Teaching and Testing Second Language Pragmatics and Interaction - A Practical Guide

TEACHING AND TESTING SECOND LANGUAGE PRAGMATICS AND INTERACTION Pragmatic ability is crucial for second language learners to communicate appropriately and effectively; however, pragmatic

TEACHING AND TESTING SECOND LANGUAGE PRAGMATICS AND

INTERACTION

Pragmatic ability is crucial for second language learners to communicate appropriately and effectively; however, pragmatics is underemphasized in language teaching and testing This book remedies that situation by connecting theory, empirical research, and practical curricular suggestions on pragmatics for learners

of different proficiency levels: it surveys the field comprehensively and, with useful tasks and activities, offers rich guidance for teaching and testing L2 pragmatics Mainly referring to pragmatics of English and with relevant examples from multiple languages, it is an invaluable resource for practicing teachers, graduate students, and researchers in language pedagogy and assessment

Carsten Roever is Associate Professor in Applied Linguistics at the University of

Melbourne, Australia

LANGUAGE PRAGMATICS AND INTERACTION

A Practical Guide

Carsten Roever

First published 2022

by Routledge

605 Third Avenue, New York, NY 10158 and by Routledge

2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN

Routledge is an imprint of the Taylor & Francis Group, an Informa business

© 2022 Taylor & Francis The right of Carsten Roever to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.

All rights reserved No part of this book may be reprinted or reproduced or utilised

in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.

Trademark notice: Product or corporate names may be trademarks or registered trademarks,

and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging- in- Publication Data

Names: Roever, Carsten, author

Title: Teaching and testing second language pragmatics and interaction: a practical guide / Carsten Roever

Description: New York: Routledge, 2021 | Includes bibliographical references and index | Identifiers: LCCN 2020054080 (print) | LCCN 2020054081 (ebook) | ISBN 9780367202811 (hardback) | ISBN 9780367203030 (paperback) | ISBN 9780429260766 (ebook)

Subjects: LCSH: Pragmatic–Study and teaching | Second language acquisition

Classification: LCC P53.62 R64 2021 (print) | LCC P53.62 (ebook) | DDC 418.0071–dc23

LC record available at https://lccn.loc.gov/2020054080

LC ebook record available at https://lccn.loc.gov/2020054081 ISBN: 978- 0- 367- 20281- 1 (hbk)

ISBN: 978- 0- 367- 20303- 0 (pbk) ISBN: 978- 0- 429- 26076- 6 (ebk) DOI: 10.4324/ 9780429260766 Typeset in Bembo

by Newgen Publishing UK

1.1 What Is Pragmatics and Why Does It Matter? 11.2 Who This Book Is For (and Some Terminology You Might Need) 3

1.3 How This Book Is Organized 4

2.1 Pragmatics 62.2 How to Talk to Whom: Pragmatic Competence 8

2.2.1 Speech Acts and Politeness 10 2.2.2 Learning L2 Speech Acts 12 2.2.3 Conclusion 14

2.3 Understanding Non- literal Speech: Implicature 15

2.3.1 Learning L2 Implicature 16

2.4 Fixed Expressions for Fixed Purposes: Routine Formulae 17

2.4.1 Learning L2 Routines 18 2.4.2 Conclusion: Implicature and Routine Formulae 18

3.3 Multiple- choice Tasks 40

3.3.1 Creating Multiple- choice Items for Routines 41 3.3.2 Creating Multiple- choice Items for Implicature 42 3.3.3 Creating Multiple- choice Items for Speech Acts 44 3.3.4 Procedure: Administering Multiple- choice Tasks 46 3.3.5 Beyond Multiple Choice: Multi- response Tasks 46 3.3.6 Resources and Further Readings 47

3.4 Discourse Completion Tasks 47

3.4.1 Types of DCTs 49 3.4.2 Designing DCTs 51 3.4.3 Procedure: Administering DCTs 53 3.4.4 Resources and Further Readings 53

3.5 Role Plays 53

3.5.1 Types of Role Plays 55 3.5.2 Target Features in Role Plays 56 3.5.3 Role Play Scenarios 56

3.5.4 Interlocutor Considerations 62 3.5.5 Procedure: How to Run Role Plays 63 3.5.6 Resources and Further Readings 64

3.6 Elicited Conversation 65

3.6.1 Designing Elicited Conversation Tasks 66 3.6.2 Procedure: How to Run Elicited Conversation Tasks 68 3.6.3 Resources and Further Readings 68

4.1 The CEFR and Other Frameworks: A Brief Introduction 704.2 A Developmentally Sensitive Pragmatics Curriculum 734.3 Pragmatics and Proficiency Levels 74

4.4 Beginners: the A1 Level Learner and Pragmatics 74

4.4.1 Getting Learners from 0 to A1 76

4.5 Upper Beginner/Low Intermediate: the A2 Level Learner and Pragmatics 79

4.5.1 Getting Learners from A1 to A2 81

4.6.1 Getting Learners from A2 to B1 86

4.7 Upper- intermediate: the B2 Level Learner and Pragmatics 88

4.7.1 Getting Learners from B1 to B2 93

4.8 Advanced: the C1 Level Learner and Pragmatics 95

4.8.1 Getting Learners from B2 to C1 98

4.9 The High Advanced Learner: Pragmatics at the C2 Level 101

4.9.1 Getting Learners from C1 to C2 103

4.10 Resources and Further Readings 1034.11 Conclusion 103

5.1 Overview: Findings and Issues in Teaching L2 Pragmatics 105

5.1.1 Effectiveness of Teaching L2 Pragmatics 106 5.1.2 Factors in Teaching Pragmatics 107

5.2 Materials for Teaching Pragmatics 1085.3 Phases of a Pragmatics Lesson 110

5.3.1 Step 1: Presenting the Target Feature 112 5.3.2 Step 2: Receptive Practice 115

5.3.3 Step 3: Productive Practice 118

5.4 Teaching a Feature Across Levels 122

5.4.1 Requests at A1 Level 122 5.4.2 Requests at B1 Level 130 5.4.3 Requests at C1 Level 137

5.5 Resources and Further Readings 140

6.1 Pragmatics Testing So Far 1426.2 Testing in Pragmatics Research 1456.3 Pragmatic Norms 148

6.4 Assessing Learning: Classroom- based Testing of Pragmatics and Interaction 149

6.5 How to Make Sure Tests Work: Validation of Large- scale Tests 1506.6 Validating “Objective” Tests 153

6.7 Validating Sociopragmatic Judgment Tests 1556.8 Validating Productive Tests and their Rating Scales 1586.9 Fairness and Bias in Testing of Pragmatics 167

6.10 Conclusion 1686.11 Resources and Further Readings 169

7.1 Pragmatics in General Language Teaching 1707.2 Specific- purposes Pragmatics: Needs Analyses, Tasks, and Indigenous Criteria 173

7.3 Making Pragmatics Tests Practical: the Role of Technology 175

inter-1.1 What Is Pragmatics and Why Does It Matter?

To put it simply, pragmatic knowledge is knowing how to adapt one’s language use

to other people Consider the following scenario:

You are in the post office and filling out a form when your pen dies There

is a woman about your age right next to you She has just finished filling out her form and is folding it You would like to borrow her pen What do you say?

It is perfectly grammatically correct in English to say any of the following:

1 “Give me the pen.”

2 “Can you pass me your pen?”

3 “My pen doesn’t work I so wish I had a pen.”

4 “Excuse me, could I borrow your pen for a second?”

5 “Forgive the intrusion but I was wondering if you wouldn’t mind letting me use your pen for just a moment I’ll return it right away.”

too direct for the situation, and the second one seems a bit abrupt and not polite enough The third one is so arcane and roundabout that the other person may not even know what the speaker wants but the fourth one is quite good, and the fifth one would probably get the job done but is possibly a bit too much

Now imagine the speaker (“you”) and the other person are best friends That would change radically what fits the situation and what does not In that case, the first one is probably okay, the second one might already be a bit too polite, the third one could be humorous, and the final two seem completely over the top

The fact that the relationship influences how we talk might be obvious to native speakers and high- proficiency speakers, but how would lower- level learners know this? How do learners know how to talk to people depending on their rela-tionship with the person and the situation? The stakes are high here: talking to someone in a way that is culturally inappropriate can make speakers appear rude even though they do not mean to be Yet, how to use language with people appro-priately is not systematically taught in most language courses, not systematically developed in most language textbooks, not included in most language tests, and quite scattered even in such an influential document as the Common European Framework of Reference for Languages (CEFR; Council of Europe, 2001, 2020)

This state of affairs is highly problematic because it leads to a phenomenon that communicative language teaching (CLT) is supposed to avoid: learners know about the target language but cannot actually use it appropriately in daily inter-action CLT has certainly made great strides in the direction of fostering ability for use through its “fluency over accuracy” paradigm But this view does not necessarily take into account that language is always used with other people and language use is designed for the person(s) we are speaking to Fluency must be supplemented with appropriateness to make learners effective communicators

The situation is made worse by the absence of pragmatics in language tests

Unfortunately, what is not tested does not get taught (much) In most language instruction settings, language teachers and learners focus on content that allows learners to pass exams and obtain certifications There is nothing in principle wrong with this as long as the tests create “positive washback,” i.e., they encourage the teaching and learning of useful material and thereby make learners better language users, not just better test takers Tests can create positive washback by including desirable material but at this point, pragmatics is not specifically assessed

in any of the big language tests be it IELTS or TOEFL for English, TestDaF for German, HSK for Chinese, or others Some of them contain some elements of pragmatics, for example, the tests in the Cambridge Main Suite assess some inter-actional abilities but represent a fairly narrow perspective on pragmatics

Why are pragmatics and abilities for social interaction so underemphasized

in language tests? Is it even possible to test such highly contextual and situation- dependent skills? And if it is, can scorers agree on what is the “right way” to say something? These questions have actually been the subject of quite a bit of

2

research, and there is now quite a bit of knowledge about testing of pragmatics but

it has not really trickled through to many large- scale tests or classrooms

It is the purpose of this book to familiarize readers with how pragmatic and interactional abilities can be taught in language classrooms and how they can be assessed in teacher- made or large- scale tests

1.2 Who This Book Is For (and Some Terminology You Might Need)

This book is primarily aimed at practicing second language teachers, ulum designers, and test makers It assumes some background in second language teaching and learning but not in- depth background in linguistics, applied linguis-tics, or second language acquisition Having done an introductory course or sub-ject on second language learning and teaching should be sufficient to follow along.The book is also useful to students in teacher training courses for ESL or other languages, or students in applied linguistics/ TESOL/ second language studies degrees who want to know more about pragmatics and how it can be taught and tested It can also be a resource for researchers who work on teaching and testing

curric-of second language pragmatics, though it does not specifically focus on research methodology Instead, it is intended as a practical guide that can help with the design of instructional sequences, teaching materials, and tests

The book is suitable for second language teachers of any modern guage: English, Chinese, German, Spanish etc Most examples in the book are in English though a few are from other languages However, pragmatics is an issue for the learning of any language, just like grammar, so everything in the book can

lan-be applied to any language

There is some terminology that I will use frequently, and which might not be entirely familiar to all readers:

• Second language: When I talk about “second language,” this means a

lan-guage other than a person’s native lanlan-guage It also includes third, fourth and further languages, and does not distinguish between language learning in the target community (second language setting, e.g., learning English in the US)

or outside (foreign language setting, e.g., learning French in Germany)

• Pragmatics: Chapter 2 will talk in detail about what “pragmatics” is but as a

shorthand definition, pragmatics is an area of language study that investigates how language relates to the real world It looks at how the context of lan-guage use influences the way we talk, and how we use language to make things happen

• Interaction: An interaction is two or more people communicating with each

other, typically by means of spoken language It is also possible to interact in writing (e.g., text chat) or through gestures and facial expressions but these modes of communication will not play a big role in this book

4

• Interlocutor/ interactant: These terms are used fairly interchangeably

to describe people involved in an interaction The term “interlocutor”

is also sometimes used more specifically to describe a tester or researcher who interacts with a learner in a role play or other elicited interaction An

“imaginary interlocutor” is the person described in an interactional task who

is not actually present, for example, the woman in the post office in section 1.1 above

• Pragmatic competence and interactional competence: Chapter 2 will

talk about these key terms in more depth but suffice it to say that throughout the book, “pragmatics” and “pragmatic competence” will be used as general terms that also include interaction and interactional competence

• Recipient design: Recipient designing an utterance means to formulate

it specifically for a particular hearer This includes levels of politeness with close friends and strangers when asking for a pen, but also background know-ledge What can the speaker expect the hearer to know and how much back-ground information do they need to provide to be understood? Speakers make decisions about this and structure their talk accordingly

• CEFR: The Common European Framework of Reference for Languages

(Council of Europe, 2001, 2020) is a lengthy document that describes what learners at different proficiency levels can be expected to do with the target language Its levels A1 and A2 are roughly beginner levels, B1 and B2 are intermediate levels, and C1 and C2 are advanced levels The CEFR will be discussed in more detail in Chapter 4

1.3 How This Book Is Organized

Chapters 2 and 3 provide background information for readers with little previous experience in second language pragmatics and interaction Chapters 4 through

7 are the core parts of the book, outlining a novel perspective on teaching and testing pragmatics Chapter 2 goes into detail about what pragmatics and inter-actional competence is, and what we know about learners learning to be prag-matically and interactionally competent It provides a quick overview of this fairly broad area to orient readers who may not have a great deal of background on pragmatics in general and second language pragmatics specifically

Chapter 3 shows a variety of teaching and testing tools for second language pragmatics that can be adapted for learners at different levels of proficiency and pragmatic competence

Chapter 4 proposes a “pragmatic curriculum” by discussing what aspects of pragmatics can be taught at various levels of language proficiency The chapter uses the CEFR’s proficiency levels as a general guide, and relates them to teaching of pragmatics and interaction This is a core chapter, proposing a curricular structure that is novel in second language pragmatics and curriculum work

Chapter 5 talks about planning sequences of instruction for pragmatics How are pragmatics lessons organized? What ways of teaching pragmatics are effective? Why is awareness raising of cultural differences such an important part of teaching pragmatics? The chapter also demonstrates how the same pragmatic feature, requests, can be taught at beginner level, intermediate level, and advanced level.Chapter 6 is about testing of pragmatics It summarizes the research done in this area, and then looks at the issue of pragmatic norms (who says what is pragmatic-ally appropriate?) and how to make sure pragmatics tests actually test pragmatics, and not just general proficiency It also shows how to construct and validate tests for different aspects of pragmatics

Chapter 7 provides an outlook on the future of pragmatics in second language teaching and testing It discusses general language teaching and teaching language for specific purposes, and makes suggestions about how technology can help make testing of pragmatics more feasible

This is all captured in a well- known formal definition by Crystal, who defines pragmatics as “the study of language from the point of view of users, especially of the choices they make, the constraints they encounter in using language in social interaction and the effects their use of language has on other participants in the act of communication.” (1997, p 301).

To illustrate what pragmatics is all about, imagine a situation where Jane’s phone is out of battery and she asks Mark to borrow his phone to make a call

Here are three ways this imaginary conversation might run:

6

1 Casual:

Jane: “Where’s your phone? Mine’s out of battery.”

Mark: [takes out his phone] “Here.”

Mark: “Of course, no problem Here you are.” [takes out his phone]

Pragmatics tries to understand under what circumstances Jane might choose to speak casually, neutrally, or politely, which is what Crystal means by “choices” and

“constraints.” What makes her choose one option over another? It is clear that the major consideration is not which option is grammatically correct: they all are So her choice is not motivated by the constraints of the grammatical system of her language, but it is based on her point of view as a language user regarding her rela-tionship with Mark Consider the following options:

• If Mark is Jane’s husband, speaking casually would be quite normal but speaking politely would be odd because speakers do not usually use a lot of politeness with people they are very close to

• If Mark and Jane are friends, Jane might choose to speak causally or neutrally, depending on how close they are With a very close friend, she can be casual, whereas with someone who is more of an acquaintance than a friend, she might prefer the more neutral option However, the polite option would be a bit too much for talking to a friend

• If Mark is Jane’s boss, the casual option would be a bit short, the neutral option might be suitable if they work together closely and know each other well, and the polite option might apply where their relationship is more distant

While much of pragmatics research cares a great deal about the speaker (Jane), the hearer also matters Note how Mark replies to Jane: his replies get longer the more Jane says, so the way Jane phrases her request seems to affect Mark’s response It

is also notable that some of Mark’s responses are chunks that language users have said and heard many times, such as “no problem,” “here you are.” The meaning of these chunks is more or less transparent and more or less limited to a particular situation: “no problem” is pretty obvious in its meaning and can be said in many different situations However, the literal meaning of “here you are” has little to do with how this expression is used, namely when handing someone something, and

it is only used in that situation

of speech acts and politeness research has tried to answer these questions.

2 How does the listener know what the speaker means? When Jane asks casually

“Where’s your phone? Mine’s out of battery,” most language users would understand this to mean that she wants to use Mark’s phone, not just know

where it is This question has been addressed in work on implicature.

3 Why do people frequently use the same set phrases, such as “here you are,” to

convey a particular meaning in a situation? Research on routine formulae

has provided insight into this question

4 How do people know when to talk, how do they avoid talking over each other, how do they know that a positive answer to a request can be short rather than involve elaborate explanation? These fundamental questions have

been investigated in work on interactional competence.

Figure 2.1 shows the areas of pragmatics which have been investigated in L2 pragmatics and interactional competence research and which will feature prom-inently in this book

The next section tackles a crucial question: how language users know that they need to adjust the way they talk depending on who they talk to

2.2 How to Talk to Whom: Pragmatic Competence

How does Jane know how to adjust the way she talks, depending on whether she

is talking to her boss, friend, or husband? This ability is commonly referred to as pragmatic competence but what does a language user need to know to be prag-matically competent? Leech (1983) described two areas of pragmatic competence interactants need to have: sociopragmatic and pragmalinguistic competence, as shown in Figure 2.2

A language user’s sociopragmatic competence is their knowledge of the

social rules and norms that govern language use What is normal and acceptable?

Pragmatics

Speech acts Implicature Routines Interaction

FIGURE 2.1 Pragmatics and its areas

8

Who deserves respect and who would be considered your equal? How do gender and age affect how you talk to someone? What is considered a small favor and what is a huge imposition on the interlocutor? As Fraser et al (1980) put it so succinctly, sociopragmatic knowledge means knowing “what you do, when, and

to whom” (p 79)— it is knowledge about how social relationships affect talk This knowledge includes knowing that you can talk casually to people who are very close, like a spouse, but need to be more polite to people who are less close, such

as an acquaintance

Sociopragmatic knowledge is largely implicit: language users do not usually have to think about what their relationship with another person is and whether they should be casual or polite They just know intuitively, at least if they were brought up in the target speech community, but second language users may struggle with sociopragmatics Not having been socialized in the target culture, they may lack an intuitive understanding of how people relate to each other and what is “normal,” and they may unwittingly give offense, for example, by misreading a formal situation as a casual situation and treating people who deserve respect and deference like close friends

The second and equally important component of pragmatic competence is

pragmalinguistic competence Pragmalinguistic competence includes

know-ledge of linguistic tools for saying what you want to say while being guided and constrained by sociopragmatic rules and norms Pragmalinguistic knowledge is essentially language knowledge, but importantly, it is knowledge of how language can be harnessed for pragmatic purposes How do you formulate a request, or suggestion? What do you say to open a story telling in a conversation? More gen-erally, what are the conventional linguistic tools for expressing certain intentions?

A second language learner can fall short in terms of their pragmalinguistics simply because their overall language competence is low For example, Jane’s casual request for the phone (“Where’s your phone?”) does not require much com-plex language, although low- level learners might still struggle with formulating

Pragmatic competence

Sociopragmatic competence Social rules

Pragmalinguistic competence Linguistic tools

FIGURE 2.2 Sociopragmatic and pragmalinguistic competence

10

a question But her neutral and polite requests are linguistically much more complex: in addition to question formation, they require modals, if- clauses, and expressions such as “I was wondering.” Being able to produce these language features requires more than a basic level of language competence If a learner is

at a low level of proficiency and has not yet acquired English modals (can, could, would etc.), their linguistic toolset for making requests in English will be limited

This has important implications for curriculum design, which will be addressed

in the next chapter

While having sociopragmatic and pragmalinguistic knowledge is crucial for successful language use, it is not sufficient Both types of knowledge need to be mapped onto each other, i.e., language users need to be able to use the right kind

of linguistic tools given the social rules For example, it is useful to know what linguistic options there are for making requests and softening or strengthening them, but this knowledge needs to be connected to sociopragmatic knowledge

of how to speak to a certain interlocutor under particular conditions In addition, language users must also have fast and effortless access to their knowledge to com-prehend and produce language under the time pressure of real- time conversation

While sociopragmatic and pragmalinguistic knowledge are useful concepts, they are not detailed enough to explain how language users adjust their language

to an interlocutor and the situation What features of the interlocutor and the communicative situation do language users take into account when deciding how

to talk?

2.2.1 Speech Acts and Politeness

The main reason people adjust the way they talk based on the listener is to tain social harmony This adjustment happens whenever we speak to another person but has been thoroughly investigated for a particular type of sociolinguistic action called a speech act (Austin, 1962; Searle, 1969, 1976) A speech act is the use

main-of language to accomplish something in the world, or in Austin’s (1962) lation, “how to do things with words.” Speech acts include such linguistic actions

formu-as requesting, apologizing, refusing, suggesting, complaining, criticizing, thanking, complimenting, congratulating, greeting, and others

How language users go about deciding the design of a speech act was investigated in a famous study by Brown and Levinson (1987), who collected data from three very different communities in Mexico, Sri Lanka, and the US Based

on earlier work by Goffman (1959), they claimed that speakers take into account two different aspects of their hearer when producing speech acts One aspect is

“positive face,” which describes people’s desire to be appreciated, liked, seen as

group members, and to live in harmony with others

The other, complementary aspect is “negative face,” which is people’s desire

not to be coerced, not to suffer loss or disadvantage, and not to do things they

do not want to do Speakers orient towards these so- called “face wants” by

do not want to do To avoid damage to the relationship, the speaker can ledge the face threat by deploying various politeness strategies Most commonly, they would avoid a direct command (“Give me your phone.”) and use a strategy known as “conventional indirectness,” which Jane used when speaking neutrally and politely by saying “Can I have your phone?” A question like “Can I have your phone?” looks on the surface like an ability question but the conventional meaning relates to willingness It is such a common strategy that a competent lis-tener would be able to infer that they are being asked for help, rather than about ability.

acknow-Speakers also use hedges to soften the request and make the burden placed

on the hearer appear smaller Jane uses a hedge when she asks about making a

“quick” call She also uses external modifications to further lower the impact of the request by providing a reason (“Mine has run out of battery.”) Finally, a com-pletely different strategic approach would be to hint (“Oh no, my phone is out of battery, and I really need to make a call.”)

All these strategies serve to make the request more polite or appear less demanding and thereby increase the likelihood of acceptance But what sociopragmatic criteria do speakers use to determine what level of politeness is needed? Brown and Levinson (1987) identified three major factors that speakers weigh up in their relationship to the hearer to determine how to make their speech acts:

• Power (P): Power describes the power differential between hearer and speaker

A high Power difference (P+) means that the hearer is in a more powerful position than the speaker, for example, the hearer might be the speaker’s boss, parent, coach, or professor Equal Power (P=) means that the hearer and speaker do not have control over each other, e.g., classmates, colleagues, teammates, or siblings Lower Power (P- ) means that the speaker is in a more powerful position over the hearer Usually, talking to someone higher in Power requires more politeness than talking to an equal or someone lower

in Power

• Social Distance (D): Social Distance is the degree of closeness between the

speaker and hearer This can refer to the degree of acquaintanceship, i.e., how well speaker and hearer know each other, but it also takes into account if they are members of the same social group For example, high Social Distance (D+) is normally the case where speaker and hearer do not know each other, e.g., between strangers However, if they have other things in common, such

as similar age, same gender, working for the same company, they may feel

12

some degree of commonality, and make Social Distance medium (D±)

Low Social Distance (D- ) exists where interactants are close to each other, e.g., family members, friends, close colleagues Interestingly, Wolfson (1983) showed that medium Social Distance actually leads to the most politeness, probably because the relationship is not very clearly defined

• Degree of Imposition (I, sometimes also shown as R for “ranking of

imposition”): Imposition is the amount of trouble for the hearer For requests,

a high Imposition request (I+) causes a lot of cost to the hearer in terms of money, time, inconvenience, or social sanctions For example, asking someone

to use their car for a weekend, borrow a large amount of money, or help the speaker cover up a crime are all high Imposition requests By contrast, a low Imposition request (I- ) only costs the hearer minimally, e.g., borrowing a small item or small amount of money, helping the speaker do an easy task, or providing some information that is readily available

These context variables are all subject to cultural variation There can be large differences between cultures in views of who is more powerful, who is considered

a peer, and what causes a great deal of cost to the hearer Knowing these culture- specific social norms is a key part of sociopragmatic knowledge

It is also worth noting that Brown and Levinson’s (1987) approach is not uncontroversial It has been criticized for being too strategic and utilitarian, making it seem like polite language use in speech acts is carefully planned

to manipulate the interlocutor (Mao, 1994; Matsumoto, 1988) It can tainly happen that speakers consciously decide how to talk to an interlocutor

cer-to achieve a particular goal, which is part of their strategic competence in Bachman and Palmer’s (2010) model of communicative competence However, certain aspects of politeness are simply required by cultural rules, e.g., polite forms in Japanese when talking to a superior or the distant address term (Sie) in German when talking to a stranger While speakers take into account the social relationship when formulating a speech style in Japanese or an address term in German, they do not do so to gain an advantage but they just act in accordance with the social relationship with the listener

Furthermore, there are other factors which impact how people talk For example, relative age is an important consideration in some cultures, and people also orient to their degree of entitlement, i.e., whether they have a right to make

a particular request (Curl & Drew, 2008) Research into the sociopragmatic rules that language users follow is still ongoing and they are by no means as fixed

as grammatical rules An approach like Brown and Levinson’s (1987) should be considered a partial explanation rather than a deterministic, immutable system

2.2.2 Learning L2 Speech Acts

Probably due to their strong prevalence in general pragmatics research, learner performance with speech acts has been extensively researched in L2 pragmatics

12

The most investigated speech acts are (in order) requests, refusals, and apologies, with others having received much less attention In terms of their development, learners move from short and simple speech acts that are not carefully crafted to

a particular interlocutor to longer, complex, and more fine- tuned speech act duction Kasper and Rose (2002) outline five stages of the development of speech act production:

pro-1 Pre- basic stage: Learners have so little L2 competence that they use ever linguistic and non- linguistic means they have to convey their speech intentions to their interlocutor There is no adaptation of utterances to different interlocutors

what-2 Formulaic stage: Learners use linguistic means to express speech acts, such as high- frequency chunks or simple verb forms Their production is simple, and there is no variation in politeness so they still do not adapt their production

to the interlocutor or situation

3 Unpacking stage: Learners have a wider range of pragmalinguistic tools able, including conventional indirectness for requests (“Could you…”) Their utterances become longer and there is some degree of variation based on situation and social context factors (Power, Distance, Imposition) though mapping of pragmalinguistics and sociopragmatics is still largely incomplete

avail-4 Pragmatic expansion: As their overall proficiency increases, learners’ range of pragmalinguistic options also widens Their speech act production becomes more complex and they can attain higher politeness levels However, there may still be mismatches between sociopragmatic and pragmalinguistic know-ledge, and learners may actually say too much as they may lack sociopragmatic constraints on their thoroughly developed pragmalinguistic ability

5 Fine- tuning: Learners have a wide range of pragmalinguistic tools available, and the mapping of pragmalinguistic tools and sociopragmatic rules is mostly accurate Their production gets increasingly adapted to different situations and interlocutors

Kasper and Rose’s (2002) sequence makes it clear that learners at early stages

of development are constrained by their overall proficiency in what they can

do pragmatically: if learners have few linguistic tools and need to deal with a high processing load in producing and comprehending language, they are not able

to formulate complex speech acts As their repertoire of pragmalinguistic tools increases and their language production becomes more automatized, requiring less processing resources, they are able to make their speech acts longer and more com-plex, using modals, integrating hedges, and deploying modifications to strengthen

or soften the speech act Overall, lower- level learners tend to be too brief and direct, and as their competence increases, their pragmatic production becomes longer and more indirect

However, longer is not necessarily better Blum- Kulka (1982) described a

“waffling phenomenon” and Shively (2011) as well as Al- Gahtani and Roever

14

(2015) showed similar trajectories where learners, especially at an intermediate stage, “talk too much,” e.g., their apologies or requests might be more detailed and effusive than the situation requires This is because their pragmalinguistic produc-tion is not constrained by their sociopragmatic knowledge to inform them how much politeness is needed in a given situation Over time, they get socialized into the target community’s ways of “thinking, feeling and acting” (Ochs, 1986, p 2)

Of course, this process is not limited to L2 learners: people entering a new munity, such as a workplace, sports team, or friendship group, get socialized into the specific rules and norms the community follows

com-Similar to other aspect of second language learning, transfer from L1 has also been found in L2 pragmatics It is nearly impossible for (adult) learners to avoid sociopragmatic transfer as they have fully developed sociopragmatic knowledge

of social relationships in their L1, which they can fall back on initially in the L2 setting The case for pragmalinguistics is different since beginning- level learners have so little pragmalinguistic repertoire that they cannot even transfer As they develop, they may transfer L1 sociopragmatic- pragmalinguistic mappings, e.g., in Shively’s (2011) study where learners used US- style greetings and requests in a service encounter setting in Spain whereas the local preference was for fast and direct task fulfillment

Also similar to general SLA, development in L2 pragmatics does not ably and smoothly run its course from low competence to near- native ability In

predict-a study of lepredict-arners of L2 Modern Greek, Bellpredict-a (2014) found thpredict-at even predict-advpredict-anced learners at CEFR C1 level were still noticeably different in their use of refusal strategies from native speakers of Greek in that they under- used some strategies and over- used others Furthermore, pragmatics is so closely connected to how people portray themselves to others and how they want to be seen that learners may sometimes choose not to follow target language pragmatic norms simply because they feel uncomfortable doing so Siegal (1996) did a classic study on Western women learning L2 Japanese, who in some cases consciously decided not

to follow the typical female Japanese speech style because they felt it was contrary

to their identity as equals to men Kim (2014) describes Korean ESL learners’

struggle to use first names with professors in a Western context, where the use of first names was the norm Some learners decided not do so as it just made them feel too uncomfortable, thereby consciously transferring the L1 norm and vio-lating the local sociopragmatic norm (see Ishihara, 2019, for an extended discus-sion of identity and transfer)

2.2.3 Conclusion

Sociopragmatic and pragmalinguistic knowledge are at the core of pragmatic ability, and connect language with social norms Language users evaluate social relationships and adjust what they are trying to express accordingly This can be difficult for second

14

language learners due to their developing sociopragmatic and pragmalinguistic ities, which may lead them to deviate from the target language norm, and they may come across as rude, too polite, or just generally odd in an L2

abil-While research on speech acts and politeness has been very productive, not everything is about politeness: in addition to indirectness or hinting in speech acts, language users also employ other ways of being less than literal

2.3 Understanding Non- literal Speech: Implicature

In addition to speaking pragmatically appropriately, interactants also need to be petent at pragmatic comprehension An important part of pragmatic comprehension

com-is to understand indirect or non- literal language use People frequently say things they do not mean literally to imply another layer of meaning For example, a parent saying to a child “I’ve asked you a thousand times not to leave your clothes on the floor.” does not mean literally that the request has been uttered exactly 1,000 times but implies annoyance for having to ask repeatedly Implying meanings beyond the surface level of an utterance is called conversational implicature, and goes back to the philosopher Grice (1975) and his conversational maxims Grice claimed that interactants generally operate under the assumption that their interlocutors are being cooperative (the Cooperative Principle) and are trying to provide the right level of information (Maxim of Quantity), to be truthful (Maxim of Quality), to be relevant (Maxim of Relation), and to be clear (Maxim of Manner) Implicature arises when

a speaker intentionally disregards (“flouts”) these maxims to imply an extra layer of meaning It is the hearer’s task to infer what additional meaning is implied In the above example of a parent claiming to have reminded their child “a thousand times,” the Maxim of Quality is flouted with the implication that they have had to remind the child more often than they should have

Flouting the Maxim of Quantity also gives rise to interesting implicatures For example, after dinner at a new restaurant, A asks B: “What did you think about dinner?” and B replies, “The waiters were friendly.” This does not provide enough information, since the quality of the service is only one component of a dining experience and arguably less important than the quality of the food, but by praising only one part of a whole, B implies criticism of the rest

Similarly, flouting the Maxim of Relation also creates implicature If an actant responds with a seemingly unrelated utterance it is up to the other person

inter-to infer what additional meanings are implied For example, responding inter-to the question “How do you like my new shoes?” by saying “I can’t believe how hot it’s been.” seems unrelated and irrelevant but a pragmatically competent inter-locutor can infer that this obvious change of topic conveys an unwillingness

to discuss the proposed topic, and most likely disapproval of the new shoes

A special case of flouting the Maxim of Relation is the Pope Q (“Is the Pope Catholic?”) and its variations Following a yes/ no question (e.g., “Are we going

16

to get in trouble for submitting the report late?” – “Is the Pope Catholic?”), the Pope Q and its derivatives imply a “yes” response since the Pope is indeed Catholic

Finally, flouting the Maxim of Manner involves responding in a roundabout way For example, if one colleague asks another “Can you proofread this report for me?” and the other replies “Well, I’m really busy on my project and was hoping

to go home early today.” the respondent has strictly speaking not answered the question but to a competent interlocutor, a negative reply is implied by giving ample reasons that prevent compliance Grice’s (1975) original analysis has since been revised by Sperber and Wilson (1995), who proposed that all implicature occurs because hearers assume the relevance of an utterance and try to come up with an interpretation that makes sense

Conversational implicature is widely used in interaction, so it is important for learners to be able to interpret these instances of non- literal language and com-prehend the implied meaning

2.3.1 Learning L2 Implicature

Taguchi and Yamaguchi (2019) provide an authoritative overview of work on L2 implicature Key findings in second language implicature research include the importance of proficiency in implicature comprehension, and the facilitative effect of conventional responses

Proficiency is a key driver of implicature comprehension Roever et al

(2014b) found that a high- proficiency learner group had five times the ture comprehension ability of a beginner- level group Compared to the ability

implica-to produce speech acts and recognize routine formulae (see section 2.4 below), implicature comprehension was the aspect of pragmatics most sensitive to pro-ficiency effects

While proficiency is helpful to learners in processing implicature, a series of studies by Taguchi and collaborators (e.g., Taguchi, 2011, 2012; Taguchi, Li &

Liu, 2013) found additional factors that contribute to implicature difficulty Most importantly, they demonstrated that implicature comprehension is easier where the implied meaning is similar to the propositional meaning, and the implied meaning is conventionally encoded In a study by Taguchi (2012), implicature in refusals, which followed a conventional format of giving a reason (e.g., “Do you want to come jogging with me?” – “I have to finish my term paper.”), was more easily comprehended than less conventionally structured implicature for giving opinions (e.g., “How was the weather during your trip?” – “I’m glad I brought a sweater.”)

Taguchi and Yamaguchi (2019) argue that more conventional implicatures are easier to interpret because listeners know implicitly that giving a statement in response to an invitation constitutes a refusal since an acceptance is normally

16

done directly and positively (see also “preference” in the section on Interactional Competence below) By contrast, the range of possible interpretations is larger for non-conventional responses, and listeners need to activate background know-ledge, i.e., that people wear sweaters when it is cold and that therefore the weather during the trip must have been cold In extreme cases of non- conventionality, such as the original Pope Q, lack of background knowledge precludes compre-hension entirely but modified versions of the Pope Q (e.g., “Do fish swim?”) are more accessible, as Roever (2005) found

Taguchi and Yamaguchi (2019) also point out that the degree of shared ventionality between L1 and L2 has an important impact on comprehension Indirect criticism by praising an irrelevant part of a whole (“What did you think

con-of his house?” – “The bathroom was pretty.”) is not a conventional pattern in some languages, making comprehension extremely difficult, but this convention can easily be taught, as Bouton (1999) demonstrated

Overall, work in L2 implicature has shown that learners need to mobilize various resources, such as their knowledge of discourse conventions, background knowledge, and language ability to interpret implied meaning The greater the distance between propositional and implied meaning and the less conventional the implicature, the more difficult it is to interpret

2.4 Fixed Expressions for Fixed Purposes: Routine Formulae

Another aspect of pragmatic competence is to know fixed expressions associated with particular situations and social roles These are known as routine formulae (Coulmas, 1981) or conventional expressions (Bardovi- Harlig, 2009), and they tend to recur often in a similar form and under similar circumstances For example, expressions like “how are you?,” “thank you for having me,” “can I get you any-thing else?” are tied to particular situations, such as greeting someone, expressing gratitude for someone’s hospitality, and a server addressing a customer in a food service encounter

Routine formulae can be tied very closely to a particular social role and ation For example, the requirement for using “no, thanks, I’m full” are quite restrictive as it can only be used in response to an offer of food by host or fellow diner in a mealtime setting By contrast, “how are you doing?” is only restricted in that it is a greeting, but it can otherwise be said in many settings

situ-Routine formulae also vary in how fixed and immutable they are The tude formula “thanks a lot” cannot be changed, i.e., *“thank you a lot.” However,

grati-“I’m full” allows some limited modification (pretty full/ quite full/ really full) and other formulae are so flexible that it is questionable whether they still qualify as formulae, e.g., “I am/ was wondering if…,” which has an infinite number of pos-sible completions

chal-a word of English will be chal-able to deploy “how chal-are you?” or one of its vchal-arieties very quickly as a greeting even though s/ he would not be able to produce other questions with inversion and a conjugated verb form for quite a while It is pos-sible for learners to use routines that exceed their general competence because learners acquire routine formulae as unanalyzed chunks, which they can store in memory and retrieve when required In the case of “how are you?” the very clear and unambiguous connection between the chunk and its meaning and as well as the chunk’s high frequency make it easy to store and use (see Eskildsen & Kasper,

2019, for a usage- based explanation) Dechert (1983) has therefore called such formulae “safety islands” (p 183) which figuratively offer learners refuge in a sea

of incomprehensible language

At the same time, the fixed nature of routines leaves little room for error

Bardovi- Harlig (2009) has shown that learners actually do make errors in their routines production, especially with longer or grammatically more complex formulae This may simply be a memory limitation as less highly frequent for-mulae are probably not as fixed in learners’ memory and must be reconstructed

This also explains why proficiency has some effect on routines knowledge, though less than on implicature and speech acts: in Roever et al.’s (2014b) study, advanced learners scored three times as high as beginners on a recognition test

of routine formulae, compared to a five- fold proficiency advantage for ture, and a four- fold advantage for speech acts However, Roever (2012) showed that even short- term exposure of three months or less leads to a strong increase

implica-in recognition of routimplica-ine formulae This is probably due to the high frequency and high utility of routine formulae for learners in the target language context, but it raises a question for teaching and learning of routine formulae in foreign language classrooms: even though such formulae are useful to learners once in the target language culture, is their teaching justified and feasible in a foreign language setting where they are not reinforced by situational exposure and con-stant use?

2.4.2 Conclusion: Implicature and Routine Formulae

Implicature comprehension and knowledge of routine formulae are smaller areas of pragmatics compared to speech acts and interactional competence (see below) However, learners do face specific challenges with these aspects

of pragmatic competence, and this makes them appropriate targets for teaching interventions

2.5 Interactional Competence

While it is convenient for researchers to investigate speech acts, routines, and implicature as individual topics, this is of course not how language users handle them Language users deploy their knowledge of speech acts, routines, and impli-cature when they interact with each other but in addition they activate a great deal more knowledge to ensure they can understand each other in interaction This competence specifically relevant to interacting has been called “interactional com-petence” (IC; Hall & Pekarek Doehler, 2011; Pekarek Doehler, 2019) As Young (2019) outlines, there are different views on what constitutes it, but from a second language perspective, the approach by Hall and Pekarek Doehler (2011) is the most fruitful

Following Hall and Pekarek Doehler (2011), IC is language users’ ability to coordinate social actions with each other in mutually recognizable ways By “social action,” IC researchers mean anything a language user says, does, or signals in social interaction A speech act like request, apology, refusal, greeting, compliment etc is

a social action, but the concept of social action goes further For example, imagine that Anna and Bella are chatting over coffee, and Anna says to Bella: “Guess what.” This is difficult to classify as a speech act but it clearly does something in con-versation: it lays the groundwork for telling news By doing the social action of a

“pre- announcement,” Anna lets Bella know that she wants to tell her some news and thereby attempts to secure Bella’s recipientship, i.e., make sure Bella is ready

to listen Both interactants need interactional competence: Anna to make the pre- announcement in a way that is recognizable to Bella, and Bella to recognize it as a pre- announcement and respond accordingly, e.g., by saying “What?”

Beyond speech acts and pre- announcements, what then constitutes IC? What

do language users need to be able to do to have IC? While it is difficult to create

an inventory of skills, some broad areas of IC have been identified by research

on L2 interactional competence as well as foundational research in conversation analysis (CA; see Wong & Waring, 2010, for an accessible introduction) Relevant aspects of IC include:

• Turn- taking: A conversation is organized by turns, e.g., first Anna speaks, then Bella, then Anna again etc It is rare that there is unintended overlap or long silences How do people know when it is their turn to speak? How do they manage trouble with turn- taking?

• Intersubjectivity: The whole point of interacting is to understand each other

If Bella does not understand what Anna is trying to do by saying “Guess what,” the social action fails Mutual understanding is called “intersubjectivity.”

• Recipient design: People adapt how they talk to their interlocutor, in terms of showing respect, closeness, deference, dominance but also with regard to the

• Repair: Sometimes things go wrong in conversation and need to be fixed

What if there is overlap with Anna and Bella “talking on top of each other,”

or there is a misunderstanding? Repair is about the mechanisms that people use to fix problems in conversation

• Adjacency pairs: Social actions evoke specific responses For example, a request needs a granting or refusing As long as neither has happened, the request remains unresolved So what are the typical ways that social actions are responded to?

• Preference organization: Interaction is where people manage their social relationships but what if they say something that might endanger the rela-tionship, e.g., Anna refuses Bella’s request to lend her a hundred dollars? How

do people organize their talk when social harmony is potentially endangered?

• Topic management: When people talk about a topic, they often move away from the original topic of conversation to another one and then to another one and to another one after that How are these shifts done? How do people manage topics in extended talk?

Many of these concepts can be illustrated with a short excerpt from a tion The sample conversation below is taken from a role play between second lan-guage learners of English at a fairly advanced level (Al- Gahtani & Roever, 2011)

conversa-Alex is working on an assignment late at night but his laptop just died, so he is asking his friend Brad, who lives in the same building, if he can borrow Brad’s laptop Unbeknownst to Alex, Brad is also working on an assignment and needs his laptop, so they will have to negotiate

Note that the transcript follows CA conventions so it is arranged by turns, which are the main unit of analysis in CA and other IC research The numbers

in parentheses (0.4) mean the length of silences, with (.) meaning a short pause

Empty parentheses () mean inaudible, untranscribed talk, and square brackets []

mean overlapping talk For example, the first syllable of “computer” spoken by Alex in line 6 occurs at the same time as Brad’s “oh” in line 7

Alex and Brad request and refusal from Al- Gahtani and Roever (2011)

1 Brad: Hi hi ((name))

2 Alex: Hi how are you ((name))

3 Brad: I’m fine yourself?

4 Alex: Just I have pr ehhh (0.4) a little problem with my

5 assignment tomorrow I have to submit at nine o’clock hh

6 and my [com]puter

7 Brad: [Oh]

8 Alex: is shut down and I have problem with that [Can I]

9 Brad: [Oh]

10 Alex: use your computer?

11 Brad: Yeah sure but actually I have ahh (0.5) an assignment ah

12 (0.4) and I need to submit it at 8 o’clock tomorrow

13 mornink ah (0.5) and I need ehhh (0.4) three or four

14 hours to finish it so maybe after that you can use my

15 computer.

16 Alex: Ok so will you (.) call me when you finish I can come

18 Brad: Yeah [sure but err (0.5)]

19 Alex: [and () your computer?]

20 Brad: I I will (0.7) go to sleep

21 Alex: Yeah?

22 Brad: after it immediately so please don’t be late when I call

you.

23 Alex: No worries

24 Brad: Thank you.

The interactants begin with a greeting sequence in lines 1– 3 before Alex starts giving background to his upcoming request by explaining what the problem is

in lines 4– 8 Alex then makes the request in lines 8– 10, and Brad gives reasons for why he cannot grant the request in lines 11– 14, thereby implying that he is refusing (without explicitly saying so) However, he suggests a possible solution

in lines 14– 15, and the two interactants then discuss the logistical details of this course of action from lines 16– 24 This conversation illustrates well many of the core concepts in conversation analysis

Turn- taking This conversation may seem fairly unremarkable: Alex needs a

laptop and negotiates with Brad to borrow his But this unremarkableness is in itself quite remarkable: how come they have such an orderly back and forth? They only very rarely overlap, and they do not have long silences between turns This is the case because they both follow the rules of turn- taking, which were outlined

in a famous paper by Sacks, Schegloff, and Jefferson (1974) Interactionally petent language users can tell that someone’s turn is coming to an end, and they then prepare to take their turn They do not just randomly jump in and start talking Of course, overlaps do occur, but they are usually in places where one speaker could conceivably finish their turn, and the other speaker anticipated that they would This is probably the case in lines 18 and 19 in Alex and Brad’s conversation

22

16 Alex: Ok so will you (.) call me when you finish I can come

18 Brad: Yeah [sure but err (0.5)]

19 Alex: [and () your computer?]

Brad probably anticipated that Alex was done speaking after line 17 because

“I can come back” is a complete unit He gets the word “yeah” in but then it turns out Alex was not done, which leads to the overlap between Brad “sure but err…” and Alex’s “and () your computer” with the inaudible part likely to

con-Intersubjectivity Another typical characteristic of conversation is that

interactants respond to a turn in a way that fits and “makes sense” They thereby show their understanding of the social action that the turn accomplished This can

be seen clearly in the first three lines Brad opens the conversation in line 1 with a variation on a conventional greeting Alex responds in a way that people normally use when someone greets them: he returns Brad’s greeting This means that Brad’s social action of greeting was successfully accomplished because Alex was able to recognize what Brad was doing They have established “intersubjectivity,” which

is a fundamental part of IC: conversations only work if interactants can interpret each other’s actions, so actions have to be done in such a way that they are easily interpretable

Recipient design Speakers can ensure intersubjectivity by doing their actions

in a way that is easy to understand for the interlocutor Speakers design their talk with the hearer in mind, taking into account

• their relationship with the hearer;

• what knowledge they can assume the hearer to have;

• how to make it easier for the hearer to understand what the core social action is

Designing talk in terms of the speaker– hearer relationship includes the eration of Power, Distance, and Imposition, which were discussed earlier in this chapter

consid-Designing talk in terms of knowledge that hearer and speaker share affects the way a speaker refers to the object of their talk When I talk to a new acquaintance

it would be very strange in the latter case to talk about “my cat,” which might

be taken to imply that there is a secret cat my partner does not know about Drew (2013) shows how interlocutors fine- tune their utterances very precisely in accordance with the recipient’s knowledge status

Recipient design is also operative in the conversation between Alex and Brad One function of Alex’s talk in lines 4– 6 is to provide background information, which ensures that both interactants are “on the same page” as otherwise Brad would have no way of knowing why Alex wants his computer:

4 Alex: Just I have pr ehhh (0.4) a little problem with my

5 assignment tomorrow I have to submit at nine o’clock hh

6 and my [com]puter

Finally, speakers design talk to prepare the hearer for what is coming and making

it easier to understand the main message They might do this by foreshadowing what the main business of the talk is and thereby make it easier for the recipient to follow them For this, they can employ a particular kind of move, called a pre- pre (Schegloff, 1980), which precedes factual background information and orients the listener to the nature of the main business of the talk by foreshadowing it Examples of pre- pres include “Can I ask you a favor?” before a request, “I’ve got something to tell you.” before a bad news telling, or “Guess what I did today!” before a story telling Alex and Brad do not use pre- pres and they are of course entirely optional

Overall, recipient design and the creation of intersubjectivity are absolutely central to interaction because otherwise interlocutors would not make sense to each other, and society as a whole would not work

Conversation opening and closing Interactants do not just start talking

about a topic when they embark on a conversation, and they do not just walk away when they have finished their topical discussion Rather, they begin a con-versation with an opening, and finish it with a closing

Openings commonly consist of a greeting and a “how are you” exchange, as Alex and Brad demonstrate in lines 1– 3

1 Brad: Hi hi ((name))

2 Alex: Hi how are you ((name))

3 Brad: I’m fine yourself?

Openings are purely social talk oriented towards maintaining good social relations, and they can be fairly self- contained, with reciprocal greetings and “how are you” questions before moving on to the main business of the talk However, in Alex and Brad’s conversation, Brad’s “how are you” question in line 3 (“yourself?”)

“See you later then.” – “See you.”, “Ok, talk to you tomorrow.” – “Catch you later.” Wong and Waring (2010) show pre- closing exchanges for various languages

The typical terminal exchange in English is “bye” – “bye” but can also be, for example, “good night” – “night night,” “bye” – “see ya.” The conversation between Alex and Brad does not have a real closing sequence, probably due to the role- played nature of the talk

Adjacency pairs Adjacency pairs are turns that frequently occur together as

a “first- pair part” (FPP) followed by a “second- pair part” (SPP), e.g., greeting–

greeting, request– grant/ refusal, question– response etc

The exchange between Alex and Brad in lines 2 and 3 (“how are you?” – “I’m fine yourself?”) is a typical adjacency pair, consisting of a routine enquiry and a routine answer Importantly, Alex’s request for the laptop in lines 8 and 10 and Brad’s negative answer in line 11 is also an adjacency pair of request– refusal, and it

is what Schegloff (2007) calls the “core adjacency pair,” around which the whole interaction is organized:

8 Alex: is shut down and I have problem with that [Can I]

9 Brad: [Oh]

10 Alex: use your computer?

11 Brad: Yeah sure but actually I have ahh (0.5) an assignment ah

However, just because an FPP has been uttered does not mean that it has to be directly followed by an SPP A sequence of talk may be inserted between the FPP and SPP, known as an “insert expansion.” Instead of replying to Alex’s request in line 11, Brad could have left the request open and launched his own FPP with a question like “How long will you need my computer for?” or “What’s wrong with your computer?” These FPPs would require their own SPP from Alex before Brad then provides a response to Alex’s original request So the structure of adjacency pairs looks something like Figure 2.3

Preference organization The discussion so far has been mostly at the level

of turns and pairs of turns There are structural principles of interaction that are broader than this, and preference organization is such a principle It comes into play particularly where social harmony is threatened The core aspect of preference organization is that actions that promote social harmony and those that threaten

24

it are done differently by interactants Specifically, people tend to delay, minimize, and push into the background social actions that potentially threaten the relation-ship between the interactants, such as requests, refusals, and disagreements This way of making a potentially problematic action less prominent is called

“dispreferred” sequential organization

By contrast, actions that are conducive to good social relations are done quickly, overtly, and immediately, such as agreements, grants of requests, and offers, and their sequential organization is called “preferred.” This does not mean that preferred actions are better than dispreferred ones or more liked by interlocutors These are simply technical terms for two ways of organizing social actions

In the conversation between Alex and Brad, Alex is doing a request and Brad a refusal, so both are organizing their actions as dispreferred Alex’s request is delayed

by the troubles telling in lines 4– 6 and 8, which functions as a prelude to the core request of borrowing the laptop and at the same time delays the request in lines 8– 10

4 Alex: Just I have pr ehhh (0.4) a little problem with my

5 assignment tomorrow I have to submit at nine o’clock hh

6 and my [com]puter

7 Brad: [Oh]

8 Alex: is shut down and I have problem with that [Can I]

9 Brad: [Oh]

10 Alex: use your computer?

The placement of the troubles telling after the greeting and before the core adjacency pair of request– refusal makes it a preliminary to the request, and it is also referred to as a “pre- expansion.” Pre- expansions are very typical in a dispreferred format and frequently include explanations and accounts, as well as hesitation phe-nomena such as cut- offs, false starts, and pauses (as Alex does in line 4)

When speakers get to their core social action, they also usually phrase it in an indirect and mitigated way, as Alex also does by using a conventionally indirect formulation for his request in lines 8– 10

Alex: “Can I borrow your computer?”

Brad: “How long will you need it for?”

Alex: “Maybe a couple of hours.”

Brad: “Yeah sure but…”

FIGURE 2.3 Adjacency pair with insert expansion

26

Dispreference applies just as much to Brad’s refusal In lines 11– 15, Brad responds to Alex’s request negatively, and his response is similarly designed to minimize the threat to social harmony

11 Brad: Yeah sure but actually I have ahh (0.5) an assignment ah

12 (0.4) and I need to submit it at 8 o’clock tomorrow

13 mornink ah (0.5) and I need ehhh (0.4) three or four

14 hours to finish it so maybe after that you can use my

Brad opens with a “token agreement” (“yeah sure but”) in line 11, which is quite a typical way for people to open refusals (Al- Gahtani & Roever, 2018) He then proceeds in lines 11– 14 to lay out the reasons why he cannot lend Alex his computer, punctuated by pauses and hesitation phenomena It is interesting that

he never actually says “no” or “cannot” but implies his inability to help by giving reasons that prevent him from doing so

All these actions by Alex and Brad serve to delay and “hide” the core social actions of request and refusal that might damage the relationship between the interlocutors By contrast, imagine how Brad might have phrased an acceptance

of Alex’s request:

Alex: Can I use your computer?

Brad: Of course, I’ll go get it.

There is no need to delay or mitigate a social action that enhances the social relationship In fact, doing so would be quite strange

Preference organization is anchored by the core adjacency pair (for example, request – refusal or complaint – apology) but it does not end with the core adja-cency pair Once the core adjacency pair has occurred, interactants can conclude the talk, or they can launch into a post- expansion, which is exactly what Alex and Brad do Brad opens the post- expansion in lines 14– 15 with his suggestion that Alex could use the computer later on, Alex accepts the suggestion in line 16, and they then discuss the logistics of the handover The overall structure of a conver-sation like Alex and Brad’s is shown in Figure 2.4

While this sequence is quite common, it is not set in stone In particular, post- expansions might not occur or can be very brief, but preference organization as

an overall structuring principle of conversation is very common, and research into how second language learners do it (or do not do it) has been quite productive as will be discussed below

Topic management A final aspect of IC is topic management How do

interactants open a topic, collaboratively work on a topic, shut down a topic,

or change topics? Looking at Alex and Brad’s conversation, Alex makes his assignment a topic of the talk in lines 4 and 5:

26

4 Alex: Just I have pr ehhh (0.4) a little problem with my

5 assignment tomorrow

This is what Button and Casey (1984) call a “newsworthy event.” Other ways

of establishing topics are via topic elicitators (e.g., “What’s new?” “What have you been up to?”)

Once a topic is set, the interactants then collaborate to extend the topic This can be mostly done by one person, or it can be a back- and- forth effort In Alex and Brad’s conversation, Alex mostly develops his computer problem as the topic

of his troubles talk by himself from lines 4 to 8 but Brad does collaborate by uttering two response tokens “oh” in lines 7 and 9 By providing these tokens, Brad signals his continued listenership and his engagement in Alex’s talk Note that he provides a token that is common with receipt of information (Heritage, 1984) and commiseration, not one of joy and excitement (“wow,” “that’s cool,”

“hilarious”) So while Brad is not talking much during the troubles telling, he is definitely oriented towards Alex’s talk as being about something unpleasant

The excerpt does not show topic extension, where the participants work together to elaborate on a topic, or topic shifts Galaczi (2014) and Wong and Waring (2010) discuss these in detail but it is worth pointing out two types of

Pre- exp

Post-FIGURE 2.4 Conversation between Alex and Brad

28

topic shifts that are particularly relevant to IC: disjunctive and stepwise topic shifts

Disjunctive topic shifts are not tightly connected to the preceding turn, and are often introduced with tokens like “anyway” or “by the way.” To illustrate, imagine that Carla and Doris are talking about a colleague of Carla’s:

1 Carla: He just moved here, and he seems like a pretty

2 nice guy, just quiet.

3 Doris: Right (0.5) Anyway, do you want to hear what I

4 just found out?

5 Carla: What?

6 Doris: It’s about my cousin…

In line 2, Doris shifts the topic away from Carla’s colleague by opening up a new topic, which she signals with “anyway,” and she subsequently starts talking about her cousin

Stepwise topic shifts take an element of the topic that is currently being discussed and use it to move to another topic In the conversation between Carla and Doris, such a shift might look as follows:

1 Carla: He just moved here, and he seems like a pretty

2 nice guy, just quiet.

3 Doris: Haha, it’s the quiet ones you gotta watch out for.

4 Carla: Haha, yes, I guess you’re right.

5 Doris: A cousin of mine was also really quiet, but then

In this version, Doris uses Carla’s description of the new colleague as being

“quiet” as a transition to change the topic to her own cousin, with the link being that he is also quiet

Repair Repair is a crucial mechanism in conversation to overcome problems

and ensure intersubjectivity, i.e., fix a problem with understanding CA distinguishes between self- repair and other- repair Self- repair means that speakers fix problems themselves, and it is by far the most common type of repair There is an instance

of self- repair in Alex’s troubles telling in line 4:

4 Alex: Just I have pr ehhh (0.4) a little problem with my

Alex seems to start to say the word “problem” but then cuts himself off and after a hesitation and pause produces “a little problem.” This illustrates that self- repair does not only fix “mistakes” but any case where a speaker adjusts production mid- stream for better recipient design This kind of self- repair, where the inter-locutor does not intervene, is technically called “self- initiated self- repair.”

Self- repair can also happen after a prompt by the interlocutor, e.g., where they signal that they do not understand something and the original speaker reformulates,

3 A: A theater performance during the day.

In this example, A’s use of the word “matinee” causes trouble for B, who initiates repair by asking what the meaning of matinee is A then completes the repair sequence by glossing the word

The second type of repair, other- repair, is also not present in the conversation between Alex and Brad Other- repair is generally less common than self- repair, and it means that the interlocutor fixes a problem, either initiatively or in response

to the original speaker asking for help An example of the former, which is called

“other- initiated other- repair,” would be:

1 A: So yesterday we went and saw a martinee.

2 B: You mean matinee?

3 A: Yes, right, matinee.

The other case of other- repair, where the original speaker initiates repair,

is called “self- initiated other- repair” and usually consists of the speaker asking for help:

1 A: So yesterday we went and saw a uhm what do you call it when

a performance is during the day?

2 B: A matinee?

3 A: Yes, right, a matinee.

Repair is an important mechanism in interaction because it restores jectivity if there was a problem in comprehension

intersub-2.5.1 Learning L2 Interactional Competence

Interactional competence is a fairly new research area in L2 pragmatics Until Kasper (2006) called for a discursive re- orientation of L2 pragmatics, most research had focused on isolated speech acts However, since then a number of social actions and types of sequential organization have been investigated

In a series of studies, Al- Gahtani and Roever (2012, 2013, 2014a, 2014b, 2015, 2018) investigated preference organization of requests and refusals in English and Arabic as second languages (see also Pekarek Doehler & Pochon- Berger, 2011, for disagreements) They found that as learners’ proficiency increased, they also produced more extended and sophisticated pre- and post- expansions Compare the conversation between Brad and Alex in section 2.5 above with the example

30

below (from Al- Gahtani & Roever, 2011), which shows the beginning of a role- played conversation between two beginning- level learners at low A2 level, Charles and Dave:

Request– refusal at the beginner level (adapted from Al- Gahtani and Roever, 2011)

1 Charles: Hi Dave

3 Charles: Can you give me your laptop?

4 Dave: No because tomorrow I do homework

The lower- level learners do produce a greeting sequence but there is no pre- expansion or any background information before Charles’s request in line 3, whereas Alex in the previous example gave concise but sufficient background information In addition, Dave’s refusal in line 4 is very direct (“no”) whereas Brad in the previous conversation never actually said “no” but implied the refusal conventionally (to follow Taguchi & Yamaguchi, 2019) by giving reasons It is notable that Dave gives a reason for his refusal in line 4 This shows that learners

at this level are capable of these strategies but they do not organize their talk as dispreferred yet

Other research in the area of interactional competence has shown that learners develop in their recipiency (Galaczi, 2014; Ishida, 2011) from apparently passive, unengaged recipients to overtly involved and contributing recipients Their col-laborative development of topics also becomes more fine- tuned with fewer dis-junctive topic shifts and more stepwise transitions (Galaczi, 2014) Studies have also investigated development in turn- taking (Cekaite, 2007), repair (Hellermann,

2009, 2011), story opening (Pekarek Doehler & Berger, 2018), and task openings

in computer- mediated communication (Abe & Roever, 2019)

Outlining a developmental trajectory of IC overall, Pekarek Doehler and Pochon- Berger (2015) describe development as an increasing diversification of linguistic and sequential tools for doing social actions, and deployment of these tools that is increasingly sensitive to the discourse context Whereas learners ini-tially have very little to work with in terms of pragmalinguistic tools and therefore use what they have indiscriminately, they develop an increasingly broader rep-ertoire and use this wider toolset in a more fine- tuned and tailored way This is quite reminiscent of the concept of greater pragmalinguistic and sociopragmatic knowledge and more accurate mapping with increasing proficiency, but the idea

of including sequential organization as part of the toolset goes beyond classic understandings of pragmalinguistics, which tend to be quite proficiency focused

2.5.2 Conclusion

Work in interactional competence has opened new perspectives on learners’ L2 pragmatic development Language users need to be able to engage in extended

30

interactions, and these have specific rules and conventions, which need to be learned In particular, work on topic management and preference organization shows how learners’ ability to engage in extended discourse develops Findings from IC work are highly relevant for informing curriculum design because they show what the developmental steps are and what is possibly learnable and there-fore teachable

2.6 Summary: Speech Acts, Implicature, Routines, and Interactional Competence

This chapter has given a condensed overview of general work on pragmatics and

IC, as well as second language research in speech acts, implicature, routines, and IC.What is the relationship between these four areas? Are they all equally important

or could they be ordered in a taxonomy? Most researchers would probably put speech acts, implicature, and routines in one category under the heading of

“pragmatics,” and interactional competence in a category of its own The reasons for this are mostly due to the different origins of these areas

Pragmatics is anchored in philosophy with work on politeness coming out

of anthropology, and most second language pragmatics researchers come from a general linguistics or second language acquisition background They are comfort-able with isolating research targets of interest (e.g., a particular speech act), using experimental manipulations or survey tools to gather data in order to answer a research question, and then run analyses with statistical and qualitative methods

By contrast, most researchers working in interactional competence come from

a conversation analysis background rooted in sociology, which is critical of the traditional speech act pragmatics paradigm Conversation analysis only works with natural data, and is fundamentally interested in how people organize their con-duct with each other so that they can live in a society Its research approach is very much bottom- up, deriving insights from actual data rather than collecting data to answer a research question, and it has little interest in learning or mental representations of language

While these traditions underlying L2 pragmatics and IC research may appear irreconcilable to the people strongly committed to one or the other, they are actu-ally quite complementary when it comes to teaching and testing Speech acts are

a good example of the connection

In reality, people do not do isolated speech acts but speech acts occur as part

of extended interactions, so a “speech acts in interaction” (Félix- Brasdefer, 2019) perspective or “discursive pragmatics” (Kasper, 2006) perspective has tried to rec-oncile speech acts and extended talk It is important for learners not only to know how a request can be pragmalinguistically formatted and made more or less polite, but also how to sequentially organize a request interaction At the same time, they need to understand how the social context of the talk influences how people interact because it does affect not just speech act formulation but also sequential