentences from a list had been part of the stories they had heard or not. An example of such a test sentence for the stories in a) and b) is the following: c) John was using the hammer to fix the birdhouse when his father came out to watch him and to help him do the work. If you compare c) with the two versions in a) and b), you can see that the third sentence does not occur in any of the two story versions. However, we would usually assume that pounding in a nail would be done with a hammer. c), then, would be in line with the world knowledge activated by the story told in a) - c) would be an implication of a). Interestingly, the informants would often think that sentences of the c) type had actually been part of the story presented to them: 244 12 The recipient <?page no="246"?> presupposi‐ tion entailment “The S’s were likely to think that information available only by implication was actually given during the […] task” ( Johnson et al. 1973: 204). Inferencing happens all the time in everyday language use. It is one of the features of the way we use language that makes language such an efficient tool. This becomes particularly apparent with the phenomenon of presupposition. The term describes an implicit assumption. i.e. something that we take for granted even though it has not been explicitly stated. The typical example in linguistics is The King of France is bald. Even though the sentence does not explicitly state that France has a king, it presupposes that this is the case. Above I mentioned that two weeks ago we took our dog home from the vet. This sentence presupposes that we have a dog, even though it does not make it explicit. Presupposition has to be distinguished from entailment. An entailment is a form of logical deduction. For instance, the sentence The King of France was murdered entails the sentence The King of France is dead. How can we, then, tell the two apart? A pre‐ supposition ‘survives’ the negation of the sentence it is triggered by, an entailment does not. So, the sentence The King of France was not murdered still presupposes that France has a king. However, we lose the entailment that the King of France is now dead. This does not mean that the king has not died from some other cause. The point is that we cannot derive that bit of information from the first sentence anymore. Why should presuppositions and presupposed information be a problem? How can presuppositions be used to manipulate people through language? The answer is simple. Presuppositions are everywhere in everyday language. They are very frequent and a part of what makes language such an efficient means of communication. We do not have to make everything explicit when we talk. We can trust the addressee to fill gaps if that is necessary. Take as an example my sentence from above that we got our dog back from the vet two weeks ago. I did not have to state explicitly that we have a dog and I did not have to state that we had taken it to the vet. All of this was understood since it was presupposed. Through presupposition I was able to communicate a lot more than I actually stated. This is very efficient. By the way: Pressupositions - efficient, but also effective In their textbook on the language of advertising Sedivy and Carlson (2011: 97-99) report on the use of presupposition in Bill Clinton’s 1996 presidential re-election campaign. Apparently, Clinton had difficulties in ‘selling’ his achievements regarding the creation of new jobs. One of Clinton’s consultants, Bob Squier, suggested a change of wording. For instance, instead of an explicit statement along the lines of “I have created seven million jobs during my last presidency”, Squier suggested to background this achievement through statements like “The seven million jobs we’ve created won’t be much use if we can’t find educated 12.2 Communication and the unpacking of information 245 <?page no="247"?> people to fill them.” (Morris 1999: 183; qtd. in Sedivy and Carlson 2011: 97) The creation of the seven million jobs is not the main point of the message, instead it is presupposed. This helped to communicate Clinton’s achievement much more effectively. Unfortunately, this efficiency comes at a price. In conversation we would usually focus on those bits of information that are stated explicitly. Presuppositions that are triggered would usually not be in focus and, as a consequence, they are not likely to be challenged. This is what makes presupposition a very useful means of manipulation. Through the exploitation of presupposition we manage to add information to your stock of world knowledge without you recognising that it is happening. For example, when I said that we took our dog back home from the vet, this presupposed that we have a dog, which, by the way, is not true. We do not have a dog! In addition, since I presented this bit of information as backgrounded it feels like it may be taken for granted, like something we both agree on, like information both of us share: “This fact gives presupposition a curious property of undeniability” (Sedivy and Carlson 2011: 99). Of course, you would also have believed me if I had explicitly stated that we have a dog, but I hope you agree that the vet part was processed more consciously than the part of us having a dog. Before you read on … … take a look at the following tweet by former US President Donald Trump. It shows that presuppositions can be nested inside other presuppositions. How many nested presuppositions can you count and which are they? Do you notice the Fake News Mainstream Media never likes covering the great and record setting economic news […]; 01-16-2018 The innermost presupposition is that there are great and record setting economic news. As with the King-of-France example, the mention of an entity, a state or an event presupposes the existence of this event. The next-level presupposition is that the Fake News Mainstream Media do not like covering these news. Note the effort it takes to challenge these presuppositions. Your first reaction to this tweet might be “Oh, I’ve never noticed that.” Afterwards you might come to think “Hang on, do the Mainstream Media really not talk about that? ” If you have arrived at that step you might, then, start to challenge the presupposition about the economic news: “Wait a minute! Do we actually have a great and record setting economy? ” I hope you can see the effort it takes to untangle all of the presupposed information that is hidden in this tweet. And if you have done all of that, if you have taken the time to challenge all of this content, it is very likely that you might have overlooked the very basic presupposition that suggests that there is such a thing as the “Fake News Mainstream Media”. I hope that this has demonstrated the danger of presuppositions. 246 12 The recipient <?page no="248"?> definite descriptions factive verbs As I said above, presupposed information is likely to enter our stock of world knowledge subconsciously. Fortunately, we can train ourselves to become more aware of presuppositions and we can learn to look out for words, phrases and structures that serve as warning signals. A useful list of such signals is provided by Sedivy and Carlson (2011: 102-104). Definite descriptions presuppose the existence of something. They involve noun phrases that either start with a determiner that expresses definiteness, i.e. the definite article or a possessive determiner, as in our initial example The King of France is bald, which presupposes the existence of a king of France. Other examples are the following, the most interesting presupposed parts are printed in bold letters: (7) Do you notice the Fake News Mainstream Media never likes covering the great and record setting economic news […]; 01-16-2018 (8) The forgotten men and women of our country will be forgotten no longer. (Trump, n.pag.) (9) The GOP is consumed by its own conspiracy theories as establishment Republican jump ship. (Stephen Colbert) We have already discussed the first of the three examples above. The definite descrip‐ tions presuppose the existence of Fake News Mainstream Media as well as that of great and record setting economic news. In example (8), the existence of forgotten men and women is presupposed. (9) presupposes that the Republican Party does in fact entertain conspiracy theories (of course, it also presupposes that there is a GOP and that there are establishment Republicans). Another way of marking something as presupposed is the use of factive verbs. They describe a mental state towards a state of affairs, which is, then, marked as a fact. These states could be knowing, regretting and so on. Consider the examples below: (10) We know that the nation that goes all-in on innovation today will own the global economy tomorrow. (Barack Obama; https: / / millercenter.org/ the-presidency/ presid ential-speeches/ january-28-2014-2014-state-union-address) (11) I understand that many Americans view the future with some fear and trepida‐ tion. ( Joe Biden; https: / / millercenter.org/ the-presidency/ presidential-speeches/ janua ry-20-2021-inaugural-address) (12) We’re proud to have donated nearly 140 million vaccines over 90 countries, more than all other countries combined, including Europe, China, and Russia combined. ( Joe Biden; https: / / millercenter.org/ the-presidency/ presidential-speeches / september-9-2021-remarks-fighting-covid-19-pandemic) In the first example, the factive verb to know introduces the content presented in the subordinated clauses as presupposed and as fact. In the second example the verb is to understand. As in the case of to know, you can only understand things that are actually the case, hence the example presupposes that “many Americans view the future with some fear and trepidation.” The factive expression in example (12) is to be proud to have done X. As in the cases above, the X, i.e. the number of vaccines donated, is presupposed. 12.2 Communication and the unpacking of information 247 <?page no="249"?> continua‐ tion/ change of a state of affairs iterative expressions subordinate clauses Similarly to factive verbs, verbs that describe a continuation or change of a state of affairs, will presuppose that this state of affairs is actually the case. Consider the ex‐ amples below. Again, the presupposed information is in bold print. (13) The divisions of the moment shouldn’t stop us from doing the right thing for the future. ( Joe Biden; https: / / millercenter.org/ the-presidency/ presidential-speeches/ ma rch-31-2021-announcing-american-jobs-plan) (14) I’m a union guy. I support unions. Unions built the middle class. It’s about time they start to get a piece of the action. ( Joe Biden; https: / / millercenter.org/ the-presidency / presidential-speeches/ march-31-2021-announcing-american-jobs-plan) (15) He supports sanctuary cities and now Biden wants to end immigration enforcement and he wants to require you to provide free healthcare for millions and millions of illegal aliens. (Donald Trump; https: / / millercenter.org/ the-presidency/ presidential-sp eeches/ june-20-2020-campaign-rally-tulsa-oklahoma) Example (13) presupposes that the government has been and is currently doing the right thing for the future while (14) presupposes that so far they have not been getting a piece of the action. (15) presupposes that immigration enforcement is going on. Iterative expressions presuppose that something has happened before. Take a look at the examples below. (16) As I recently released the key parts of my pandemic preparedness plan so that America isn’t caught flat-footed when a new pandemic comes again—as it will—next month, I’m also going to release the plan in greater detail ( Joe Biden; https: / / millercente r.org/ the-presidency/ presidential-speeches/ september-9-2021-remarks-fighting-covid -19-pandemic) (17) And it threatens to return to decades of war that ravaged Europe before the international rule-based order was put in place. We cannot go back to that. We cannot. ( Joe Biden; https: / / millercenter.org/ the-presidency/ presidential-speeches/ march-26-2 022-remarks-support-people-ukraine) (18) Tonight, I’m announcing we will increase the average pace of shipment across the country of free monoclonal antibody treatments by another 50 percent. ( Joe Biden; https: / / millercenter.org/ the-presidency/ presidential-speeches/ september-9-202 1-remarks-fighting-covid-19-pandemic) In (16) the adverb again presupposes that a new pandemic has come before at least once. Example (17) through using the verb return instead of just turn suggests that there have at least once been decades of war that ravaged Europe. In example (18) the presupposition trigger is another, indicating that the average pace of shipment has already been increased by some (not necessarily specified) amount at least once. With subordinate clauses introduced by temporal expressions such as after, before or while the content of the subordinate clause is marked as presupposed. (19) After being exposed for telling numerous lies about his identity and resumé, Rep. George Santos joins Stephen Colbert live on The Late Show to truthfully answer the question everyone is asking: who is George Santos? (https: / / www.youtube.com/ w atch? v=KJr7SwQ03pg - Video description) 248 12 The recipient <?page no="250"?> (20) I watched a lot of different elections before they decided to go with this big, massive election with tens of millions of ballots going out to everybody — in many cases, totally unsolicited. (Donald Trump; https: / / millercenter.org/ the-presidency/ president ial-speeches/ november-5-2020-remarks-2020-election) (21) You know, the truth is that we have coronavirus cases now in 46 American states. And while the risk of serious illness of the coronavirus remains low, we want to encourage every American to practice common sense, practice good hygiene, go to the CDC’s website to see what the guidance is for your community or for the American people broadly. (Donald Trump; https: / / millercenter.org/ the-presidency/ presidential-s peeches/ march-13-2020-press-conference-about-coronavirus) In example (19), for instance, the subordinate temporal clause, is marked as presup‐ posed, thereby marking the fact that Santos has been exposed as common knowledge. You may have noticed that the verb to expose also triggers a presupposition, namely that Santos has told numerous lies about his identity and resumé. The subordinate clause in (20) presupposes that the group of people referred to by the personal pronoun they did indeed decide “to go with this big, massive election.” Note how the wording also suggests that this was somehow out of the ordinary even though it was part of the normal procedures, at least as far as I can tell. Similarly, example (21) presents the statement that “the risk of serious illness of the coronavirus remains low” as presupposed. Subordinate clauses in general lend themselves easily to present information as presupposed. Take a look at the examples below. (22) Xi is blatantly spying on America because he does not fear or respect @JoeBiden. (Twitter, @SenRickScott; 03.02.2023) (23) And even though we’re bringing manufacturing jobs back to America, we’re creating more good-paying jobs in education and health care and business services; (h ttps: / / obamawhitehouse.archives.gov/ the-press-office/ 2013/ 12/ 04/ remarks-president -economic-mobility) (24) Fox News host Tucker Carlson explains why permanent Washington is tired of President Biden's leadership on 'Tucker Carlson Tonight.' (https: / / www.youtube.c om/ watch? v=MgXq8S02NJc - Video description) (25) Bill Maher explains how Michael Jackson perfectly reflects our country (https: / / www.youtube.com/ watch? v=efBKINY58Ho - Video description) The mechanism is similar every time. Subordinate clauses usually do not contain the main part of the message. In (22), for instance, the main piece of information is that Xi is spying on the USA, the information provided in the causal subordinate clause is presented as presupposed. Similarly, the concessive subordinate clause in (23) presupposes that the Obama government is bringing manufacturing jobs back to the USA. If someone explains why something is the case, as in example (24), this presupposes that this something is the case, i.e. the content of the subordinate clauses is marked as information that we all agree on. Similarly, if something tells us how 12.2 Communication and the unpacking of information 249 <?page no="251"?> particular grammati‐ cal expres‐ sions just asking a question something can be the case, this something must be the case. Example (25), accordingly, presupposes that Michael Jackson perfectly reflects the USA. Finally, sometimes particular grammatical expressions will mark parts of a message as presupposed. This is often done through structures that put particular elements in focus, thereby backgrounding others and, as a consequence, marking them as some‐ thing speaker and addressee both agree on. In example (26), for instance, this effect is achieved through it-extraposition, which places the focus on the people of Europe, taking the fact that they fought to free themselves for decades for granted. (26) Let’s remember: The hammer blow that brought down the Berlin Wall, the might that lifted the Iron Curtain were not the words of a single leader; it was the people of Europe who, for decades, fought to free themselves. (Joe Biden; https: / / millercenter.org/ th e-presidency/ presidential-speeches/ march-26-2022-remarks-support-people-ukraine) 12.2.3 Don’t think of an elephant! Before you read on … … don’t think of an elephant! The task above is a task that George Lakoff sets all of his first-year students. What is more, he made this the title of one of his books on language and politics (Lakoff 2004). What is so important about this sentence? Well, as you have just experienced, the request not to think about elephants makes you think about elephants. This is what Lakoff wants to emphasise: whenever you negate something, whenever you question something or when you doubt that something is the case, you create a representation of this something in your mind. You cannot help it. Manipulative discourse makes use of this feature of human reasoning in the so-called ‘Just-asking-a-question’ technique (cf. Kreyer in press). The idea is that you can ask questions about anything, but while you ask a question about something you make your addressees create a mental representation of this something. The technique is very efficient when it comes to spreading any kind of (mis-)information, rumours and half-truths. The example below (cf. Kreyer in press) is taken from the Fox News show Tucker Carlson, where the host says the following about a series of possibly COVID-19-related traffic cancellations: (27) We can tell you that the shutdown of Southwest Airlines over the weekend was a direct consequence - it was a reaction to Joe Biden's vaccine mandates. […] At least two Amtrak train routes in the Northeast were cancelled over the weekend, and so is a regularly scheduled car ferry in Washington state […]. Were these also protests against the Biden shot mandates? We can't say for certain. It certainly wouldn't surprise us. (Carlson 2021; 8: 36-9: 14; my emphasis) 250 12 The recipient <?page no="252"?> Questions like the one in bold print are not as innocent as they seem even though Carlson wants us to believe just that: “We're allowed to ask questions and demand answers. That's why we live here. That's your birthright." (Carlson 2021 n. pag.) As Nick Harper from the Minnesota Reformer points out, this technique serves as a Trojan Horse into the mind of the addressee (cf. Kreyer in press): “Just asking questions,” to the uninformed observer, might seem serious, informed, or based in common sense. But in reality, the questioner simulates curiosity in order to plant seeds of doubt in the minds of others. The questioner provides an opportunity for uninformed listeners to be radicalized in pursuit of an ulterior motive. The conversation becomes a Trojan horse for spreading false information, rather than a method for fighting it.-(Harper n. pag.) What is more, the spreading of false information will usually go unnoticed by the recipient. As Harper says, propaganda is “smuggled” into the minds of those that listen and those that ask questions can always maintain that they have never made any claims, as they were just asking questions. This chapter has explored a number of features of communication and message processing that make recipients vulnerable to manipulation through language. The next chapter will take a closer look at the producer and we will shed light on some of the most important kinds of manipulation through language. New tools in your toolkit Humans as rational animals ▶ You are now aware that humans are not as rational in their decision-making as we would assume. This is one feature that makes humans vulnerable to manipulation through language. Cooperative Principle ▶ You have learned that human communication is governed by a general understanding that all interlocutors behave cooperatively. Fortunately, this is usually the case. Unfortunately, this basic assumption of cooperativeness can be exploited for manipulation through language. Whataboutism ▶ You have learned about this frequently used manipulative technique that violates the Maxim of Relevance by steering away from the topic at issue and drawing the recipients’ attention to something else. Inferencing ▶ You are now aware of this general feature of language processing. Recipients will usually learn more from a message than what has been stated explicitly. Inferences are usually of a more subconscious kind and, accordingly, more likely to be accepted as true. Presupposition ▶ You are now aware of the special kind of inferencing which presupposes certain bits of information and portrays them as being understood or agreed on. You know that presuppositions are not likely to be challenged and, accordingly, are a good way to manipulate people into taking something for granted. You are also aware of a large range of lexical items that can serve as presupposition triggers. This will help you to become more aware of presupposed information. 12.2 Communication and the unpacking of information 251 <?page no="253"?> Don’t think of an elephant ▶ You have learned that you cannot not think of something. As soon as something is mentioned, your mind will create a mental representation and with it all the associations that are connected to it. Just-asking-a-question ▶ You have learned about this as a manipulative technique that exploits the fact that you cannot not think of something. You know that it seems innocent but actually serves to spread disinformation and doubt. If you want to know more A R C H E R , Dawn, Karin A I J M E R & Anne W I C H M A N N (2012): Pragmatics. An Advanced Resource Book for Students. London. Routledge. ▶ Introduces key terms and concepts of pragmatics and introduces influential articles on the respective topics and puts them in context. Also provides an extensive companion website. B E A V E R , David I., Bart G E U R T S & Kristie D E N L I N G E R (2021): "Presupposition". In: Edward N. Zalta (ed.): The Stanford Encyclopedia of Philosophy (Spring 2021 Edition). plato.stanford.edu/ ar‐ chives/ spr2021/ entries/ presupposition/ . ▶ Provides a very detailed account of presupposition. Kreyer, Rolf (2020): “Der sprachwissenschaftliche Blick auf Kommunikation - und was Lehr‐ personen davon haben“. In: N. Meister, U. Hericks, R. Kreyer & R. Laging (eds.): Zur Sache. Die Rolle des Faches in der universitären Lehrerbildung. Das Fach im Diskurs zwischen Fachwissenschaft, Fachdidaktik und Bildungswissenschaften. Wiesbaden, 79-100. ▶ Provides a more detailed account of the communication process in the sense of packing, transmitting and unpacking information packages. S E D I V Y , Julie & Greg N. C A R L S O N (2011): Sold on Language. How Advertisers Talk to You and What This Says About You. Hoboken, NJ: Wiley-Blackwell. ▶ Provides many illustrative examples of language use in advertising. Chapter 4 focuses on presupposition. Additional online content. 252 12 The recipient <?page no="254"?> manipula‐ tive 13 The producer The pen is mightier than the sword. (E D W A R D B U L W E R -L Y T T O N ) What this chapter is about In the previous chapter we explored the recipient side of manipulation through language. In terms of the conduit metaphor of communication we found that there are many processes and assumptions that come into play when unpacking the package. This chapter turns to the producer and the packing of the package. The conduit metaphor suggests that there is more or less one way to pack the package, i.e. to put the same thought into language. We will see that this is very far from the truth. The producer has at their disposal many ways in which words and structures can be used to achieve certain effects in the recipient. Before you read on … … take a look at the following campaign slogans. I claim that they use manipulative language. What might be regarded as manipulative here? a. Make America Great Again! (Trump - Presidential Campaign) b. Let’s take back control (Brexit Campaign) c. Remettre la France en Ordre (Marine Le Pen - Presidential Campaign) To answer the question above, of course, we first need to discuss what the term ‘ma‐ nipulative’ should mean in this context. Merriam-Webster’s defines the term as follows: “serving or intended to control or influence others in an artful and often unfair or selfish way”. The Cambridge Dictionary says: “tending to influence or control someone or something to your advantage, often without anyone knowing it”. Both definitions focus on the assumption that manipulative behaviour involves influencing or controlling someone else. What is important in this respect is that the other is usually not aware that they are being influenced. This is, according to van Dijk (2006: 361), what distin‐ guishes “(illegitimate) manipulation [… from] (legitimate) persuasion”. In which sense, then, could we say that language can influence us without us being aware that we are being influenced? One way in which this can happen is that a text leads us to believe something even though we are not fully aware that this is happening. <?page no="255"?> choosing words This may sound almost impossible but, as we will see below, a lot of language processing is subconscious and a lot of language processing goes beyond what is actually being expressed in the message itself (as we have seen in the previous chapter). All of the three examples above, for instance, influence us to accept certain beliefs about the past of America, the UK and France: America once was great, the UK once had control and France once was in order. None of the statements, though, are made explicitly. They hinge on the use either of adverbs or prefixes. To make something great again it must have been great at some point in time before. To take back control somebody must have had control before. To bring something back to order (remettre not mettre) it must have been in order before. All of the three examples above are instances of ‘presupposition’ (see section 12.2). Presupposition works because, as we have seen, humans are ‘trained’ to accept presupposed material as true and we readily do so. This chapter is about further options that the producer has at their disposal to present content in such a way as to manipulate the recipient. 13.1 Packaging information In section 12.2 we have introduced Michael Reddy’s ‘conduit metaphor’: A sender has some thoughts in their mind which they put into a package. This package is sent through a channel (or conduit) to the receiver. After opening the package, the information from the sender’s mind is now also stored in the receiver’s mind. While the previous chapter focused on the recipient, we will now focus on the producer and the packing of the package. Again, the conduit metaphor suggests that there is more or less one way to pack the package, i.e. to put the same thought into language. This chapter wants to demonstrate some of the many ways in which this is not true. 13.1.1 The right word at the right time Typically, we are not quite aware of the huge number of options a person has at their disposal when it comes to describing a person, a scene or an event, because usually we do not search for the right word consciously. In everyday conversations we often use the first words that come to mind when we talk to our partner or our friends. As a student of the English language and after having read the first 12 chapters of this book, you might be a little more aware of the options that a language provides. When it comes to choosing words, Frege’s distinction of extension and intension comes to mind. Please recall section 5.3, where we discussed the example of the ex‐ pressions morning star and evening star. Both expressions mean different things (in the sense of intension), namely ‘the last star in the morning sky’ and ‘the first star in the evening sky’, respectively, but they refer to the same object in the outside world, namely the planet Venus, their extension. Other examples are easy to find: the expression the 254 13 The producer <?page no="256"?> Professor of Modern English Linguistics at the University of Marburg refers to me, Rolf Kreyer, but so does the expression husband to Anna Kreyer. To make reference to Olaf Scholz, you could use the expression der neunte Bundeskanzler der Bundesrepublik Deutschland or you could say Hamburgs Erster Bürgermeister von 2011 bis 2018. In chap‐ ter 5 we explored the idea that words and expressions form huge networks in our minds, which also involve encyclopaedic knowledge. As a consequence, different expressions for the same entity might not necessarily be connected to the same bits of knowledge stored in our mind. If I refer to Scholz as the German Chancellor this might evoke associations different from those that are evoked when I think of him as Hamburg’s First Mayor. The first, for instance, might make us think about the way he handles the war in Ukraine, the second about his possible involvement with the Warburg Bank and their cum-ex-deals. Before you read on … … think about this. In Japan the term kokoro no kaze, meaning ‘cold of the soul’, was introduced into public discourse fairly recently. What do you think this term might refer to? I stumbled upon this term as I was watching the interesting arte documentary Krankeiten nach Maß (you can still find it on youtube, e.g. https: / / youtu.be/ -wOMemp uHu0). The term kokoro no kaze is an alternative term for utsubyō, which is the medical term for depression. The renaming was part of a larger campaign aimed at dealing with the stigma associated with depression in Japanese culture, since Japanese psychiatrists were worried people suffering from depression might not get the treatment they needed because they were too ashamed to seek help. The new term, cold of the soul, was skilfully chosen for several reasons (see Krankheiten nach Maß, 30: 00 - 31: 00): Firstly, colds are very common and everyone has had a cold several times in their lives. So, having a cold of the soul is nothing to be ashamed of. Secondly, Japanese people seem to be used to taking pills even for simple colds. Accordingly, the threshold for taking medication against depression might sink if a depression is understood in terms of a cold. Finally, the soul is a poetic term that touches Japanese people emotionally. All in all, this contributed to a change in the way the people of Japan look at depression. To some extent the whole campaign can be regarded as too successful. More and more people now diagnosed themselves as depressed and demanded medication as a treatment. One psychiatrist who originally supported the whole campaign later claimed: The slogan, depression is like a ‘cold of the soul,’ has convinced far too many people to seek medical treatment for something that is often not an illness. Perhaps we could start saying that depression is like a ‘cancer of the soul.’ That would be more accurate and perhaps not so many people would be willing to adopt that belief. (Watters 2010: 246) 13.1 Packaging information 255 <?page no="257"?> nomination contested nomination The examples above make clear that different names can evoke different associations. This can be exploited to express a particular attitude or a particular perspective through the terms that we use. In this case we speak of ‘nomination’. Consider the name we give to the generation of people born somewhere between the early 1980s and the late 1990s. They are often referred to, neutrally, as Millennials, making reference to the time that they grew up in. The equally neutral alternative term, Generation Y or Gen Y, does not focus on this aspect but distinguishes this group of people more clearly from Boomers, Gen X and Gen Z. Another alternative term, Generation Snowflake, is not neu‐ tral but expresses an unfavourable attitude towards this generation and can be found as an insult in political or public discourse, since it alleges that this generation has an extreme sensitivity, vulnerability and lack of resilience. As another example consider the expression seed money that is frequently used instead of donation in US evangelical circles, particularly those of the prosperity gospel persuasion. It is manipulative for three reasons. Firstly, it suggests that you do not simply donate but that, by donating, you invest in your own future. The money you donate is like a seed that you plant and, like a seed, it will bear fruit far beyond what you have ‘invested’. This is the feel-good side of the metaphor. However, secondly, you cannot expect to harvest fruit if you do not plant a seed. So, the not-so-feel-good side of this metaphor is that it suggests that planting the seed, i.e. donating to the church, is a necessity if you want to benefit. Thirdly, the more you sow the more you can expect to harvest. As a consequence you should try and maximise your donation (even if you do not have a lot of money) since, this way, you will maximise your harvest. Obviously, words have a lot of power. When two or more (political or public) actors used different nominations for the same entity or concept, we speak of ‘contested nomination’. A drastic example is the distinction between ‘war on Ukraine’ as opposed to “special military operation in Uk‐ raine”. Other examples can be found in climate-related discourse. Particularly insight‐ ful, in this context, are the changes that The Guardian made to its style guide. Climate change is no longer considered to accurately reflect the seriousness of the overall situation; use climate emergency or climate crisis instead to describe the broader impact of climate change. […] The OED defines a sceptic as “a seeker of the truth; an inquirer who has not yet arrived at definite conclusions”. Most “climate sceptics”, in the face of overwhelming scientific evidence, deny climate change is happening, or is caused by human activity, so ‘denier’ is more accurate. […] ‘Global heating’ is more scientifically accurate [than ‘global warming’]. Greenhouse gases form an atmospheric blanket that stops the sun’s heat escaping back to space. (www.thegua rdian.com/ environment/ 2019/ oct/ 16/ guardian-language-changes-climate-environment) Contested nomination is highly relevant when it comes to public discourse, as we will see further below. 256 13 The producer <?page no="258"?> transitivity 13.1.2 Beyond the word level In addition to options on the lexical level language also provides us with options on the level of syntax. A thorough exploration of these aspects is provided in Halliday’s (1985) concept of ‘transitivity’. When we describe something that has happened we can distinguish three major elements. The first is the ‘happening’ itself. Halliday calls this the ‘process’. He uses this term fairly generally to, basically, encompass anything that can be described by verbs, e.g. hitting, reading, thinking, being etc. Processes typically involve ‘participants’, e.g. the agent and the patient of an action, and ‘circumstances’, e.g. time, place or manner. Transitivity provides the language user with a large range of options regarding the representation or non-representation of elements relevant to an event and regarding the order in which these elements occur. Take a look at example (5), which provides different headlines relating to the Uvalde mass shooting on the 24 th of May 2022. (5) a) School shooting in Uvalde, Texas (www.cbsnews.com) - b) Texas shooting: How a sunny Uvalde school day ended in bloodshed (www.bbc.co m) - c) Uvalde school shooting ends with 19 children, 2 adults dead (www.texas-tribune.o rg) - d) Uvalde school shooting: At least 21 killed at Robb Elementary (www.ksat.com) - e) 19 children, 2 teachers killed in Texas elementary school shooting (www.nbcnews. com) - f) Uvalde, Texas elementary school shooting: 21 dead, including 19 kids (www.fox5at lanta.com) To fully appreciate the extent to which the different headlines above exploit the system of transitivity it makes sense to start off from a ‘neutral’ description of what has happened. The process referred to is that of someone shooting someone else, resulting in their death. The participants of that process are the shooter and those being shot, the victims. With regard to those being shot we have a number of confirmed deaths and we know how many children and adults died. There are several circumstances accompanying the event: it happened in Uvalde, Texas, it happened at a Robb Elementary School, and it was sunny. Before you read on … … which of the elements above are represented in the six headlines shown in example (5)? What might be the effects resulting from the differences in representation? Table 13.1 below provides an analysis with regard to the (non-)representation of the most relevant elements relevant to the event. 13.1 Packaging information 257 <?page no="259"?> nominalisa‐ tion Element (5a) (5b) (5c) (5d) (5e) (5 f) shooting + + + + + + shooter - - - - - victims - - + + + + number of deaths - - + + + + children and adults - - + - + + Uvalde + + + + - + Texas + + - - + + school + + + + + + elementary school - - - + + + Robb Elementary - - - + - - Table 13.1: (Non-)representation in headlines relating to Uvalde shooting. All headlines mention the process, i.e. the shooting, itself. This does not come as a surprise since this is what the headlines are about. In addition all headlines mention one circumstantial element, namely the fact that the shooting happened at a school. However, only three of the headlines make explicit that the location of the event was in an elementary school and only one headline gives the full identity of that school. With the exception of one headline the town of the shooting is made explicit, four out of the six headlines mention the state of Texas. Interestingly, none of the headlines mention the shooter himself. Headlines a) and b) do not even mention the victims, while the other four make explicit the number of deaths. Three of these four provide additional information as to the number of children and adults killed. Note how the different degrees of specificity and detail can influence the response of the readers. Readers might be expected to react more strongly to a headline which specifies that the location of the incident was not just a school but an elementary school. Readers might also be more shocked if they are provided with details on the number of victims, particularly if they learn that 19 of them were children. Mentioning the state of Texas, or not mentioning it, can lead to a significant difference in readers’ reactions, since Texas belongs to the most liberal states as far as the Second Amendment is concerned. A particularly interesting phenomenon that we find in all of the six headlines is that of ‘nominalisation’ (careful, not to be confused with ‘nomination’ above). The term refers to an instance of language use where we describe a process in terms of an object. In the examples above the process is that of shooting, captured by the verb to shoot. It is a mono-transitive verb, which means that the description of the process, as in 18-year-old Salvador Ramos shoots 21 people, needs at least a subject (18-year-old Sal‐ vador Ramos) and a direct object (21 people). All of the headlines above, however, talk 258 13 The producer <?page no="260"?> passive clause agent deletion about the shooting. What originally was a process represented by a clause has now become an object represented by a noun phrase. Note that the constituents of subject and direct object, which were obligatory in the clause, are no longer obligatory in the noun phrase. It is true, we can add them, for instance in the shooting of 21 people by 18-year-old Salvador Ramos, but grammar does not demand that we do so. This way headlines are a lot freer with regard to the amount of detail that they provide. In some of the above examples the victims of the shooting are mentioned in the headline, but not in all of them. None of the headlines mention the perpetrator. This is only possible through nominalisation. The content of clauses can also be presented in other types of phrases. Politicians like to use the phrase without alternative, or in German alternativlos. Again, underlying these words is a process of someone searching for alternatives but not finding any. If a politician said something to the effect of we have been looking for alternatives but we couldn’t find any, the media and the public might be likely to respond well, you just have to look harder. Contrast this with the statement this is without alternative or das ist alternativlos. The process has now become a feature or an attribute of a thing (this or das). Attributes are not easily changed and, accordingly, we are less likely to blame those that were supposed to present us with alternatives. The examples in (5) also show another way of ‘hiding’ the agent of an action, namely the passive clause. You may recall from section 10.2.2 that a passive clause puts one of the objects of a clause in initial position while deferring the agent to final position in a by-phrase. So, 18-year-old Salvador Ramos kills 21 people becomes 21 people killed by 18-year-old Salvador Ramos. You may also recall that passives allow agent deletion, i.e. the by-phrase becomes optional, which is exactly what we see in examples (5d) and (5e). 13.1.3 Of swarms and floods - the power of metaphor Before you read on … … during the height of the refugee crisis in the summer of 2018 the then British Prime Minister came under heavy criticism (and rightly so) for the following statement on ITVnews: […] You have got a swarm of people coming across the Mediterranean, seeking a better life, wanting to come to Britain because Britain has got jobs, it’s got a growing economy, it’s an incredible place to live […]. What do you think people associate with the term swarm and what do you think might be a possible effect on viewers when they hear migrants and refugees being referred to as a swarm of people? 13.1 Packaging information 259 <?page no="261"?> metaphors in cognitive linguistics The term swarm is usually associated with large number of insects or small animals, such as bees or ants. In figurative language, the term can also be used to describe a large number of persons or animals, particularly when they are in turmoil or move as one mass. In both of these cases, the term implies something that is difficult to control and, hence, scary. Even if used in the context of persons, the insect understanding of the term would usually also be activated in our mental lexicon and consequently any associated negative connotations. A swarm of bees, for instance, the most frequent kind of swarm in the British National corpus (BNC), while unproblematic if looked at from a distance turns into something dangerous and potentially life-threatening if you are surrounded by the swarm. A swarm of locusts, the second-most frequent kind of swarm in the BNC, might eat all your crop, then fly away and leave the land barren and the people starving. Obviously, the metaphorical use of the term swarm can have an impact on the way we see and understand the described situation. The present section will explore this phenomenon in more detail. When we think of metaphor, most likely we think of novels or poems. Already in school we learned that a metaphor is a shortened simile, i.e. instead of saying Achilles is like a lion (which would be a simile) we say the lion Achilles, which is a metaphor. So far so good. Over 40 years ago cognitive linguists George Lakoff and Mark Johnson (1980) suggested a different view on metaphors. Metaphor is for most people a device of the poetic imagination and the rhetorical flourish - a matter of extraordinary rather than ordinary language. Moreover, metaphor is typically viewed as characteristic of language alone, a matter of words rather than thought or action. For this reason, most people think they can get along perfectly well without metaphor. We have found, on the contrary, that metaphor is pervasive in everyday life, not just in language but in thought and action. Our ordinary conceptual system, in terms of which we both think and act, is fundamentally metaphorical in nature. (1980: 1) The above quote, taken from the beginning of the seminal book Metaphors we Live by, makes two interesting claims. Firstly, metaphor occurs in everyday normal language, not just in poetic language. Secondly, metaphor is not just a matter of words but also a matter of thoughts and actions. The way we understand our world, the way we interact with our world, the way we think is based on metaphor. Both claims may seem bold at first sight but they make a lot of sense if we take a moment to think about them. Example (6) shows a handful of everyday expressions that you might hear at your workplace, in the library or at any restaurant. (6) a) Lions are at the top of the food chain. - b) He fell from grace. - c) I’m under a lot of pressure. - d) I’m glad I’ve left that behind. - e) We have to go our separate ways. (6a and b) illustrate that our understanding of (social) power and hierarchies is based on the concepts ‘top versus bottom’ or ‘up versus down’. Example (6a), for instance, 260 13 The producer <?page no="262"?> structural metaphor target and source domain mapping does not literally mean that somehow lions sleep and stay in elevated places or that they are taller than their prey. Of course, this is obvious, but I’m stressing this point so much because a sentence like (6a) is so common and so much a part of our everyday language that we find it hard to believe that there is a metaphor underlying it. (6b) is similar. If someone falls from grace they lose someone’s favour or they have to leave a position of power and honour. None of this has anything to do with an actual downward movement. It is only through our metaphorical conceptualisation of power in terms of up and down that the verb to fall makes sense here. When you are under a lot of pressure (example (6c)), you express that the amount of pressure you are feeling right now to some extent has gained control over your life. Again, being in control is conceptualised as being up or on top, whereas being controlled by somebody or by something relates to being below this somebody or something, which is why you are under pressure. Example (6d) illustrates two kinds of metaphors. First, we have a metaphor which understands past and future in terms of back and front. This is in part based on a second metaphor, namely an understanding of our personal life as a journey. If we move in a forward direction, the future is in front of us while things of the past are behind us. Another example of the journey metaphor is (6e). The expression to go our separate ways is used to describe the end of a relationship between two people, such as a romantic relationship or a professional relationship. The relationship is conceptualised as a journey together, the end of the relationship is conceptualised as parting ways. The metaphor L OV E I S A J O U R N E Y is a prime example of what Lakoff and Johnson call a ‘structural metaphor’. In structural metaphors “one concept is metaphorically struc‐ tured in terms of another” (Lakoff and Johnson 1980: 14). The first concept is called the ‘target domain’. The target domain typically is abstract and not easy to comprehend, as, for example, love or a romantic relationship. The ‘source domain’, in the above case the journey, helps us to understand the nature of love/ romantic relationships. This is done by taking our idea of and experiences with journeys and relating the different aspects of a journey to different aspects of a love relationship. This is what is meant by metaphorically structured. The structure of the source domain helps us understand the target domain. Lakoff and Johnson say that structural metaphors are based on mappings between source domain and target domain. Figure 13.1 shows some map‐ pings between the source domain J O U R N E Y and the target domain L OV E in the metaphor L OV E I S A J O U R N E Y (cf. Kövecses 2002: 7). Fig. 13.1: Mappings of the conceptual metaphor L O V E I S A J O U R N E Y . 13.1 Packaging information 261 <?page no="263"?> The lovers in a romantic relationship are the travelers on a journey together and their relationship is the vehicle in which they travel. They have common goals which are conceptualised as the common destination of the journey and they will encounter difficulties in reaching these goals, these are likened to impediments on the road. Of course, there are other possible mappings, for instance that of major decisions in a relationship and crossroads on a journey, but I hope the general idea of structural metaphors has become clear. Before you read on … … think, again, about David Cameron’s expression swarm of people when referring to refugees. To many people this might trigger an association with locusts. What would be the effect of such an association? What would a mapping for the metaphor R E F U G E E S A R E L O C U S T S look like? As described above, a conceptual metaphor uses a source domain to understand and conceptualise a target domain. The source domain in this case is L O C U S T S . First, we have to identify elements of the source domain, i.e. what comes to mind when we think of a swarm of locusts. Maybe the first thing is that such a swarm consists of an incredibly large number, in which the individual insect is no longer perceived. Secondly, as single animals locusts are completely harmless. It is only when they gather together in large swarms that they pose a problem and a threat. Thirdly, a swarm will migrate and invade neighbouring countries. Fourthly, as a swarm the locusts will travel across the lands of a given country and eat all the crops they find. Fifthly, when they move on, they leave nothing behind but barren fields, possibly causing severe food crises and even famines. Finally, humans are more or less helpless against locust swarms, leaving them only with the option to try and drive them away from their fields as much as they can. Having identified the pertinent elements of the source domain, in a second step we have to find out how the metaphor maps these six elements to the target domain, i.e. R E F U G E E S . I am very much aware of the extent to which these mappings are to be despised but I have chosen to comment on this metaphor since it comes from authentic political discourse (of course, we do not know whether Cameron had this effect in mind) and since a discussion of this metaphor is particularly instructive because it is so contemptible. When the metaphor conceptualises refugees as a swarm of locusts, one main effect is that refugees become de-individualised. No longer are they considered as individual persons that have their own individual stories, their own individual sufferings and hopes, but they are regarded as a faceless mass. Secondly, while people might feel 262 13 The producer <?page no="264"?> sympathetic towards refugees as individuals, they start perceiving them as a problem or a threat if their numbers increase beyond some (subconscious) threshold. Thirdly, in conceptualising refugees as locusts the metaphor suggests that refugees enter a country irrespective of whether they have been invited or not and maybe even against the will of this country’s inhabitants. It suggests further that once inside a country, refugees will not contribute to the well-being of this country but they will only take from the resources of that country and, fifthly, if there is nothing more to get, if they have ‘sucked the social systems dry’, they will move on. Food crises and famines would be mapped onto an alleged intolerable strain on our health and social systems, which might cause these to break down. Finally, since the inhabitants are basically helpless once a swarm of locusts is in their country, the metaphor suggests that the only option is to keep refugees out of a country. The mapping between the two domains looks as shown in figure 13.2. Fig. 13.2: Mappings of the conceptual metaphor R E F U G E E S A R E L O C U S T S . The question, of course, is to what extent that is really relevant. You might say “even if a phrase like swarm of people triggers associations with locusts, I still see the individuals in the refugees, and I am aware of their problems and I am aware that it is our responsibility to help”, and I would say “yes, you are right, and you might be able to do that, but you do it against or despite of the interpretation that the metaphor suggests. The image of a swarm of locusts is activated in your mind. You cannot help it.” This point, in my view, is very important. It’s the same point that Lakoff wants to make with his task “Don’t think of an elephant! ” You cannot not think of something. This goes to show that once a concept is activated in your mind, it is activated, you cannot help it, and the same is true for images that are activated through metaphors. Of course, the use of particular metaphors does not determine the way we think about something, but metaphors might provide us with something like ‘conceptual guardrails’. That is, they direct our thinking process in particular tracks, often without us being aware that this is happening, which makes it even more difficult for us to actively conceptualise people, events or situations in a different way. With regard to the locust metaphor above, for instance, the metaphor itself does not leave any room for dealing with a swarm of locusts other than hoping that it will not invade our country. This suggests a particular response to any refugee situation. Even if you, like me, think that a metaphor of this 13.1 Packaging information 263 <?page no="265"?> 2 This section is largely based on Kreyer (in press). kind is absolutely despicable, the images it conjures up will become activated. It is important that we are aware of this, as it gives us a much better understanding of how this particular use of manipulative language works. Before you read on … … take a look at the water-related metaphors below. Water-related metaphors are very frequent in the context of public discourses related to immigration and refugees. What are possible effects of the metaphorical expressions shown below? What, in particular, is the difference between floods and tsunamis? a) Flood of Refugees From Ukraine War Tests Europe’s Capacity to Welcome Them (The Wall Street Journal) b) Mass Migrant Crossing Floods Texas Border Facilities (The New York Times) c) Tsunami of refugees in Europe triggered by US policy in Middle East (TASS) d) Syria: UN cites utter desperation behind ‘tsunami’ of refugees into Europe (UN News) The metaphor, by presenting refugees and migrants as floods and tsunamis, suggests that immigration and refugees are a threat. Like the swarm metaphor from above, the metaphors in a) to d) de-individiualise refugees. Refugees are seen like an amorphous mass, just like a body of water. As a consequence, there is no room for the suffering and the tragedies of the individual human beings. Both, floods and tsunamis lead to major destruction. The difference is mainly in the abruptness of the two events. While floods usually build up and end slowly and steadily, tsunamis hit a coastal area suddenly and with little prior warning. This leaves the inhabitants completely defenceless in the case of tsunamis. With slow-building floods you try your best to keep the water out of your home, e.g. by using sandbags as a barricade. Elisabeth Wehling (2016, n. pag.), one of George Lakoff students, puts it very succinctly: Build dams, put sandbags in front of your doors, pump out your cellar and your groundfloor. […] In case of a flood, it would be absurd to distribute incoming water to the different rooms of your home or to try and find out how much water would fit into each room. (my translation) It is obvious what this metaphor suggests when it comes to how we should deal with immigration and refugees. I think it is important to be aware of this. 13.1.4 Framing 2 Much of what we have discussed above can be regarded as instances of framing. Framing is a process where we choose our words in a way that make our listeners look at what we say through a particular pair of glasses. A simple example is looking at a 264 13 The producer <?page no="266"?> glass half empty or half full. Another example was given above: framing immigration as a flood conceptualises immigrants as a threat. The concept of framing has been advocated forcefully by George Lakoff who contributed a lot to our understanding of language and political and public discourse. One of his claims is that liberals in the US have fallen into the “[t]rap of Enlightenment Reason” (2010: 72). He writes: It is a common folk theory of progressives that ‘The facts will set you free! ’ If only you can get all the facts out there in the public eye, then every rational person will reach the right conclusion. It is a vain hope. Human brains just don't work that way. Framing matters. (2003: n. pag.) He defines frames as “mental structures that shape the way we see the world” (2004: xv). According to him, they correlate with neural structures in our brain. A particular expression can activate a particular frame, which means a particular part of our brain is now active. We have already seen above that human decision-making is not always as rational as we think it is. Lakoff goes a lot further when he claims: The old view claimed that reason is conscious, unemotional, logical, abstract, universal, and imagined concepts and language as able to fit the world directly. All of that is false. Real reason is: mostly unconscious (98 %); requires emotion; uses the ‘‘logic’’ of frames, metaphors, and narratives; is physical (in brain circuitry); and varies considerably, as frames vary. (2010: 72) This seems to leave little room for facts and we do actually see examples of this happening in political discourse. The retired Admiral William H. McRaven, in an opinion piece in the Washington Post, accuses the then President Donald Trump of undermining the institutions of the USA. (Then) Fox News’ host Chris Wallace (W) confronts Trump (T) with these accusations and the following exchange takes place. (7) W: Bill McRaven, retired Admiral, Navy SEAL, 37 years, former head of US special operations - T: Hillary Clinton fan - W: … special operations … - T: Excuse me, Hillary Clinton fan - W: … who led the operations, commanded the operations that took down Saddam Hussein, and that killed Osama bin Laden, says that your sentiment is the greatest threat to democracy in his life time. - T: Okay, he’s a Hillary Clinton backer, erm, and an Obama backer … and … frankly … - W: … He’s a Navy SEAL … - T: wouldn’t it have been nice if we got Osama bin Laden a lot sooner than that … (njnewsfeed 21: 08-21: 40) As you can see, Trump does not enter into a rational debate about McRaven’s accusations. Instead he forcefully tries to establish a frame against which the accusation 13.1 Packaging information 265 <?page no="267"?> should be viewed: they are not to be taken seriously since they come from a “Hillary Clinton fan” and a Clinton and Obama “backer”. Contrary to Lakoff ’s assessment that liberals fall victim to the idea of Enlightenment Reason, the following example shows that they, too, understand the power of framing. Trump has repeatedly stressed the importance and the necessity of a wall on the border to Mexico. (8) a) […] we need a WALL! In 2018, 1.7 million pounds of narcotics seized, 17,000 adults arrested with criminal records, and 6000 gang members, including MS-13, apprehended. A big Human Trafficking problem. (Donald Trump, Twitter, 01/ 05/ 2019) - b) The damage done to our Country from a badly broken Border---Drugs, Crime and so much that is bad---is far greater than a Shutdown […] (Donald Trump, Twitter, 01/ 13/ 2019) - c) […] without a Wall it all doesn’t work. Our Country has a chance to greatly reduce Crime, Human Trafficking, Gangs and Drugs. […] (Donald Trump, Twitter, 01/ 24/ 2019) Enlightenment Reason would address these claims through facts, i.e. how many narcotics are actually passing the border, what is the exact number of gang members, how extensive is human trafficking, and so on. Kamala Harris, the US Vice President at the time of writing, on various occasions, followed a different approach when she frames the whole issue as Trump’s vanity project: (9) […] there’s plenty of empirical evidence that that [sic] there are statements that are being made that are just simply not the truth and are frankly […] propaganda and we have to call it what it is. […] We have enough problems, one doesn’t need to create a problem, one does not need to create a problem. This issue is about a vanity project for this president. (ABCtheView, 2019: 0: 55-1: 26) (10) […] the President’s vanity project (CNN, 2019a: 1: 08-1: 09) (11) […] this President’s medieval vanity project called a wall (CNN, 2019b 4: 16-4: 20) Again, the effect is that a rational debate about the arguments is not necessary since, after all, Trump’s wall only serves his vanity. Before you read on … … think about a frequently cited example by Lakoff: tax relief. He argues, and I think rightfully so, that the term in itself frames taxes in a negative way. Why is that? The answer to this question lies in the noun relief. What are things you are relieved of ? If you don’t know, you can use your corpus-linguistic skills to find out: A search in the BNCweb for R1-R5 noun collocates of the verb to relieve yields mostly negative things, such as pressure, tension, pain, burden, anxiety, stress, suffering and so on. Because of the frequent co-occurrence of the verb with negative things, the verb and its derived noun obtains a negative semantic prosody (see section 5.3.4). Anything that occurs with the 266 13 The producer <?page no="268"?> verb or the noun as tax in tax relief will, by association, also be viewed as something negative. Taxes, therefore, will be regarded as negative. To make things worse, even if you argued against tax cuts, e.g. by saying Tax relief is wrong! , you would still strengthen the view that taxes are bad. So, what can we do? Lakoff (2004b: n.pag.) claims that we need alternative frames to counter existing frames. Taxes, he suggests, should be seen as a membership fee: Taxes are what you pay to live in America - to have democracy, opportunity, government services, and the vast infrastructure build by previous taxpayers - the highways, the internet, the schools, scientific research, the court system etc. Taxes are membership fees used to maintain and expand services and the infrastructure. Unfortunately, though, establishing frames takes time (Lakoff 2010: 73). It is not “a short-term messaging issue” (74). Once frames are established, on the other hand, they are very powerful, as Lakoff (73) makes clear. Have you ever wondered why conservatives can communicate easily in a few words, while liberals take paragraphs? The reason is that conservatives have spent decades, day after day building up frames in people’s brains, and building a better communication system to get their ideas out in public. The power of frames has been demonstrated quite shockingly in the January 6 th insurrection. The whole idea of stolen elections was prepared for more than one and a half year, as Spring (2020: n.pag.) shows. Already in April (2020) Trump tweeted: “GET RID OF BALLOT HARVESTING, IT IS RAMPANT WITH FRAUD. THE USA MUST HAVE VOTER I.D., THE ONLY WAY TO GET AN HONEST COUNT! ”. However, even this claim needed fertile ground to strike root and grow. This ground, in my view, is provided by the more general idea of the ‘deep state’, which in Trump-supporting circles is understood as “organized resistance within the government, working to subvert his presidency”. (Abramson 2017: n. pag.) With this idea deeply rooted in the minds of Trump supporters it seems only natural that behind the scenes everything would be done to steal the election from Trump. 13.1.5 (Faulty) logic and argumentation Logic plays an essential role when it comes to argumentation. Through logical reasoning we try to move from known information, the premises, to new information, the conclusion. Consider the syllogism, a pattern of logical reasoning that consists of two premises and a conclusion that can be derived from these. A classic example is the following: (12) Premise 1: All men are mortal. - Premise 2: Socrates is a man. - Conclusion: Socrates is mortal. 13.1 Packaging information 267 <?page no="269"?> incorrect premises Given the two premises the conclusion must be true. This might become a little clearer if we rephrase (12) as follows. (13) Premise 1: If somebody is a man, he is mortal. - Premise 2: Socrates is a man. - Conclusion: Socrates is mortal. Ideally, all argumentation is logically sound and factually correct. Unfortunately, though, this is not always the case. What is more, fallacies, i.e. errors in reasoning, sometimes go unnoticed and may thus lead to wrong conclusions. In some cases, this might be because of bad reasoning on the part of the speaker, in other cases, it might be done deliberately to deceive the receiver. Whatever the reason, it is useful to be aware of logical fallacies as they occur frequently in public discourse. As a first example let us consider the syllogism again. Before you read on … … let’s assume the following (hypothetical) syllogism: - Premise 1: If you pay Kreyer 100 Euros, he will let you pass. - Premise 2: I paid Kreyer 100 Euros. - Conclusion: He will let me pass. Let us further assume that even though you have paid me 100 Euros I have still failed you. What’s wrong? The syllogism is an example of rigorous logical reasoning. As such it is only interested in deriving new information from the premises. The problem is we can derive new information from incorrect premises. The reasoning in the task above is logically sound, and the conclusion is logically valid. However, at the same time the conclusion is fac‐ tually incorrect since at least one of the premises is wrong: I would not let you pass if you were trying to bribe me. A syllogism of the above kind is not easy to challenge. After all, the main point is the conclusion, which is logically sound. The premises are of a rather backgrounded nature. Often this would be something that we agree on. Things obviously get tricky when premises are presented as common ground, even though they are debatable. A good example is the following statement by right-wing commentator Matt Walsh titled ‘Feminism dies at the dinner table’: (14) If you don’t need no man, as you suggest, then I sure hope that you don’t expect men to pay for your meals when you go out on a date. I hope when that check comes, your reaching for it every single time. But, as we know, classic thing, we’ve … I I have never heard of a feminist who remains a feminist when the check hits the table. (https: / / www.youtube.com/ shorts/ J8TnMvAkL50) 268 13 The producer <?page no="270"?> ad hominem The underlying structure (in this case a three-premise argument) would be something like this. (15) Premise 1: If a woman is a feminist, she does not need a man. - Premise 2: If a woman does not need a man, she pays for her meals everytime she goes out with a man. - Premise 3: You do not pay for your meals everytime you go out with a man. - Conclusion: You are not a feminist. Again, the conclusion is logically valid. Premises 1 and 2 are of the kind ‘If A is the case, then B is the case as well’. A premise of this kind is called an ‘implication’ and can be reversed, i.e. ‘If B is not the case, then A cannot be the case either’. This is called the ‘contraposition’. For example, if it rains, the streets get wet. Let us assume you go outside for a walk and all the streets are dry, then you know that it has not rained. Figure 13.3 shows the implications of (15) together with their contrapositions: Fig. 13.3: Implication and contraposition. Walsh’s claim rests on the contrapositions on the right-hand side of figure 13.3. If you insert premise 3 “You do not pay for your meals everytime you go out with a man” at the bottom of the right-hand side, you get the following chain of implication: You do not pay for your meals → you need a man → you are not a feminist. Again, the conclusion is fully valid. However, the premises and, therefore, the contrapositions are debatable: With regard to premise 2, there may be many reasons beyond needing a man why a woman would let a man pay for her meal. Similarly, I know many women who would describe themselves as feminists but would not agree to premise 1 (depending, of course, on what is meant by to need a man). The ad hominem fallacy is a rhetorical ploy that is aimed at attacking the person presenting an argument, not the argument itself (ad hominem translates to to the per‐ son). An example was given in the previous section in example (7), here repeated as (16) for convenience. (16) W: Bill McRaven, retired Admiral, Navy SEAL, 37 years, former head of US special operations - T: Hillary Clinton fan - W: … special operations … - T: Excuse me, Hillary Clinton fan 13.1 Packaging information 269 <?page no="271"?> appeal to authority W: … who led the operations, commanded the operations that took down Saddam Hussein, and that killed Osama bin Laden, says that your sentiment is the greatest threat to democracy in his life time. - T: Okay, he’s a Hillary Clinton backer, erm, and an Obama backer … and … frankly … - W: … He’s a Navy SEAL … - T: wouldn’t it have been nice if we got Osama bin Laden a lot sooner than that … (njnewsfeed 21: 08-21: 40) When confronted with a series of allegations by retired Admiral Bill McRaven, Trump did not attempt to refute these allegations but simply dismissed them by emphasizing that McRaven was a Hillary Clinton and Obama backer. Having said that, ad hominems can be justified. This depends on whether the quality of the speaker that is under attack is relevant or not. We could maybe argue that Trump had a point: if McRaven really was a Clinton and Obama fan, it might make sense to assume that he does not have a neutral view of Trump’s policies. Still, we need to be aware even in those cases that the argument itself has not been refuted, we have not even considered that it might be correct. As another example, consider the present situation. You are reading a book written by a professor of English linguistics, who argues that a thorough training in linguistics can be useful for your teaching career. It would be easy for you to dismiss everything you have read so far through ad hominem reasoning: He has never worked as a language teacher, how is he supposed to know what is useful for me as a future teacher? You would be right. The fact that I have never worked as a teacher is definitely relevant when it comes to deciding whether I am the right person to suggest what future teachers should do. However, what is also relevant is the fact that I have talked to many teachers that made clear to me that they find linguistics highly relevant for their jobs and that former students of mine told me how helpful linguistics is for them. Finally, even though I have never worked as a language teacher I still know a little bit about language and language learning. So, after all I might not be completely unqualified to write this book. This goes to show that it is not always easy to decide whether an ad hominem is justified or not. Still, when you discard something on an adhominem basis, be aware that you may just have ignored a really interesting and valid point only because the wrong person brought it forward. So, whenever you hear something along the lines of S/ he says X but s/ he is just a student/ a professor/ a feminist/ divorced/ an activist/ a social justice warrior/ an old white man/ … this does not automatically prove that the X is wrong. The opposite of the ad hominem fallacy is the argument-from-authority or ap‐ peal-to-authority fallacy. In example (7)/ (16) above Chris Wallace appeals to the au‐ thority of McRaven when he says: (17) Bill McRaven, retired Admiral, Navy SEAL, 37 years, former head of US special operations, who led the operations, commanded the operations that took down Saddam Hussein, and that killed Osama bin Laden, says that your sentiment is the greatest threat to democracy in his life time. 270 13 The producer <?page no="272"?> bandwagon Again, whether an appeal to authority is fallacious or not depends on the authority themself and whether their qualities are relevant with regard to a given argument. In the above example an appeal to authority might be justified since McRaven’s career makes him an authority when it comes to threats to democracy. Similarly, I might be considered an authority regarding the relevance of linguistics for language teaching. Still, this is not the same as evidence for the claim or claims that are put forward. A case in point is the claim that the sun revolves around the earth. As is well known, the then authority, the Roman Catholic Church and the Inquisition, based on a literal reading of the Bible, put Galileo Galilei on trial for his claims in 1633, sentenced him to house arrest and forced him to stop teaching or defending heliocentrism. It took the Roman Catholic Church more than 450 years to rehabilitate Galileo Galilei (in 1992). So, authority never is an appropriate substitute for rational argumentation. Related to the appeal to authority is the ‘bandwagon argument’. The name comes from the expression to jump on the bandwagon, which refers to joining a movement or an activity because it has become fashionable or popular. Another, self-explanatory, term is ‘appeal to popularity’, i.e. something must be right because so many people think it is right. Most likely you have used the bandwagon argument as a child. For instance, “But I need a new phone. Everyone in my school has a new phone! ! ! ” or “I want to stay up till 12. All of my friends stay up that long! ! ! ” Obviously, the logic behind this argument is far from compelling as your parents might have made clear to you from early on. I have heard the classic “And would you jump off a bridge if all of your friends did? ” many, many times and have always found it difficult to argue against. Indeed, it demonstrates the fallaciousness of an appeal to popularity quite convincingly. Still, the bandwagon argument is often used in advertisements or political discourse. Example (18) is taken from the Nicer Dicer Plus commercial. (18) Millions of the original nicer dicer have been sold around the world and now the very best is even better. (www.youtube.com/ watch? v=931Yg6GYT3w; 5: 32-5: 37) While the example above should be self-explanatory, the next example might be a little less obvious. Still, I think it illustrates an appeal to popularity, albeit a little more indirectly. The example is taken from Donald Trump’s speech on the 6 th of January, 2021, before the riot on the Capitol. He says: (19) I’ve been in two elections. I won them both and the second one, I won much bigger than the first. OK. Almost 75 million people voted for our campaign, the most of any incumbent president by far in the history of our country, 12 million more people than four years ago. (https: / / www.npr.org/ 2021/ 02/ 10/ 966396848/ read-trum ps-jan-6-speech-a-key-part-of-impeachment-trial) Being even more popular than four years ago, Trump has even more of a right to be the President - at least, that seems to be a plausible argument for him. The fallacy of this argument becomes evident when we take into consideration that Joe Biden won 81 million votes. 13.1 Packaging information 271 <?page no="273"?> false dicho‐ tomy straw man In political discourse we frequently find what is called ‘false dichotomy’ or ‘false dilemma’. A false dichotomy presents a situation such that only two choices are pos‐ sible. Take a look at the following examples. (20) Every nation, in every region, now has a decision to make. Either you are with us, or you are with the terrorists. (Bush; https: / / millercenter.org/ the-presidency/ presidential-speeches/ september-22-20 01-address-us-response-attacks-september-11) (21) What choices are we gonna make to reach that goal? Either we ask the wealthiest Americans to pay their fair share in taxes or we’re going to have to ask seniors to pay more for medicare. (Obama; https: / / www.youtube.com/ watch? v=4QepFGJj74o; 4: 54 - 5: 00) Example (20) is taken from George W. Bush’s Address on the U.S. Response to the Attacks of September 11 from the 22 nd of September 2001. He reduces possible reactions to the 9/ 11 attacks to two options only. Neutrality, for instance, is not an option. What is more, if a nation does not exactly follow the USA, this means you automatically side with the terrorists. In example (21) Obama talks about how to finance medicare. He reduces a possibly fairly large range of options to only two. The straw man fallacy, too, hinges on incorrect representation, this time an incorrect representation of the original argument. Often it involves gross overgeneralisation, simplification or exaggeration. As a consequence the misrepresented argument is much easier to attack. The straw man fallacy is very commonly used and you might have experienced it yourself. Here are a couple of examples from my own experience: (22) A: We should really try to reduce our carbon dioxide emissions. - B: Well, but we cannot all go back to living in caves! (23) A: Germany, as the economically most powerful European country, has a particular responsibility when it comes to refugees. - B: Well, but we cannot take all of them! In both cases B attacks an argument that A has never made. Bill Maher illustrates the straw man argument like this: (24) If we are, for example, discussing how technology has made our lives easier and I say “The smartphone is a wondrous device” and you say “So, you support toddlers working in sweatshops”, that’s a straw man argument […]. (www.youtube.com/ watch? v=lTFnj-9EY4M; 1: 17-1: 33) It is a clever strategy for two reasons. The misrepresentation and distortion of the original claim links your opponent to a possibly outrageous claim while at the same time making you the winner of the whole argument. It is a very frequently used technique that we see in social media but also in other news outlets. A particularly instructive example, in my view, is Jordan Peterson’s interview with Channel 4’s Cathy Newman (https: / / www.youtube.com/ watch? v=aMcjxSThD54). The whole interview (a transcript can be found here: https: / / scrapsfromtheloft.com/ psychology/ jordan-peter 272 13 The producer <?page no="274"?> sons-channel-4-interview-cathy-newman-transcript/ ) is 30 minutes long and shows multiple uses of the straw-man strategy by Newman. (25) JP: Women want deeply men who are competent and powerful. And I don’t mean power in that they can exert tyrannical control over others. That’s not power. That’s just corruption. Power is competence. And why in the world would you not want a competent partner? Well, I know why, actually, you can’t dominate a competent partner. So if you want domination … - CN: So you’re saying women want to dominate, is that what you’re saying? - JP: No, I’d say women who have had their relationships impaired with, impaired, their relationships with men impaired and who are afraid of such relationships will settle for a weak partner because they can dominate them. But it’s a suboptimal solution. In (25) Peterson makes a nuanced statement about competence and domination in partnerships, which Newman incorrectly sums up as “women want to dominate men”. Similar examples can be found all over the interview. Typically, Newman introduces them by something along the lines of “So, you’re saying …” to which Peterson would typically reply “no”, “I’m not saying that …” or something to that effect. Here is another example, this time about the gender pay gap in the UK. (26) CN: That 9 percent pay gap, that’s a gap between median hourly earnings between men and women. That exists. - JP: Yes. But there’s multiple reasons for that. One of them is gender, but that’s not the only reason. If you’re a social scientist worth your salt, you never do a univariate analysis. You say women in aggregate are paid less than men. Okay. Well then we break it down by age; we break it down by occupation; we break it down by interest; we break it down by personality. - CN: But you’re saying, basically, it doesn’t matter if women aren’t getting to the top, because that’s what is skewing that gender pay gap, isn’t it? You’re saying that’s just a fact of life, women aren’t necessarily going to get to the top. - JP: No, I’m not saying it doesn’t matter, either. - CN: You’re saying that it’s a fact of life… - JP: I’m saying there are multiple reasons for it. - CN: Yeah, but why should women put up with those reasons? - JP: I’m not saying that they should put up with it! As a final example consider the following exchange: (27) CN: Okay, so rather than denying that the pay gap exists, which is what you did at the beginning of this conversation, shouldn’t you say to women, rather than being agreeable and not asking for a pay raise, go ask for a pay raise. Make yourself disagreeable with your boss. - JP: Oh, definitely, there is that. But I didn’t deny it existed, I denied that it existed because of gender. See, because I’m very, very, very careful with my words. 13.1 Packaging information 273 <?page no="275"?> slippery slope CN: So the pay gap exists. You accept that. I mean the pay gap between men and women exists—but you’re saying it’s not because of gender, it’s because women are too agreeable to ask for pay raises. - JP: That’s one of the reasons. - CN: Okay, one of the reasons. So why not get them to ask for a pay raise? Wouldn’t that be fairer? - JP: - I’ve done that many, many, many times in my career. […] So one of the things you do as a clinical psychologist is assertiveness training. […] So I’ve had many, many women, extraordinarily competent women, in my clinical and consulting practice, and we’ve put together strategies for their career development that involved continual pushing, competing, for higher wages - and often tripled their wages within a five-year period.- - CN: And you celebrate that? - JP: Of course! Of course! The examples, I think, speak for themselves. More examples can be found in the interview. It is highly instructive to take a closer look and we can learn a lot from it. The slippery-slope fallacy also involves a distortion of reality. It involves a chain of causation aptly described in Dowden (2020: 286) as follows. (28) A leads to B. B leads to C. C leads to D. D leads to … - … which leads to HELL. - We don’t want to go to HELL. - So, we should not take that first step A. Many children, I assume, have heard slippery slope argumentation when debating with their parents whether to do their homework. It goes something like this: (29) If you don’t do your homework, your grades will get worse. If your grades get worse, you will fail this class. If you fail this class, you will not graduate. If you don’t graduate, you cannot go to university. If you can’t go to university, you will not get a job. If you do not get a job, you will have no income. If you don’t have income, you will end up homeless. If you are homeless, you will die on the streets. You don’t want to die on the streets, do you? So, get your homework done! The argument works because people are generally not really good with probabilities, as Dowden (2020: 286) points out. If each of the individual implications had a likelihood of 80 % (i.e. If you do not do your homework, there is an 80 % probability that your grades will get worse), this does not mean that the implication ‘If you don’t do your homework, you will die on the streets’ has the same probability. Quite the opposite: with a chain of eight implications the probability is actually 0.8 8 , which is roughly 0.17. Slippery slope argumentation is frequently found in discussions concerning major decisions about important areas of life, such as immigration, the legalisation of marihuana or same sex marriage. The following example is taken from a report on gay marriage in Australia (www.youtube.com/ watch? v=R4HhpMXdKM4), where an anti-gay-marriage protester says the following. 274 13 The producer <?page no="276"?> (30) The right should extend to gay marriage, so why shouldn’t it extend to polygamous relationships, why shouldn’t it extend to bestiality, where should it end? (0: 43 - 0: 50) As you can see, the slippery slope is presented in terms of a sequence of necessary consequences: If we allow gay marriage, there is no reason not to allow polygamous marriages. This hidden claim, of course, is not true. There may be many reasons against polygamy in addition to those that some people put forward against gay marriage. Similarly, moving from polygamy to bestiality may involve additional questions and concerns that are not pertinent to polygamy. As pointed out above, the relevant question when it comes to challenging a slippery slope argument is to question the validity of each of the implications of the argument. How likely is it that if we allow gay marriage, we will also have to allow polygamy? How likely is it that if we allow polygamy, we will also have to allow bestiality? 13.2 Walking uphill In the previous chapter we have seen how susceptible humans are to manipulation through language. This chapter has shown a wide range of ways in which the producer can package information to influence the way their recipients interpret the message. Language can be manipulative and it is used to manipulate. Much of this manipulation happens without us being aware of it because it often exploits features of language processing and communication that make our languages the excellent tools that they are. What is more, in the age of social media, the amount of messages that we have to process every day has increased tremendously. As a consequence the time we take to process all of this input will be reduced drastically. Take as an example Trump’s tweet about the “Fake News Mainstream Media[’s]” reluctance to report on alleged “record setting economic news” from one of the tasks in section 12.2.2. We have seen how much effort it will take to disentangle this tweet, and identify and (hopefully) challenge all of the presuppositions in that tweet. Who would be willing to invest all the time needed to do this with hundreds of messages that need to be processed? The odds are definitely not in the favour of those who want to keep manipulation through language at bay. The situation, in my view, is similar to the situation that Aza Raskin, former employee of Firefox and Mozilla Labs, co-founder of the Center for Humane Technology and inventor of Infinite Scroll, describes in the excellent film The Social Dilemma: You can imagine these things are sort of like… they… they tilt the floor of… of human behavior. They make some behavior harder and some easier. And you’re always free to walk up the hill, but fewer people do, and so, at scale, at society’s scale, you really are just tilting the floor and changing what billions of people think and do. (Aza Raskin, n. pag.) Challenging manipulative language can be done, but it takes conscious effort. The good news - it gets easier with practice. But before it can be practiced, people have to be 13.2 Walking uphill 275 <?page no="277"?> aware of the mechanisms and techniques explored in this and the previous chapter. This is why I have included them in this book and this is where you come into play: as future language teachers you will have the opportunity to teach hundreds of students of yours about the power of language. You can make hundreds of young adults aware of manipulation through language and you can make hundreds of young adults a little less susceptible to manipulation. I think it’s worth a try. New tools in your toolkit (Contested) nomination ▶ A new understanding of the semantic terms intension and extension has made it clear(er) to you how many options people have at their disposal to describe one and the same thing, including the associations that come with each description. You have also learned that this can be used to convey a particular attitude. Transitivity ▶ With regard to the description of events you are now more aware of the many options that the English language provides when it comes to (non-)representation of aspects related to said event. Metaphor ▶ You have learned that metaphor is pervasive in everyday language and that it influences the way we think and act. In particular, metaphors suggest ways in which we look at and react to something. Framing ▶ The more general idea of framing draws your attention to the fact that language can be used to portray things in such a way as to suggest a particular interpretation together with a particular response. Logical fallacies ▶ You now know about some of the most important logical fallacies that are frequently applied in public discourse. As a consequence you are a little less vulnerable to these rhetorical ploys and you can help your students to become a little less vulnerable. If you want to know more D O W D E N , Bradley H. (2020): Logical Reasoning. Availiable under: https: / / www.csus.edu/ indiv/ d/ dowdenb/ 4/ logical-reasoning-archives/ Logical-Reasoning-2020-05-15.pdf ▶ A comprehensive textbook (almost 600 pages) on logical reasoning with chapters on defending against deception and detecting fallacies. F L O W E R D E W , John & John E. R I C H A R D S O N (eds.) (2020). The Routledge Handbook of Critical Discourse Studies. London: Routledge. ▶ Provides a comprehensive overview of critical discourse studies including different theoretical and methodological approaches as well as applications to a wide range of social questions and media. K R E Y E R , Rolf (in press): “Make political discourse rational again. Language and Polarization”. ▶ This paper, on which parts of this chapter are based, provides many other examples of metaphors and framing in US American political discourse. 276 13 The producer <?page no="278"?> L A K O F F , George & Mark J O H N S O N (1981). Metaphors We Live By. Chicago: The University of Chicago Press. ▶ The classic on conceptual metaphor theory. L A K O F F , George (2004). Don’t Think of an Elephant. Know Your Values and Frame the Debate. Hartford, VT: Chelsea Green Publishing. ▶ A very accessible introduction to Framing Theory. W O D A K , Ruth & Bernhard F O R C H T N E R (eds.) (2018): The Routledge Handbook of Language and Politics. London: Routledge. ▶ Provides a comprehensive overview of theoretical and methodological approaches to language and politics as well as many illustrative case studies. Additional online content. 13.2 Walking uphill 277 <?page no="280"?> References A A R T S , Bas & Sean W A L L I S (eds.) (2006): The British Component of the International Corpus of English. [Release 2. CD-Rom]. London: Survey of English Usage. ABCT H E V I E W (2019): “Sen. Kamala Harris Calls President Trump's Proposed Border Wall a 'Vanity Project'.” The View, YouTube, 8 Jan. 2019, https: / / www.youtube.com/ watch? v=P-cEU 8MlP-U. accessed 28.02.2022. A B R A M S O N , Alana (2017): “President Trump's Allies Keep Talking About the 'Deep State.' What's That? ” TIME. time.com/ 4692178/ donald-trump-deep-state-breitbart-barack-obama/ . Accessed February 28, 2022. A I T C H I S O N , Jean ( 3 2003): Words in the Mind. Malden, MA: Blackwell Publishing. A N D E R S O N , Richard C., Ralph E. R E Y N O L D S , Diane L. S C H A L L E R T & Ernest T. G O E T Z (1977): “Frameworks for comprehending discourse”. In: American Educational Research Journal 14, 367-381. A N T H O N Y , Laurence (2022): AntConc (Version 4.2.0) [Computer Software]. Tokyo, Japan: Waseda University. Available from https: / / www.laurenceanthony.net/ software. A R I S T O T L E . Politics. A Treatise on Government. Transl. by William Ellis. www.gutenberg.org/ file s/ 6762/ 6762-h/ 6762-h.htm. Accessed February 28, 2022 B A L L , Catherine N. (1994): “Automated Text Analysis: Cautionary Tales”. In: Literary and Linguistic Computing 9(4). 295-302. B A R L O W , Rich (2014): “BU Research: A Riddle Reveals Depth of Gender Bias”, https: / / www.bu.e du/ articles/ 2014/ bu-research-riddle-reveals-the-depth-of-gender-bias/ , accessed 12.01.2023. B E H A G H E L , Otto (1932): Deutsche Syntax. Eine geschichtliche Darstellung. Band IV. Wortstellung. Periodenbau. Heidelberg: Carl Winters Universitätsbuchhandlung. B E H A G H E L , Otto (1909): "Beziehungen zwischen Umfang und Reihenfolge von Satzgliedern". In: Indogermanische Forschungen 25, 110-142. B E R N A D I N I , Silvia (2002): “Exploring new directions for discovery learning”. In: Bernard Kette‐ mann & Georg Marko (eds.) Teaching and Learning by Doing Corpus Analysis. Proceedings of the Fourth International Conference on Teaching and Language Corpora, Graz 19-24 July, 2000. Amsterdam: Rodopi. 165-182. B E R N A R D I N I , Silvia (2000): “Systemising serendipity. Proposals for concordancing large corpora with language learners”. In: Lou Burnard & Tony McEnery (eds.): Rethinking Language Pedagogy from a Corpus Perspective. Frankfurt a. M.: Peter Lang. 225-234. B I B E R , Douglas & Susan C O N R A D ( 2 2019): Register, Genre, and Style. Cambridge. CUP. BNC C O N S O R T I U M (2001): The British National Corpus, version 2 (BNC World). Distributed by Oxford University Computing Services on behalf of the BNC Consortium. URL: http: / / www .natcorp.ox.ac.uk/ . B O L I N G E R , Dwight (1968): Aspects of Language. New York: Harcourt, Brace, and World. <?page no="281"?> B R A N S F O R D , John D. & Marcia K. J O H N S O N (1972): “Contextual prerequisites for understanding: some investigations of comprehension and recall”. In: Journal of Verbal Learning and Verbal Behaviour 11, 717-726. B R O W N , Penelope and Stephen C. L E V I N S O N (1987): Politeness. Some Universals in Language Usage. Cambridge: Cambridge University Press. C A R L S O N , Tucker (2021): “We’re allowed to ask questions and demand answers”, Fox News, 07/ 20/ 2021. https: / / www.foxnews.com/ opinion/ tucker-carlson-vaccine-questions-demand-a nswers. Accessed 28.02.2022. C A R T E R , Ronald & Michael M C C A R T H Y (2006): Cambridge Grammar of English. A Comprehensive Guide. Spoken and Written Grammar and Usage. Cambridge: Cambridge University Press. C L A R K , Eve V. & Herbert H. C L A R K (1977): Psychology and Language. An Introduction to Psycholinguistics. New York: Harcourt, Brace, Jovanovich. C O L L I N S , A.M., & Elizabeth F. L O F T U S (1975): “A spreading-activation theory of semantic processing”. In: Psychological Review 82, 407-428. C O L L I N S , Allan M. & M.Ross Q U I L L IA N (1969): “Retrieval time from semantic memory”. In: Journal of Verbal Learning and Verbal Behavior 8, 240-247. C O R D E R , Stephen Pit (ed.) (1981): Error Analysis and Interlanguage. Oxford: Oxford University Press. C H A F E , Wallace L. (1982): "Integration and involvement in speaking, writing, and oral literature". In: Deborah Tannen (ed.) Spoken and Written Language. Exploring Orality and Literacy. Norwood, NJ: Ablex. 35-53. C H O M S K Y , Noam (1965): Aspects of the Theory of Syntax. Cambridge, MA: MIT Press. C R U S E , D. Alan (1986): Lexical Semantics. Cambridge: CUP. D A N E Š , František (1974): "Functional sentence perspective and the organization of the text". In: František Daneš (ed.) Papers on Functional Sentence Perspective. The Hague: Mouton. 106-128. D A V I E S , Mark (2017): Corpus of US Supreme Court Opinions. Available online at https: / / www.en glish-corpora.org/ scotus/ D A V I E S , Mark (2011-): Corpus of American Soap Operas. Available online at https: / / www.english -corpora.org/ soap/ . D A V I E S , Mark (2008-): The Corpus of Contemporary American English (COCA). Available online at https: / / www.english-corpora.org/ coca/ . D A V I E S , Mark (2004): British National Corpus (from Oxford University Press). Available online at https: / / www.english-corpora.org/ bnc/ . D E B E A U G R A N D E , Robert-Alain & Wolfgang U. D R E S S L E R (1981): Introduction to Text Linguistics. London: Longman. D I J K , Teun A. van (2006): “Discourse and Manipulation”. In: Discourse & Society 17, 359-383. D O W D E N , Bradley H. (2020): Logical Reasoning. Availiable under: https: / / www.csus.edu/ indiv/ d/ dowdenb/ 4/ logical-reasoning-archives/ Logical-Reasoning-2020-05-15.pdf E D G E , Julian (1988): “Applying linguistics in English language teacher training for speakers of other languages”. In: ELT Journal 42(1), 9-13. F I R T H , John Rupert (1957; 1968): “A synopsis of linguistic theory, 1930 - 55”. In: F.R. Palmer (ed.): Selected Papers of J. R. Firth 1952 - 59. London: Longmans, Green and Co, 168-205. 280 References <?page no="282"?> F R E G E , Gottlob (1892; 4 2022): “Über Sinn und Bedeutung”, In: Zeitschrift für Philosophie und philosophische Kritik 100, 25-50 [reprint in: Karel Berka & Lothar Kreiser (eds.): Logik-Texte. Berlin-: Akademie-Verlag, 423-442] G A B E L E N T Z , Hans Georg Conow v. d. (1901; 1972): Die Sprachwissenschaft, ihre Aufgaben, Meth‐ oden und Bisherigen Ergebnisse. Tübingen: TBL. G A R N E S , Sara and Zinny S. B O N D (1976): “The relationship between semantic expectation and acoustic information”. In: Phonologica 3, 285-293. G ÖT Z , Sandra (2013): Fluency in Native And Nonnative English Speech. Amsterdam: John Benja‐ mins. G R I C E , H. Paul (1975): “Logic and conversation”. In: Peter Cole & Jerry L. Morgan (eds.) Syntax and Semantics, Vol. 3: Speech Acts. New York: Academic Press. 41-58. H A L L I D A Y , Michael A.K. (1989): Spoken and Written Language. Oxford: OUP. H A L L I D A Y , Michael A.K. & Ruqaiya H A S A N (1976): Cohesion in English. London: Longman H A R P E R , Nick (2020): “No, you’re not ‘just asking questions.’ You’re spreading disinfor‐ mation.“ Minnesota Reformer. minnesotareformer.com/ 2020/ 12/ 17/ no-youre-not-just-ask‐ ing-questions-youre-spreading-disinformation/ / . Accessed 28.02.2022. H A W K I N S , John A. (1994): A Performance Theory of Order and Constituency. Cambridge: Cam‐ bridge University Press. H I L L , Jimmie (2000): “Revising priorities. From Grammatical Failure to Collocational Success”. In: Michael Lewis (ed.): Teaching Collocation: Further Development in the Lexical Approach. Oxford: Oxford University Press: 47 -69. H O E Y , M. (2005): Lexical Priming. A New Theory of Words and Language. London: Routledge. H O F F MA N N , Sebastian, Stefan E V E R T , Nicholas S M I T H , David Y.W. L E E and Ylva B E R G L U N D P R Y T Z (2008): Corpus Linguistics with BNCweb---a Practical Guide. Frankfurt am Main: Peter Lang. H O F F M A N N , Sebastian & Stefan E V E R T (2006): “BNCweb (CQP-edition): the marriage of two corpus tools”. In: Sabine Braun, Kurt Kohn und Joybrato Mukherjee (eds.): Corpus Technology and Language Pedagogy. New Resources, New Tools, New Methods. Frankfurt a. M.: Peter Lang. 177-195. J A M E S , Carl (1998): Errors in Language, Learning and Use. Exploring Error Analysis. London: Routledge. J O H N S O N , Marcia K., John D. B R A N S F O R D & Susan K. S O L O M O N (1973): “Memory for tacit implica‐ tions of sentences”. In: Journal of Experimental Psychology 98(1), 203-205. J O H N S , Tim (2002): “Data-driven learning. The perpetual challenge”. In: Bernard Kettemann & Georg Marko (eds.) Teaching and Learning by Doing Corpus Analysis. Proceedings of the Fourth International Conference on Teaching and Language Corpora, Graz 19-24 July, 2000. Amsterdam: Rodopi. 105-117. J O H N S , Tim (1991): “Should you be persuaded: Two examples of data-driven learning materials”. In: English Language Research Journal 4, 1-16. K ÖV E C S E S , Zoltán (2002): Metaphor. A Practical Introduction. Oxford: Oxford University Press. K R A S H E N , Stephen D. (1981): Second Language Acquisition and Second Language Learning. Oxford: Pergamon. References 281 <?page no="283"?> K R E Y E R , Rolf (in press): “Make Political Discourse Rational Again. Language and Polarization”. In: W. Zacharasiewicz, C. Birkle & M. Prisching (eds.): Polarization in the North Atlantic Triangle. Wien: Verlag der ÖAW. K R E Y E R , Rolf (2010a): “How much wheat is there in the chaff ? Issues concerning the use of concordancers in the classroom”. In: Language Forum 36, 155-173. K R E Y E R , Rolf (2010b): Introduction to English Syntax. Frankfurt a.M.: Peter Lang. K R E Y E R , Rolf (2004): Inversion in Modern Written English. Syntactic Complexity, Information Status and the Creative Writer. Tübingen: Gunter Narr. K R E Y E R , Rolf & Barbara G ÜL D E N R I N G (2016): “‘But sometimes this potential is not used at all’. Views on linguistics in EFL teacher training and teaching”. In: R. Kreyer, S. Schaub & B. Güldenring (eds.): Angewandte Linguistik in Schule und Hochschule. Neue Wege für Sprachunterricht und Ausbildung. Frankfurt a.-M., 265-297. L A K O F F , George (2010): “Why it Matters How We Frame the Environment.” Environmental Communication 4, 70-81. L A K O F F , George (2004a): Don’t Think of an Elephant. Know Your Values and Frame the Debate. Hartford, VT: Chelsea Green Publishing, 2004. L A K O F F , George (2004b): What’s your law“; edge.org/ response-detail/ 10696. Accessed 17.01.2023. L A K O F F , George (2003): “Metaphor and War, Again.” UC Berkeley. escholarship.org/ uc/ item/ 32b962zb/ . Accessed February 28, 2022. L A K O F F , George & Mark J O H N S O N (1980): Metaphors we Live by. Chicago: University of Chicago Press. L A N G E R , Ellen, Arthur E. B L A N K & Benzion C H A N O W I T Z (1978): ”The mindlessness of ostensibly thoughtful action: The role of ‘placebic’ information in interpersonal interaction”. In: Journal of Personality and Social Psychology 36, 635-42. L E N N O N , Paul (1991): “Error: Some problems of definition, identification, and distinction”. In: Applied Linguistics, 12(2), 180-196. L E W I S , Michael (1993): The Lexical Approach. The State of ELT and a Way Forward. Hove: Language Teaching Publications. L I P K A , Leonard (2002): English Lexicology. Lexical Structure, Word Semantics & Word Formation. Tübingen: Gunter Narr. L O F T U S , Elizabeth (1975): “Leading questions and the eyewitness report”. In: Cognitive Psychology 7, 560-72. L O F T U S , Elizabeth & John C. P A L M E R (1974): “Reconstruction of automobile destruction. An example of the interaction between language and memory”, Journal of Verbal Learning and Verbal Behaviour 13, 585-589. L O V E , Robbie, Claire D E M B R Y , Andrew H A R D I E , Vaclav B R E Z I N A & Tony M C E N E R Y (2017): “The Spoken BNC2014. Designing and building a spoken corpus of everyday conversations”. In: International Journal of Corpus Linguistics 22(3). 319-344. L Y O N S , John (1977): Semantics, Vol 1. Cambridge: CUP. M A N N , William C. & Sandra A. T H O M P S O N (1988): “Rhetorical Structure Theory: Toward a functional theory of text organization”. Text 8(3), 243-281. 282 References <?page no="284"?> M C R A V E N , William H. (2020): “Opinion: Trump is actively working to undermine the Postal Serv‐ ice.” The Washington Post, 08/ 16/ 2020, www.washingtonpost.com/ opinions/ the-countrys-futu re-could-hinge-on-postal-workers/ 2020/ 08/ 16/ c0f7b97e-dfca-11ea-8dd2-d07812bf00f7_story. html/ . Accessed 28.02.2022. M O R R I S , Richard S. (1999): Behind the Oval Office. Getting Reelected against all Odds. St. Martin’s Press. Folkestone: Renaissance Books. N A T T I N G E R , James R. & Jeanette S. D E C A R R I C O (1992): Lexical Phrases in Language Teaching. Oxford: Oxford University Press. P A W L E Y , Andrew & Frances Hodgetts S Y D E R (1983): “Two puzzles for linguistic theory. Nativelike selection and nativelike fluency”. In: Jack C. Richards & Richard W. Schmidt (eds.): Language and Communication. London: Longman: 191-226. R E D D Y , Michael J. (1979): “The conduit metaphor: A case of frame conflict in our language about language”. In: Andrew Ortony (ed.) Metaphor and Thought. Cambridge: Cambridge University Press. 284-310. R O S C H , Eleanor & Carolyn B. M E R V I S (1975): “Family resemblances: Studies in the internal structure of categories”. In: Cognitive Psychology 7(4), 573-605. R O S E N B E R G , Paul (2014): “This is why conservatives win: George Lakoff explains the importance of framing -and what Democrats need to learn”, Salon, www.salon.com/ 2014/ 11/ 22/ this_is_ why_conservatives_win_george_lakoff_explains_the_importance_of_framing_and_what_d emocrats_need_to_learn/ , accessed 12/ 30/ 2022. S E D I V Y , Julie & Greg N. C A R L S O N (2011): Sold on Language: How Advertisers Talk to You and What This Says About You. Hoboken, NJ: Wiley-Blackwell. S E L I N K E R , Larry (1972): “Interlanguage”. In: IRAL - International Review of Applied Linguistics in Language Teaching 10(3), 209-231. S I N C L AI R , John (1996): “The search for units of meaning”. In: Textus 9, 75-106. S I N C L AI R , John (1991): Corpus, Concordance, Collocation. Oxford: Oxford University Press. S I N C L AI R , John (1983): “Naturalness in Language”. In: Jan Aarts & Willem Meys (eds.): Corpus Linguistics. Recent Developments in the Use of Computer Corpora in English Language Research. 203-210. S P R I N G , Marianna (2020): “BBC News: ’Stop the steal’: The deep roots of Trump’s ‘voter fraud’ strategy.” www.bbc.com/ news/ blogs-trending-55009950. Accessed February 28, 2022. S T A N D F I E L D , Robert A. & Rolf A. Z W AA N (2001): “The effect of implied orientation derived from verbal context on picture recognition”. In: Psychological Science 12(2), 153-156. S T E I N D L , Christina, Eva J O N A S , Sandra S I T T E N T H A L E R , Eva T R A U T -M A T T A U S C H & Jeff G R E E N B E R G (2015): “Understanding psychological reactance. New developments and findings“, Zeitschrift für Psychologie 223, 205-214. T V E R S K Y , Amos & Daniel K A H N E M A N N (1981): “The Framing of Decisions and the Psychology of Choice”, Science 211, 453-458. W A T T E R S , Ethan (2010): Crazy Like us. The Globalization of the American Psyche. New York: Free Press. References 283 <?page no="285"?> W E H L I N G , Elisabeth (2016): “Die verkehrte Sprache. Deutschland redet sich eine solidarische Flüchtlingspolitik aus: Die Debatte bestimmt von Sprachbildern, die keine Empathie für Flüchtende zulassen”, Spiegel Online. Accessed 08.01.2023. W E I L , Henri (1844; 1978): The Order of Words in the Ancient Languages Compared with that of the Modern Languages. Amsterdam: John Benjamins. W I L K I N S , David A. (1972): Linguistics in Language Teaching. Cambridge, MA: MIT Press. W R A Y , Alison (2002): Formulaic Language and the Lexicon. Cambridge: Cambridge University Press. 284 References <?page no="286"?> Glossary * character. A → wild card that substitutes for zero or more characters. Accordingly, a search for ‘*man’ will yield all occurrences of man, woman, chairman, Roman etc. ? character. A → wild card that substitutes for one single character. Accordingly, a search for ‘characteri? e’ will yield all occurrences of characterise and characterize. | character. A character used to indicate alternatives in a → corpus query, e.g. ‘man|woman’ would identify all occurrences of the string man and the string woman. Acceptability. A standard of textuality. It focuses on the attitude of the text receiver, who should view the text as having → cohesion and → coherence and be in some way meaningful to him or her. Acceptability also demands that the text receiver’s effort to decode the text should be reasonable. Activation spreading. An important component in → network models that describes the fact that a fully active node in a network will contribute to the activation of its neighbouring nodes. Ad hominem argument. A rhetorical strategy that does not attack the argument but attacks the person presenting the argument. Allophone. A specific variant of a phoneme, such as the ‘clear’ and ‘dark’ l’s as two allophones of / l/ . Allophones either occur in → free variation or in → complementary distribution. Analytic language. A language that expresses grammatical relations through word order. English is a (mostly) analytic language. Compare → synthetic language. Antonymy. A → sense relation that describes oppositeness of meaning. Different kinds of antonymy distinguish between → binary or → complementary a., → gradable a., → directional a. and → relational a. Appeal-to-authority argument. A rhetorical strategy that does not substantiate the argument by facts but by reference to the authority of the person presenting the argument. Assimilation. A phonological process where the articulation of a sound is influenced by a neighbouring sound, making the first more similar to the latter. In regressive assimilation one sound influences the pronunciation of a preceding sound (e.g. handbag is pronounced as [hæmbæg]), in progressive assimilation one sound influences the pronunciation of a following sound (e.g. would you is pronounced as [wʊdʒuː]) . Bald on record. A subtype of the → on-record strategy for doing → face-threatening acts where the speaker, in contrast to → positive politeness and → negative politeness strategies, does not make use of any redressive action to mitigate the face threat. Bandwagon argument. A rhetorical strategy that does not substantiate the argument by facts but by reference to the fact that a large number of people is of the same opinion. Binary antonymy. A → sense relation that describes a special kind of → antonymy where only two states are possible, e.g. dead and alive. Also called ‘complementary antonymy’. Citation form. In contrast to → weak form the fully stressed realisation of a word for example the use of the word and in “the word and has three letters” as opposed to its use in rock’n’roll. <?page no="287"?> Coherence. A standard of textuality that demands that the elements of the textual world are connected in meaningful ways (within the rules of the world created by the text). Cohesion. A standard of textuality that has to do with how elements of the text surface are connected to one another. According to Halliday and Hasan (1976: 11) cohesion occurs when “the interpretation of any item in the discourse requires making reference to some other item in the discourse […].” Colligation. The typical co-occurrence of a word with a grammatical category or construction, e.g. say + NP + PP as opposed to tell + NP + NP. Collocation. The typical co-occurrence of a word with another word, e.g. handsome man. Competence. A term coined by Noam Chomsky to describe the knowledge that an ideal speaker-listener, who knows their language perfectly, has of this language. Contrast with → performance. Complementary antonymy. A → sense relation that describes a special kind of → antonymy where only two states are possible, e.g. dead and alive. Also called binary antonymy. Complementary distribution. A feature of allophones that are not in → free variation but mutually exclusive, i.e. they never occur in the same environment. The → phoneme / l/ , for example, has two allophones, clear ‘l’ ([l]) and dark ‘l’ (ɫ). The former occurs before vowels or [j], the latter occurs before a consonant or at the end of a syllable. Conceptual metaphor. A concept introduced by George Lakoff and Mark Johnson to describe the phenomenon that humans understand one idea (the target domain) in terms of another idea (the source domain), e.g. a romantic relationship in terms of a journey. In contrast to the traditional conception of metaphor in literary studies Lakoff and Johnson emphasise that conceptual metaphors are everywhere in language and that they influence the way we think and act. Concordance. A basic research tool in → corpus linguistics. A concordance provides all occurrences of a search term together with its context in a given corpus. Conduit metaphor. A metaphor introduced by Michael Reddy (1979) that conceptualises communication as a process similar to the sending of a package. Ideas are packed into language objects (encoded in words and sentences), these are sent through a channel to a receiver, who unpacks (decodes) the language objects and now has the sender’s ideas in her or his mind. Consonant. A sound that, in contrast to → vowels, is produced by obstructing the airstream. Constant theme. An instance of → thematic progression where the → rheme of one clause becomes the → theme of a number of following clauses. Contraposition. The reversal of an implication which involves the negation of the individual terms and the reversal of their order. If the implication is ‘If A then B’, the contraposition is ‘If not B then not A’. Conventionalised indirectness. → Off-record utterances that have become conventionalised as forms of politeness. For example, the question “Can you pass me the salt? ” is conventionally understood as a polite request to pass the salt. Conversational implicature. Additional information that can be derived from an utterance if the speaker → flouts one of the maxims of the → Cooperative Principle. 286 Glossary <?page no="288"?> Cooperative principle. A principle suggested by Herbert Paul Grice to describe the generally cooperative nature of communication. Cooperativeness is understood in terms of four maxims relating to the areas → quality, → quantity, → relation and → manner. Corpora. The plural of → corpus. Corpus linguistics. A branch of linguistics that relies on → corpora as their data source and often involves statistical analysis. Corpus. A large collection of (typically computerised) authentic texts that are used for linguistic analysis. Cue validity. The degree to which a given feature serves a cue for a category. For instance, the feature ‘has feathers’ has a high cue validity for the category ‘bird’. Data-driven learning. A type of inductive learning advocated by Tim Johns where language learners discover language rules by analysing → corpus data. Compare → discovery learning. Defining feature. An attribute that is used to describe an entity. Typically a defining feature has a high → cue validity. Derivational morphology. The branch of → morphology that is concerned with the creation of a new word from a given one, e.g. the morpheme {-er} can be used to derive (create) a new word from verbs. The new word, in this case, describes the doer of the action that is described by the verb, e.g. a singer ({sing} + {-er}) is someone who sings. Contrast with → inflectional morphology. Directional antonymy. A → sense relation that describes a special kind of → antonymy that concerns opposite directions of movement or orientation, e.g. come and go or up and down. Discovery learning. A type of inductive learning advocated by Silvia Bernardinin where language learners discover language rules by browsing a → corpus. Compare → data-driven learning. Encyclopaedic relations. A broad cover term for any kind of background knowledge that is connected to a unit in the mental lexicon, e.g. the fact that Penguin is also the name of a biscuit and a publishing house. Endocentric phrase. A type of phrase that in contrast to → exocentric phrases can be reduced to one word only. This word is called the ‘head’ of the phrase. An example of an endocentric phrase is the noun phrase. For example, the NP the old man from downstairs can be reduced to the noun man. Entailment. A form of logical deduction such that the truth of one sentence requires the truth of another sentence. Entailments, in contrast to → presuppositions, do not survive negation. The sentence The King of France was murdered entails The King of France is dead. This entailment is not necessarily true anymore if the King of France was not murdered. Error. 1) A deviation from an assumed target norm that is due to lack of knowledge of a language. 2) After Lennon (1991: 182): “a linguistic form or combination of forms which in the same context and under similar conditions of production would, in all likelihood, not be produced by the speakers’ native speakers counterparts.” Errors have to be distinguished from → mistakes. Glossary 287 <?page no="289"?> Exocentric phrase. A type of phrase that in contrast to → endocentric phrases cannot be reduced to one word only. In the English language the prepositional phrase is an exocentric phrase. It always needs a preposition and the prepositional complement and cannot be reduced to the preposition only. Extension. The description of the meaning of a term by listing all the elements that fall under that term. Face. A concept used by Brown and Levinson to describe the public self-image that every member of a community wants to claim for himor herself. See → negative face; see → positive face. Face-threatening act. Any speech act (or other) that threatens the addressee’s → positive or → negative face. False dichotomy argument. A rhetorical strategy that reduces a complex situation to the choice between only two alternatives. Final devoicing. An articulatory process in German where voiced obstruents (i.e. / b, d, g, v, z, ʒ, dʒ/ ) are devoiced when they occur in syllable-final position. Flouting. Blatantly not following a maxim of the → Cooperative Principle so that the addressee is aware that the speaker does not follow the maxim. Flouting leads to → conversational implicature. Fragmentation. A concept suggested by Wallace Chafe to describe a feature of spoken language, where usually information units (or ideas) are presented one at a time. Compare → integration. Frame. According to Lakoff (204: xv), “mental structures that shape the way we see the world.” Frames have neural correlates in our brain. Framing. The use of language in such a way that particular → frames are activated during processing. Free variation. A feature of allophones that are interchangeable. The voiceless alveolar plosive [t] and the glottal stop [Ɂ] are in free variation in some accents of English. The phoneme / t/ as in water, for instance, can be realised in one of the two allophones in accents spoken in the London area. See → complementary distribution. Given-before-new principle. A principle of text processing that describes a tendency of sentences to start with something known or given and to proceed to something unknown or new. Gradable antonymy. A → sense relation that describes a special kind of → antonymy that allows in-between states, e.g. freezing and hot. Holonymy. A → sense relation in which one word is the whole to another word that is the part/ a member, e.g. body and head. Compare → meronymy. Homography. A relation between words that are written the same but pronounced differently, e.g. read (present tense) and read (past tense). Compare → homophony and → homonymy. Homonymy. A relation between words in which two or more words are pronounced and written the same, e.g. bat (the animal) and bat (for baseball). Not to be confused with → polysemy. 288 Glossary <?page no="290"?> Homophony. A relation between words that are pronounced the same but written differently, e.g. right and write. Compare → homography and → homonymy. Hyper(o)nymy. A → sense relation in which one word (the hyper(o)nym) is superordinate to another word (the hyponym), e.g. mammal and dog. Compare → hyponymy. Hypertheme. An instance of → thematic progression where all clauses of a text are elabora‐ tions of a common (not necessarily explicit) theme. Hyponymy. A → sense relation in which one word (the hyponym) is subordinate to another word (the hyper(o)nym), e.g. dog and mammal. Compare → hyper(o)nymy. Idiom principle. Describes the fact that a “language user has available to him or her a large number of semi-preconstructed phrases that constitute single choices, even though they might appear analyzable into segments.” (Sinclair 1991: 110) Understood in complementation to → the open-choice principle. Incidental learning. A term that refers to the acquisition of knowledge about an item even though the learner did not intent to learn about that item. Similar to → serendipitous learning. Inference. Information derived from a message that is not explicitly stated in the message itself. Inflectional morphology. The branch of → morphology that is concerned with the creation of a word form of a given word, e.g. the morpheme {-s} can be used to create the plural form cats of the singular word cat. Contrast with → derivational morphology. Informativity. A standard of textuality that makes reference to the degree to which the information provided in the text is known/ expected vs. new/ unexpected. Input. The totality of language utterances encountered by a learner. It is contrasted with → intake. Institutionalised expression. A special kind of → lexical phrase which, like a → polyword, cannot be varied, usually is continuous but longer and constitutes a whole utterance, e.g. How do you do? Intake. The subset of → input which becomes part of the language competence of a learner. Integration. A concept suggested by Wallace Chafe to describe a feature of written language, where several information units (or ideas) are represented in more syntactically complex structures, for example, complex noun phrases. Compare → fragmentation. Intension. The description of the meaning of a term by listing all its defining features. Also called → sense. Intentionality. A standard of textuality that demands that the text producer intends their text to be cohesive and coherent and to fulfil some goal with regard to the receiver. Interlanguage. A term introduced by Larry Selinker to denote the language system of a foreign language learner, which contains features, both of the learner’s first language and of the target language. An interlanguage follows its own rule system and is not some random mix of native and target language. Intertextuality. A standard of textuality that describes potential relations that a text has to already existing texts as well as the fact that every text is usually regarded as an instance of a particular text type. Intrusive ‘r’. The insertion of an r-sound between two words even though there is no ‘r’ in writing, e.g. law and order, [lɔː ɹǝn dɔːdǝ]. Glossary 289 <?page no="291"?> Just-asking-a-question. A technique where the speaker uses a seemingly innocent question to spread misinformation, rumours, doubts and half-truths. KWIC. The acronym stands for key word in context and refers to the most typical representation of → concordance data with the search term in one column at the centre of the output and the context to the left and right of this column. Lemma. The set of all forms that a word (as you would find it in a dictionary) can take. The lemma of the verb to go, for instance, consists of the five (written) forms go, goes, went, going and gone. Lemmatization. A process whereby all word forms of one word that we can find in a → corpus are assigned to their → lemma. Lexical approach. An approach to language teaching advocated by Michael Lewis. It is driven by the idea that language consists of grammaticalised lexis and not lexicalised grammar, i.e. lexis is primary and should also be primary in language teaching. Lexical phrase. The concept was introduced by Nattinger and DeCarrico (1992) and describes patterns of language use of varying sizes and degrees of variability that fulfil specifiable communicative functions. Lexicalised sentence stem. The concept was introduced by Pawley and Syder (1983) and describes a larger partly variable language pattern that constitutes the basis of a complete sentence and is stored as a single unit of form and meaning in the → mental lexicon. Lexico-Grammar. An approach which in contrast to the traditional strict separation of → vocabulary and grammar suggests a continuum between lexis and grammar with the two endpoints interacting with and influencing each other. Lexis. The store of lexical items, including individual words but also larger units. In the → Lexical Approach it is contrasted with the idea of → vocabulary, which is traditionally seen as a store of mere words and their meaning. Liaison. One form of → linking words in pronunciation by inserting a linking sound between two words. → linking ‘r’, → intrusive ‘r’ Linear progression. An instance of → thematic progression where the rheme of the previous clause becomes the theme of the current clause. Linguistics. The scientific study of language in general or specific languages. Linking ‘r’. The pronunciation of a word-final letter ‘r’ in non-rhotic accent of English if the next word begins in a vowel, e.g. far away, [fɑː ɹ əweɪ] as opposed to far, [fɑː] Manipulation. In contrast to → persuasion, manipulation involves influencing or controlling the addressees without them being aware that they are being influenced or controlled. Manner of articulation. Describes the way in which the airstream is obstructed during the articulation of → consonants. Manner, Maxim of. A maxim of the → Cooperative Principle that demands that the speaker’s contribution is easy to understand, i.e. it avoids overly complex vocabulary and syntax. Mental lexicon. The representation of our knowledge of words and other lexical units in a speaker’s mind. Meronymy. A → sense relation in which one word is the part/ member of another word that is the whole, e.g. head and body. Compare → holonymy. 290 Glossary <?page no="292"?> Minimal pair. Two words that differ in one sound only. The minimal-pair test is used to identify the phonemes of a language. Mistake. A deviation from an assumed target norm that, in contrast to → errors, does not result from a lack of knowledge of a language but from language-external factors, such as sleepiness or lack of concentration. Morphology. The scientific study of the internal structure of word forms. Word-form struc‐ ture is described by breaking words down into their smallest meaningful elements, the morphemes. For example, the word cats consists of two morphemes {cat} and {-s}. See → derivational morphology; see → inflectional morphology. Negative face. A concept used by Brown and Levinson to describe the fact that every member of a community wants to have freedom of action and freedom from imposition. Compare → positive face; see → face. Negative politeness. A subtype of the → on-record strategy for doing → face threatening acts where the speaker, in contrast to the → bald on-record strategy uses redressive action to mitigate the threat to → negative face. Compare → positive politeness. Network model. A very influential way of modelling the → mental lexicon as a combination of connected nodes. Nominalisation. A term that describes instances of language use where a process (represented in a clause) is described in terms of an object (represented in a noun phrase). Nomination. Making reference to an entity, a concept, a situation etc. in such a way that it conveys a particular attitude, e.g. the term generation snowflake to describe people born roughly between 1981 and 1995, thereby alleging an extreme vulnerability of the group. Non-canonical clause patterns. Variants of the typical SVX-clause patterns of English that involve the movement of clause constituents to positions they do not usually occur, e.g. fronting as in Distinctly brilliant I thought him, where the object complement occurs in initial position. Nucleus. In → Rhetorical Structure Theory the nucleus of two (or more) clauses is that clause which contains the main message. It is contrasted with the → satellite, which supports the nucleus. For example, in the sentence I cannot come to the meeting because I am sick the first clause is the nucleus and the causal subordinate clause is the satellite. Off record. A strategy for doing face-threatening acts where the speaker phrases their utterance in such a way that it could also be interpreted as being fully irrelevant with regard to the face needs of the addressee(s). For example, saying “it’s cold in here” to make someone close the window. Compare → on record. On record. A strategy for doing face-threatening acts where the speaker in contrast to the → off-record strategy phrases their utterance in such a way that it involves a face-threat. See → bald on record, → positive politeness, → negative politeness. Open-choice principle. A way to describe → well-formedness in the sense that language text is “a result of a very large number of complex choices. At each point a unit is completed (a word or a phrase or clause), a large range of choice opens up and the only restraint is grammaticalness” (Sinclair 1991: 109). Also called slot-and-filler model. Understood in complementation to the → idiom principle. Glossary 291 <?page no="293"?> Performance. A term coined by Noam Chomsky to describe the actual use of language by the ideal speaker-listener. Performance is understood as an application of → competence. Persuasion. In contrast to → manipulation, persuasion involves influencing the addressees with them being aware that they are being influenced. Phone. The realisation of a phoneme in a sound that you can actually hear. In contrast to → phonemes phones do not have distinctive function. Phoneme. The abstract idea of a sound. The smallest unit that distinguishes meaning, the smallest distinctive unit. Phonemes are identified through → minimal pairs. Phrasal constraint. A special kind of → lexical phrase which is similar in length to a → polyword but has at least one variable slot, e.g. ‘as far as I ___’ Pipe character |. See → | character. Place of articulation. Describes the position in which the airstream is obstructed during the articulation of → consonants. Polysemy. A → sense relation in which one word has two or more related meanings, e.g. neck (of a shirt) and neck (of the body). Not to be confused with → homonymy Polyword. A special kind of → lexical phrase consisting of a sequence of around three or four words which is fully fixed and can be understood as a single word, e.g. in a nutshell or by and large. Positive face. A concept used by Brown and Levinson to describe the fact that every member of a community claims a positive consistent self-image for themselves, which they want to be appreciated and approved of. Compare → negative face; see → face. Positive politeness. A subtype of the → on-record strategy for doing → face threatening acts where the speaker, in contrast to the → bald on-record strategy uses redressive action to mitigate the threat to → positive face. Compare → negative politeness. Precision. A measure that describes how precise a → corpus query is. Precision is understood in terms of the amount of junk that it produces. A high precision can lead to low → recall. Presupposition. An implicit assumption, something that is taken for granted even though it is not explicitly stated. The sentence The King of France is bald, for instance, presupposes the existence of a King of France. Presuppositions, in contrast to → entailments, survive negation: The King of France is not bald still presupposes the existence of a King of France. Principle of Early Immediate Constituents. A principle similar to the → Principle of End-weight that describes a tendency of clause constituents to be arranged in such a way as to increase the processibility of a sentence. Principle of End-focus. A principle of text processing that describes a tendency of sentences to provide the most important information towards the end of the sentence. Principle of End-weight. A principle of text processing that describes a tendency of heavy and complex constituents to occur towards the end of the sentence. Quality, Maxim of. A maxim of the → Cooperative Principle that demands that the speaker does not say anything that they know to be false or anything for which they lack evidence. Quantity, Maxim of. A maxim of the → Cooperative Principle that demands that the speaker provides exactly the amount of information that is needed at the current point of the communication. 292 Glossary <?page no="294"?> Recall. A measure that describes how good a corpus query is in identifying relevant cases in a given → corpus. Striving for high recall can lead to a high amount of junk, i.e. to low → precision. Register. Language variation according to the situational context in which a text is used. Relation, Maxim of. A maxim of the → Cooperative Principle that demands that the speaker’s contribution is relevant with regard to the current point of the communication. Relational antonymy. A → sense relation that describes a special kind of → antonymy where the same situation is viewed from opposite perspectives, e.g. buy and sell. Representativeness. Refers to the feature of → corpora that they, being a sample, are supposed to stand for a whole genre, register, variety or language. Resyllabification. A form of → linking between two words where the first word ends in a consonant and the second begins in a vowel. The word-final consonant of the first word is pronounced at the beginning of the second word, e.g. nude art is pronounced like new dart. Rheme. That part of a sentence that provides (usually new) information about the → theme. Rhetorical Structure Theory. An approach to the description of text structure that expresses meaning relations between clauses in terms of a semantically superordinate → nucleus and a semantically subordinate → satellite. The main message of the whole text is the one clause that functions as the nucleus to all other clauses. Satellite. In → Rhetorical Structure Theory the satellite of two clauses is that clause which supports the main message expressed in the → nucleus. For example, in the sentence I cannot come to the meeting because I am sick the first clause is the nucleus and the causal subordinate clause is the satellite. Secondary clause patterns. A set of additional clause structures that the English language provides on top of the major clause patterns (realised canonically or non-canonically). They involve moving clause constituents and additional elements. An example is the passive clause; compare → non-canonical clause patterns. Selectional restrictions. A set of syntactic rules that impose semantic restrictions regarding a word and its possible complements, e.g. the verb to drink demands a direct object that has the feature [+liquid]. Semantic preference. The typical co-occurrence of a word with a neutral semantic feature, e.g. undergo [+medical term]. Semantic prosody. the typical co-occurrence of a word with an evaluative semantic feature, e.g. cause [+negative] Semi-preconstructed phrase. A larger language pattern that consists of a combination of fully fixed, fully variable and partly variable elements, e.g. Who (expl. expr.) do NP i think PRON i B E ? Sense relations. Relations of one word to other words in the mental lexicon, e.g. the relation of → synonymy, i.e. sameness of meaning, as in couch and sofa. Sense. The description of the meaning of a term by listing all its defining features. Also called → intension. Sentence builder. A special kind of → lexical phrase which allows for variation and provides a framework for a full sentence, e.g. ‘Let my start by saying X’ Glossary 293 <?page no="295"?> Serendipitous learning. A kind of → incidental learning that involves drawing connections between hitherto unconnected parts in the learner’s stock of knowledge. Situationality. A standard of textuality that demands that a text be appropriate for the situation in which it is used. Compare → register. Slippery-slope argument. A rhetorical strategy that incorrectly assumes a chain of causation which results in a disastrous outcome if a small first step is taken. Slot-and-filler model. → open-choice principle. Sound inventory. The set of all sounds that are part of a given language. Source domain. That part (usually concrete and/ or easier to understand) of a → conceptual metaphor that helps us understand the → target domain. Straw-man argument. A rhetorical strategy that does not attack the original argument but a distorted version of the argument usually involving gross overgeneralisation, simplification or exaggeration. Sub-categorisation rules. A set of syntactic rules that makes sure that a verb has as many or as few constituents in a clause as it needs, e.g. the verb to drink demands one more constituent, the direct object. Syllogism. A pattern of reasoning in which a conclusion is derived from two premises. For example. Premise 1: All men are mortal; Premise 2: Socrates is a man; Conclusion: Socrates is mortal. Synonymy. A → sense relation that describes sameness of meaning, e.g. couch and sofa. Synthetic language. A language that expresses grammatical relations through inflectional markers. Latin and German are synthetic languages. Compare → analytic language. Tag set. The set of all tags that are used to describe the word forms in a → corpus in more detail. The size of the tag set determines the degree of detail in which the word forms can be described. Tag. A piece of information (usually the word class) attached to each word form in a → corpus. Tagging. A process whereby a piece of information is attached to word forms in a → corpus. Usually the term refers to → word-class tagging. Target domain. That part (usually abstract and/ or more difficult to understand) of a → conceptual metaphor that we want to understand. Thematic progression. A model that describes text structure as a network of clauses connected through links between → themes and → rhemes of different clauses. Theme. That part of a sentence that tells us what the sentence is about. The theme often contains given or old information. Transitivity. In Systemic Functional Grammar a set of choices that the language user has at their disposal when it comes to the representation or non-representation of elements relevant to an event and the order in which these elements occur. Trap of Enlightenment Reason. An assumption that humans are rational animals that will reach the right conclusion if they are given all of the facts. Vocabulary. Traditionally regarded as the store of words and their meaning. The → Lexical Approach advocates the more comprehensive concept of → lexis, which in addition to individual words also includes larger units. 294 Glossary <?page no="296"?> Voicing. Describes the state of the glottis during the articulation of sounds. If the glottis is closed, the vocal folds vibrate, which leads to voiced sounds. If the glottis is open, the vocal folds do not vibrate, which leads to voiceless sounds. Vowel. A sound that, in contrast to → consonants, involves modulation but no obstruction of the airstream. Weak form. An unstressed realisation of a function word which typically involves the reduction or omission of one or more sounds of → the citation form, e.g. [n] instead of [ænd] as the weak form of the word and. Well-formedness. The feature of an utterance/ sentence/ text to be in line with the rules of a grammar. Whataboutism. A rhetorical strategy that involves steering away from the issue currently in focus by asking a question of the kind “Yes, but what about XYZ? ” Wild card. In → corpus analysis the term refers to a character that can substitute for one or more character in a search string. Typical wild card characters are → ? and → *. Word-class tagging. A kind of → tagging where information on the word class is attached to every word form in a → corpus. Glossary 295 <?page no="298"?> Index Ad hominem fallacy-269 Agent deletion-200 Allophone- in free variation-62 Allophones- final devoicing-63f. in complementary distribution-62 realisation of / l/ -65f. realisation of / ŋ/ -64f. Allophonic contrasts-40 Analytic language-198 Antonymy-92 binary-92 complementary-92 directional-92 gradable-92 relational-93 Apicodental fricatives-46 Appeal-to-authority fallacy-270 Aspiration-60 Assimilation- and negation prefixes-69 and regular plural-70 progressive-69 regressive-68 Bandwagon fallacy-271 Behaghel- Law of Growing Elements-196 Second Law-191, 193 Citation form-71 Clause patterns- non-canonical-200 secondary-202 clear and dark 'l'-66 Coherence-186 Colligation-97 Collocation-96 and corpus queries-152ff. in the classroom-156 measuring association strength-153 Concordance-136-140 sorting output-137-140 Conduit metaphor-254 Connected speech-40, 67-75 Consonants- articulation-43-46 comparison English-German-46-49 manner of articulation-43f. place of articulation-44f. voicing-41f. Contraposition-269 Conventionalised indirectness-223 Corpus-133-136 and representativeness-134ff. and teaching materials-159f. as advanced ‘dictionary’-155 classroom challenges-160-164 in the classroom-155-158 precision of queries-162ff. recall of queries-160ff. Cue validity-89 Data-driven learning-155 Dative alternation-196 Discovery learning-155 Early Immediate Constituents-197 End-focus principle-193ff. End-weight principle-196ff. Enlightenment Reason- trap of-265 Error- and mistake-32 as source of information-29f. <?page no="299"?> definition by P. Lennon-33 definition of-32ff. Existential-there construction-201 Extension-254 and intension-89 Face-219ff. Face-threatening acts-221 False dilemma fallacy-272 Final devoicing-63f. Framing-264-267 Fronting-199 Full-verb inversion-200 Given-before-new principle-191ff. given-new contract-191 Grammaticalised lexis-121 Heavy-NP shift-196 Holonymy- and meronymy-93 Homography-94 Homonymy-94 Homophony-94 Hyper(o)nymy-93 Hypertheme-179 Hyponymy-93 Idea map-178 Idea web-→ Idea map Idiom principle-111ff. Incidental learning-134 Institutionalised expressions-114 Intension-254 and extension-89 Interlanguage-30ff. Intrusive 'r'-75 it-extraposition-202 KWIC-display-136 Lemmatization-141 and corpus queries-149-152 Lexicalized sentence stems-112 Lexical phrases-113-118 and necessary topics-116f. and social interaction-115f. as discourse device-117f. forms-113ff. functions-115-118 Lexico-grammar-121 Linguistics- as skill area-18, 25-28 opinions on-17f. Linking- and approximants [ɹ] and [ɻ]-75 and approximants [w] and [j]-74 Linking 'r'-75 Logical fallacy-267-275 ad hominem-269f. appeal-to-authority-270f. bandwagon-271 false dilemma-272 slippery slope-274 straw man-272ff. Mental lexicon- as network-84-87 co-occurrence relations-95-98 defining features-88f. encyclopaedic relations-90f. sense relations-91-95 Meronymy- and holonymy-93 Metaphor- and manipulation-264 conceptual in contrast to traditional-260 insect-related-260, 264 power as up-260 structural-261 water-related-264 Minimal pairs-61 Mirror neurons-49 298 Index <?page no="300"?> Negative face-220f. Negative politeness-222-226 / ŋ/ - realisation of-64f. Nominalisation-258 Nucleus-180 Open-choice principle-108-111 Passive clause-200 and manipulation-259 Patterns- and complex ideas-119 and corpora in the classroom-158 and fluency-119 and predictability-119 and processing-120 and the classroom-121-128 awareness of-128 examples from conversation-123 noticing-123-128 Phoneme- and allophone-62f. and phone-61f. definition-61 Phonemes-61 Phrasal constraints-114 Phrase structure rules-109 Pipe character ‘|’-148 Politeness- and Face-221-226 baldly on-record-223 off-record-222f. on-record-223 redressive action-223 Polysemy-93 Polywords-113 Positive face-220f. Positive politeness-222-225 Problem of too much choice-120 Psychological predicate-191 Psychological subject-191 Register-211 Resyllabification-73 Reyllabification- and final devoicing-73 Rheme-175 Rhetorical Structure Theory-180-185 multinuclear relations-184 r-sounds-47 Satellite-180 Selectional restrictions-111 Semantic preference-97 Semantic prosody-97 Sentence builders-114 Slippery slope fallacy-274 Slot-and-filler model-108 Sound inventory-40 Source domain-261 Spoken language- and events as processes -213-217, 219 and fragmentation-216 and Principle of Early Immediate Constituents-217 and speaking as process-212 Spreading activation-85 Straw man fallacy-272ff. Sub-categorisation rules-110 Syllogism-267 Synonymy-92 Synthetic language-198 Tagging-141f. and corpus queries-149-152 Target domain-261 Text- definition of-170 Textuality- standards of-172 acceptability-171 coherence-171 cohesion-170 informativity-172 Index 299 <?page no="301"?> intentionality-171 intertextuality-172 situationality-172, 209 Textual world-190 Thematic progression-175-179 constant theme-176 linear progression-175 Theme-175 Theme and Rheme-174-180 th-sounds-46 Transitivity-257ff. Voiced alveolar approximant-47 Voiced bilabial approximant-49 Vowel chart-50ff. Vowels- articulation of-50ff. ash-sound [æ]-54 comparison English-German-54ff. lip rounding-50 long central vowel [ɜː]-54 long vs. short-50 open schwa [ɐ]-54 centring diphthongs-55 tongue position-50 Weak forms-71f. elision-71 vowel reduction-71 Wild cards-149 asterisk ‘*’-147f. in the classroom-156 question mark '? '-146 Written language- and events as products-212-217, 219 and integration-216 and text as product-212 300 Index <?page no="302"?> List of Figures Fig. 2.1: Interlanguage in the learner’s progression from his or her native language to the target language (adapted from James 1998: 3). . . . . . . . 30 Fig. 3.1: voiced alveolar plosive [d] and voiced retroflex approximant [ɻ] Sources: https: / / commons.wikimedia.org/ wiki/ File: Voiced_alveolar_plosive.svg and https: / / commons.wikimedia.org/ wiki/ File: Voiced_retroflex_approximant.svg. Tavin, Nardog, CC BY-SA 4.0, https: / / creativecommons.org/ licenses/ by-sa/ 4.0, via Wikimedia Commons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48 Fig. 3.2: The vowel chart as a representation of the range of tongue positions in the oral cavity. Source: https: / / commons.wikimedia.org/ wiki/ File: Voiced_bilabial_plosive.svg Tavin, Nardog, CC BY-SA 4.0, https: / / creativecommons.org/ licenses/ by-sa/ 4.0, via Wikimedia Commons, trapezoid added. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51 Fig. 3.3: The vowels of RP. The blue circles indicate that there is some degree of variation with regard to the the exact tongue position. Note that some scholars and many dictionaries use the symbol [e] to refer to the vowel sound in set. The red circles mark those vowels that are most likely to cause problems for German learners of English. Illustration based on https: / / www.britannica.com/ science/ phonetics/ Vowels#/ media/ 1/ 457255/ 3598. . . . . . . . . . . . . . . . . . . . . . . . . . 52 Fig. 3.4: The vowels of German . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53 Fig. 3.5: The three centring diphthongs of RP and their incorrect German counterparts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55 Fig. 4.1: The English allophones of / l/ : clear l and dark l. Source: https: / / commons.wikimedia.org/ wiki/ File: Voiced_bilabial_plosive.svg Tavin, Nardog, CC BY-SA 4.0, https: / / creativecommons.org/ licenses/ by-sa/ 4.0, via Wikimedia Commons, velarization added on the right. . . . . . . . . . . . . . . . . . . . . . . . 66 Fig. 5.1: Penguin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82 Fig. 5.2: An example of a taxonomic network model (after Collins and Quillian 1969: 241) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Fig. 5.3: An example of an associative network model (after Collins and Loftus 1975: 412) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 Fig. 5.4: A lexical network around the node ‘canary’ . . . . . . . . . . . . . . . . . . . . . . . 85 Fig. 5.5: A lexical network around the node ‘milk’ . . . . . . . . . . . . . . . . . . . . . . . . . 86 Fig. 5.6: Hardy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 Fig. 5.7: Four kinds of patterns with naked eye. . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 <?page no="303"?> Fig. 5.8: Richter Scale and Gazump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Fig. 5.9: Penguin, Panda and Boat Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Fig. 5.10: Obscure, Present and Tense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Fig. 6.1: An example of a syntactic tree, the rewrite rules that created it and possible words to be filled in . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Fig. 6.2: The traditional conception of grammar as separated from vocabulary and the concept of lexico-grammar . . . . . . . . . . . . . . . . . . . 122 Fig. 6.3: Possible steps to increasing your students’ pattern noticing ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127 Fig. 7.1: The most frequent adjectives with assumption + that-clause. . . . . . . . 138 Fig. 8.1: The top ten results of the query ‘ADJ HELP_nn’ in www.english -corpora.org (with kind permission of Mark Davies). . . 151 Fig. 8.2: Specifying collocation parameter for man in BNCweb (top) and in www.english-corpora.org (bottom) (with kind permission of Sebastian Hoffmann and Mark Davies) . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Fig. 8.3: Looking for adjective collocates of the noun change in a window of L2-R3 in www.english-corpora.org (with kind permission of Mark Davies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Fig. 8.4: Looking for noun collocates of the verb to make in a window of R1-R3 in the BNCweb (with kind permission of Sebastian Hoffmann) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Fig. 8.5: Five steps to introduce working with a corpus to your students . . . . . 158 Fig. 8.6: A simple corpus-based fill-in-the-gap exercise on the light verbs to make and to do . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Fig. 8.7: A simple corpus-based fill-in-the-gap exercise on noun collocates with the light verb to make . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Fig. 9.1: An example of linear progression. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Fig. 9.2: An example of constant theme progression . . . . . . . . . . . . . . . . . . . . . . 176 Fig. 9.3: Themes and rhemes in the Silmarillion text . . . . . . . . . . . . . . . . . . . . . . 177 Fig. 9.4: An idea web of the Silmarillion text. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Fig. 9.5: An example of hypertheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Fig. 9.6: An example of an RST diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Fig. 9.7: Alternative analyses in Rhetorical Structure Theory. . . . . . . . . . . . . . . 183 Fig. 9.8: RST analysis with multinuclear relations . . . . . . . . . . . . . . . . . . . . . . . . 185 Fig. 10.1: The textual world after having read sentences a and b (left) and after having read sentence c (right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190 Fig. 10.2: The two passive transformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Fig. 10.3: The existential-there construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 Fig. 10.4: It-extraposition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 Fig. 11.1: A traffic sign. Illustration based on www.k2ksigns.com.au. . . . . . . . . . 209 Fig. 11.2: A comparison of spoken sentences in (6) and their written counterpart in (7) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 302 List of Figures <?page no="304"?> Fig. 11.3: Strategies for doing FTAs (after Brown and Levinson 1987: 61). . . . . . 223 Fig. 13.1: Mappings of the conceptual metaphor LOVE IS A JOURNEY . . . . . . . 261 Fig. 13.2: Mappings of the conceptual metaphor REFUGEES ARE LOCUSTS. . 263 Fig. 13.3: Implication and contraposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 <?page no="305"?> List of Tables Table 3.1: Manner of articulation ( 1 RP, 2 GenAm). . . . . . . . . . . . . . . . . . . . . . . . . 43 Table 3.2: Places of articulation. *Note that [w] has two places of articulation and is, hence, described as a voiced labial-velar approximant ( 1 Scot. English, 2 GenAm). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 3.3: English and German consonants. *Note that [w] is described as a voiced labial-velar approximant, i.e. it has two places of articulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Table 4.1: Final devoicing in German and its absence in English. . . . . . . . . . . 63 Table 4.2: [ŋ] in German and [ŋ] or [ŋg] in English. . . . . . . . . . . . . . . . . . . . . . 64 Table 4.3: Some weak forms of English. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Table 4.4: Examples of connected speech, including assimiliation, weak forms and linking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Table 5.1: A list of relations that a word may enter into. . . . . . . . . . . . . . . . . . . 98 Table 6.1: An overview of formal kinds of lexical phrases (based on Nattinger and DeCarrico, 1992: ch. 2). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Table 6.2: An overview of selected functions of lexical phrases related to social interaction (based on Nattinger and DeCarrico 1992: 60-63). . . . . . 116 Table 6.3: An overview of selected functions of lexical phrases related to frequent topics in conversations (based on Nattinger and DeCarrico 1992: 63-64). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Table 6.4: An overview of selected functions of lexical phrases related to discourse devices (based on Nattinger and DeCarrico 1992: 64-65). 118 Table 6.5: Examples of patterns in authentic conversation. . . . . . . . . . . . . . . . . 123 Table 6.6: Collocations with to make and their German counterparts. . . . . . . 125 Table 7.1: Examples of sentenceand KWIC-format. . . . . . . . . . . . . . . . . . . . . . 136 Table 7.2: The node and other positions in the KWIC-format. . . . . . . . . . . . . . 137 Table 7.3: Sorting on R1 and then on R3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Table 8.1: Examples of differences in query syntax. . . . . . . . . . . . . . . . . . . . . . . 149 Table 8.2: The 10 most frequent noun collocates of the adjectives tall and large in the BNCweb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Table 8.3: The top 15 noun collocates (R1 to R3) of to make and to do in the BNCweb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Table 9.1: An overview of important relations between nucleus and satellite. 182 Table 9.2: Three important multinuclear relations in RST. . . . . . . . . . . . . . . . . 184 Table 10.1: Examples of fronting (cf. Kreyer 2010b: 123). . . . . . . . . . . . . . . . . . . 199 Table 10.2: Examples of full-verb inversion (cf. Kreyer 2010b: 123-24). . . . . . . . 200 Table 11.1: verbs, their clause types and their obligatory clause elements. . . . . 213 304 List of Tables <?page no="306"?> Table 12.1: Proportion of successful requests in dependence on extent of favour and quality of information (Langer et al. 1978). . . . . . . . . . . . . . . . . 234 Table 13.1: (Non-)representation in headlines relating to Uvalde shooting. . . . 258 List of Tables 305 <?page no="307"?> ISBN 978-3-8233-8611-7 In contrast to literary or cultural studies linguis�cs is not taught in the EFL classroom, yet, it plays a major role in any English language teaching degree. Given this discrepancy it does not come as a surprise that students some�mes ask: “I want to be a teacher! Why do I need all this? ” The main goal of this textbook is to demonstrate the relevance of linguis�c exper�se for the EFL classroom. It explores a wide range of topics (phone�cs/ phonology, lexis, corpus linguis�cs, text linguis�cs and the power of language) with a clear focus on providing a convincing answer to the ques�on above. With its highly accessible style and layout, a wealth of examples and exercises as well as a large range of addi�onal innova�ve online materials this textbook sets out to convince its readers that they will be be�er teachers if they are good linguists.