Multiple Perspectives on Language Play 9781501503993, 9781501511844

Citation preview

Nancy Bell (Ed.) Multiple Perspectives on Language Play

Language Play and Creativity

Editor Nancy Bell

Volume 1

Multiple Perspectives on Language Play Edited by Nancy Bell

ISBN 978-1-5015-1184-4 e-ISBN (PDF) 978-1-5015-0399-3 e-ISBN (EPUB) 978-1-5015-0396-2 ISSN 2363-7749 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2017 Walter de Gruyter, Inc., Boston/Berlin Typesetting: Compuscript Ltd., Shannon, Ireland Printing and binding: CPI books GmbH, Leck ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

Contents Nancy Bell Introduction 1

01

Neal R. Norrick Language play in conversation

11

Thorsten Huth 2 Playing with turns, playing with action? A social-interactionist perspective 47 Tony Veale 3 The shape of tweets to come: Automating language play in social networks

73

Elizabeth Holt 4 “This system’s so slow”: Negotiating sequences of laughter and laughables in call-centre interaction 93 Olga Zayts and Stephanie Schnurr 5 Laughter as a “serious business”: Clients’ laughter in prenatal screening for Down’s syndrome Michael Haugh 6 Jocular language play, social action and (dis)affiliation in conversational interaction

143

Valeria Sinkeviciute 7 “Everything he says to me it’s like he stabs me in the face”: Frontstage and backstage reactions to teasing 169 Emi Otsuji and Alastair Pennycook 8 Cities, conviviality and double-edged language play David Hann 9 Building rapport and a sense of communal identity through play in a second language classroom 219

199

119

vi 

 Contents

Jet Van Dam and Anne Bannink 10 The first English (EFL) lesson: Initial settings or the emergence of a playful classroom culture 245 Søren W. Eskildsen 11 The emergence of creativity in L2 English: A usage-based case-study 281 Jiyun Kim 12 Teaching language learners how to understand sarcasm in L2 English 317 Natalie Lefkowitz and John S. Hedgcock 13 Anti-language: Linguistic innovation, identity construction, and group affiliation among emerging speech communities Julia McKinney and Elaine W. Chun 14 Celebrations of a satirical song: Ideologies of anti-racism in the media Index

403

377

347

Introduction I originally began my career reading and producing research that fell under the specific term “humor,” but was soon introduced to the broader term “language play,” which was just garnering interest from applied linguists working in second language acquisition as I was finishing my graduate work. Since that time, I have found myself reading work in an increasingly broad swath of scholarship, including various sub-fields within linguistics, as well as research in other disciplines, such as psychology, sociology, anthropology and even, much to my chagrin, neurobiology. Although this has often been challenging, it has convinced me that the study of creative language has much to contribute to many of the big questions we ask and theorize about. It has implications for language change, origins, structure, processing, and practices. It seems that others may be coming to this same conclusion. A search of the Linguistics and Language Behavior Abstracts for “creative” or “creativity” shows a spike of research in the mid-1970s, when language play and development in children became a topic of interest. This was followed by a decline, then, beginning in the 2000s, the term saw renewed interest. This is also when studies that had the terms “humor” and “language play” as key words began to increase, and many of these articles examined the topic with respect to second language (L2) users. During this time period we have seen the publication of a special issue of Applied Linguistics devoted to linguistic creativity (2007) and a special issue of Pragmatics and Cognition that highlighted research into prosody and humor (Attardo, Wagner, and Urios-Aparisi 2011), as well as a number of edited collections and reference works devoted to language play, humor, and linguistic creativity (e.g. Attardo 2014, 2016; Jones 2015, Maybin and Swann 2007, Raskin 2008, Swann and Maybin 2007). While much of the work that will be familiar to linguists has discourse analysis as a method to examine humor and language play, there is in fact a wide variety of approaches to these topics across disciplines that has the potential to inform, enrich, and complement all research in this area. Purposeful examination and consideration of alternative research paradigms is referred to by King and Mackey (2016) as “layering” of research perspectives. Working from the interdisciplinary field of applied linguistics, they describe layering as an approach to research that “demands the explicit consideration of research problems from a range of distinct epistemological perspectives” (p. 210). I see their argument as highly relevant to inquiry into humor and innovative language practices, as well. Such a view is already apparent in the journal Humor, as well as at the conference for the International Society for Humor studies, where lively discussion of humor

DOI 10.1515/9781501503993-001

2 

 Introduction

theory takes place among scholars from diverse fields. However, for many whose focus is on linguistic creativity and humor, specifically, such cross-disciplinary discussion has yet to be taken seriously. With this volume, I hope to broaden that scope and encourage scholars interested in language play and innovative language practices to consider “layering” in their research. Each author in this volume has approached linguistic creativity, humor, and/or language play from a different perspective using different methods. Thus, reading them together will, I hope, prompt readers to seek new and varied resources and perspectives to strengthen their own inquiry, to engage with familiar questions in new ways, and to consider new questions they might ask about playful linguistic practices. While I leave it to individual authors to select and define appropriate terms to characterize their work, let me briefly present my own orientation to each of the terms that commonly appear in this text: creativity, language play, and humor. Each of the three terms can be seen as one side of a continuum. Creative language lies on one side of a continuum with conventional or formulaic language on the other. Similarly, language may be used with varying degrees of playfulness or seriousness; utterances may be very humorous or not at all funny. It is also important to note that each of these terms, while universal in terms of human experience, is also ultimately defined differently across cultural groups and through time (on creativity specifically, see Pope 2005). Thus, what is seen as creative, playful, or amusing in one context, might not be considered so in another. Furthermore, not only are these notions culturally situated, but they are also dynamic. Today’s innovative formulation is tomorrow’s stale, overused formula. Finally, as with all language use, humorous, playful, or creative utterances are co-constructed. While we may isolate production or interpretation of nonserious language through controlled studies, in everyday interaction meanings are negotiated moment-to-moment, through mutual response and adaptation of interlocutors. Any definition of creativity will include innovation as its hallmark. Creative language, specifically, is the innovative use of linguistic forms. Ronald Carter (1999, 2004; Carter and McCarthy 2004) drew our attention to the mundane creativity that he found is quite commonly constructed in everyday talk. Prior to his work, the topic had mainly been investigated in terms of literary texts, with very little attention paid to instances of creativity in daily spoken discourse. Creativity in everyday conversation can involve repetition and patterning within a turn or across speakers, as they echo each other’s words and syntax. It can also involve unusual or marked formulations. Typically, the meaning and functions of these will be recoverable within the context, but it is certainly possible for a formulation to be highly unusual – overly creative – and thus opaque to the hearer.



Introduction 

 3

Language play is closely related to linguistic creativity, as it often involves pattern manipulation and marked linguistic choices. However, unlike creative language use, to play with language typically assumes that the behavior is pleasurable ­(although see Lantolf 1997). David Crystal (1996; 1998) and Guy Cook (1994, 1996, 1997, 2000) are prominent among linguists who helped to draw scholarly attention to this phenomenon and its communicative functions in everyday discourse. Humor, then, is the most specific of the terms, as it involves non-serious and often creative utterances that are intended to elicit a specific type of pleasurable emotion: mirth. Although the precise nature of humor remains up for debate, in general it is considered to involve the juxtaposition of incongruous language, schema, or concepts, which, upon the hearer working out the meaning, should be amusing. For those who focus on the linguistics of humor, Victor Raskin (1985) and Salvatore Attardo (1994) will be familiar names. Conventional forms of language contrast with creative, non-serious forms, and although they do not guarantee it, the conventional forms exist to facilitate clear, easy, and efficient communication. The use of innovative or ambiguous language creates a risk of miscommunication, and also presents a challenge in terms of processing (Giora et al. 2015) so why use this type of language at all? While it is becoming clear that playing with language can bestow cognitive (e.g. Bell 2012, Feingold and Mazzella 1991), social (e.g. Baynham 1996; Fraley and Aron 2004; Holmes 2000; Pomerantz and Bell 2007), and emotional benefits (e.g. Martin 2007; Pogrebin and Poole 1988), the precise nature of these is not yet clear. A fuller understanding of how, when, and under what conditions humans are receptive to language play will help us begin to answer these questions in more detail. An understanding of linguistic creativity and its limits is also necessary for a more complete understanding of structures and processes of language change and variation, interaction, and language itself. The chapters in this volume touch on some of these topics and reflect at least a portion of the diversity of inquiry that can be said to fall under the heading of language play and linguistic creativity. Investigations in this area must contend with certain methodological challenges related to defining and identifying what counts as creative, playful, or humorous language, and the first three chapters illuminate matters of methodology and questions as to what constitutes language play, which are regularly acknowledged to be particularly challenging when grappling with innovative language. In the first chapter, Neal Norrick presents two different conceptualizations of language play, exploring how language may be the object or the medium of play. This delineation has been acknowledged as play in versus play with a language (e.g. Bell 2012; Haugh this volume; Vandergriff and Fuchs 2009, 2012), but here Norrick explores play with language as the object

4 

 Introduction

in terms of language games, specifically. He then examines play as the medium through conversation, here noting as well how the two types can overlap, particularly in interaction. The notion of language games is central to the next chapter, as well, but in this case, influenced by Wittgenstein (1973), Thorsten Huth considers the ways in which interlocutors (here, L2 users) display metalinguistic awareness of the interactional mechanisms that structure talk. In doing so, Huth adds an additional dimension to Norrick’s “language as object” versus “language as medium” of play. He conceptualizes language as action and his is one of several chapters to draw on conversation analysis (CA) in his examination of language play. CA methods highlight the organization of talk-in-interaction and help us to explain how conversation unfolds sequentially. Despite some cultural differences, time and space necessarily constrain how talk is constructed, allowing some basic principles of interaction to be described. In addition to demonstrating the ways that speakers may play with conversational structure, Huth considers whether the extent to which we engage in language play should be considered a marked, or unusual behavior, or a fundamental condition of interaction. Working at the juncture of cognitive and computational linguistics, Tony Veale’s chapter examines the human reception of machine-created tweets. The tweets come from two Twitterbots which generate metaphors by slotting lexical items into templates typically used for metaphors. For example, @Metaphor Minute fills the formula “X is a Y” (e.g. Life is a box of chocolates) without reliance on semantics, resulting in tweets such as “A specialization is a thirstiness.” Its counterpart, @MetaphorMagnet employs more sophisticated methods by relying on meaning and pragmatics. Both yield metaphor-like sequences for which human raters are able to construct (often ironic) interpretations. His work demonstrates not only how formulaic sequences can be the basis for playful language, but also how the familiarity of such sequences facilitates our ability to render even highly innovative language comprehensible. The next two chapters also touch on methods, as laughter is their focus. Too often research on humor has relied on everyday understandings of laughter, which associate it with amusement, when a growing body of work demonstrates that the interactional functions of laughter are much more complex, fulfilling a number of often serious functions in talk. Both chapters use CA, a method whose detailed transcription practices and close attention to the contingent, sequential nature of conversation has already proven very useful in revealing the multifunctional nature of laughter (e.g. Glenn 2003; Glenn and Holt 2013; Haakana 2010). Elizabeth Holt’s data consists of calls to a gas supply company in which laughter is regularly elicited over remarks about the slow computer, despite these utterances not being overtly framed as humorous. This laughter, she argues, is not e­ xpressing



Introduction 

 5

mirth, but constructing affiliation, as interlocutors depart from the typical institutional talk that directly addresses the caller’s problem. Not only does her analysis demonstrate how serious work can be conducted through playful talk, but also how speakers manage that shift into non-serious talk over several turns. Olga Zayts and Stephanie Schnurr also examine institutional talk, but in this case the context is medical, specifically sessions in which a patient receives advice from a nurse about prenatal screening for Down Syndrome. Their prior work (Zayts and Schnurr 2011) has found that in these meetings, nurses may initiate laughter in response to uncomfortable situations, such as when a patient refuses to take what the nurse sees as the best course of action. In this chapter, they consider patient responses to this laughter, and specifically the ways in which a contribution of shared laughter may function either affiliatively or disaffiliatively. It is the latter which is of particular interest, as it is less commonly acknowledged in analyses of interaction involving laughter. Zayts and Schnurr demonstrate how patients may use disaffiliative laughter as a way of negotiating knowledge claims and expressing agency in making medical decisions. In addition, their well-considered use of smiling in their analysis, rather than merely laughter, adds nuance to our understanding of these interactional practices. Although humor and language play often fulfill serious, transactional functions, these have tended to be given somewhat less attention. The next three articles work to redress the balance. When serious functions of humor are discussed, it is most often with reference to teasing, which often minimally conceals a serious effort to regulate the behavior of the target. Michael Haugh, however, turns our attention to jocular language play, a type of talk typically closely associated with enjoyment, entertainment, and affiliation, demonstrating that it, too, can fulfill non-playful functions. Specifically, he examines the ways that disaffiliation can be displayed through such jocular play. Data from the reality game show, Big Brother, inform Valerie Sinkeviciute’s cross-cultural analysis of teasing. The structure of the show involves the presentation of participants’ daily interactions, as well as their retrospective reactions to events in the shared household in a video diary format. Through these varying perspectives, Sinkeviciute was able to examine the actual performance of teases, as well as obtain perceptions of teases after the fact by those involved, both from the video diaries and subsequent frontstage talk about the teasing event. Her careful analysis suggests that the lines between staged, semi-staged, and everyday performances may be blurrier than has been typically portrayed in linguistic scholarship, as unconscious social norms, here, specifically those relating to the reception of teases, permeate all of our interaction. Emi Otsuji and Alastair Pennycook take a second look at the type of playful language that they have observed as a component of multicultural,

6 

 Introduction

urban ­language practices, which they refer to as metrolingualism, in order to examine the ways in which it can encourage conflictual interaction and reinforce normativity, in contrast to the convivial talk that they have focused on in previous research (Otsuji and Pennycook 2010). While the power of humor as a vehicle for effecting change is often touted, a survey of the research suggests that it may instead be much more commonly – and successfully – used to reinforce normative values and behaviors (Bell 2015: 147). Otsuji and Pennycook demonstrate how (often multilingual) banter that is light-hearted on the surface may simultaneously function as a way of upholding racial and gendered divisions in public spaces, and perhaps even providing an opening for animosity to develop. The focus on multilingual language users continues in the next three articles, all of which consider aspects of humor and language play in L2 classrooms. David Hann’s case study followed low proficiency learners of business English to examine how they used humor to create affiliation and community. The intense contact of the educational context he researched was conducive to such work, as the curriculum required students to study together in very small groups for long hours over several days. In this context, he was able to trace the genesis and subsequent development of an in-joke, finding that, despite their limited repertoires of L2 linguistic resources, these L2 users were readily able to joke around in English and to use those jokes to position themselves in particular ways. Jet Van Dam and Anne Bannink similarly examine the ways that affiliation is developed among members of a new class, which in their study is made up of early adolescent learners of English in the Netherlands. They focus on the role of the teacher, specifically, and the ways in which she begins to shape the culture  of the classroom as a place where play is encouraged. Their analysis demonstrates not only how the teacher’s planned activities allowed for students to engage with their L2 in a variety of (playful) ways, but also how her responses to unplanned classroom events, such as student-initiated play and the presence of a wasp in the class, also laid the groundwork for student access to a wider variety of roles, voices, and stances. Research into L2 language play has emphasized experimentation, identity construction, and access to a broader range of L2 voices and linguistic resources; however, little work has been done to establish a clear role for language play in L2 development. Søren Eskildsen’s chapter works to redress that imbalance. With usage-based linguistics as his lens, his longitudinal study examines the processes through which creativity emerges in L2 speech, documenting the ways that learners begin to use existing L2 constructions in new ways and the role that language play may have in facilitating L2 development. Not only does this work highlight socio-cognitive developmental processes, but it also foregrounds the



Introduction 

 7

interplay between recurrent chunks of language and novel formulations within these processes. Humor and language play differ cross-culturally, yet little work has been done to document the specific ways in which this is so, and, further, to do so in ways that distill the work into findings that language instructors can draw on in constructing lessons to address these differences. Jiyun Kim’s work represents a rare exception, as it does both. Using a series of carefully planned activities designed to raise the awareness of the underlying conceptual basis of American sarcasm for Korean learners of English, she demonstrated development in their understanding through pre- and post-tests. Here, however, she focuses mainly on an analysis of the students’ interview data, which illuminates the changes in their thinking. Through the analysis, the reader is also provided with a glimpse into the actual classroom processes and access to sample materials. Thus, the chapter also provides a model that instructors could use in developing their own materials. Although I suggested above that the potential of humor to effect change may be overstated, its use as a means of resistance is better established. Highly creative linguistic practices can not only mark in- and out-group status, but also, and particularly in the case of subordinate or marginalized groups, can function as a way of expressing resistance to and critiquing dominant norms and values. Natalie Lefkowitz and John Hedgcock bring together diverse instances of innovative language practices under the umbrella of anti-language. While at first glance the varieties they analyze together may seem un- or only very loosely related, ­Lefkowitz and Hedgcock provide specific features that characterize anti-­ languages and examine their use in a variety of contexts. A commonality across these is the use of linguistic creativity to mark defiance and resistance to “standard” ­language use. In the final chapter, Julia McKinney and Elaine Chun analyze a popular video response to a “racist rant” posted online by a white California college student. Their work illustrates how humor and language play can express resistance and critique, but also how the humorous challenge to the student’s diatribe against Asians was taken up in diverse ways. In particular, they question the celebratory reactions from the media, which, while lauding the critique, also largely reinforced reductive conceptualizations of racism. This piece illustrates well the challenges of researching language play and linguistic creativity, as well as responses to these types of language practices, in new media. My hope is that the perspectives found here help the reader to see both how these various works all envision and approach the same topic quite differently, and how these differences are complementary. At the same time, it is also worth noting that this volume, while diverse in methods, approaches, and questions,

8 

 Introduction

does not represent the full range of topics that can, should, and have been applied to the study of playful and innovative language use and understanding. Considering the topic through a variety of approaches will help us gain a more complete picture of linguistic creativity, language play, and language itself.

References Attardo, Salvatore. 1994. Linguistic theories of humor. New York: Mouton de Gruyter. Attardo, Salvatore. 2014. Encyclopedia of Humor Studies. Los Angeles: Sage. Attardo, Salvatore. 2016. Routledge handbook of the linguistics of humor. New York: Routledge. Attardo, Salvatore, Manuela Maria Wagner & Eduardo Urios-Aparisi. 2011. Special issue on prosody and humor. Pragmatics and Cognition 19(2). Baynham, Mike. 1996. Humor as an interpersonal resource in adult numeracy classes. Language and Education 10(2–3): 187–200. Bell, Nancy. 2012. Formulaic language, creativity, and language play in a second language. Annual Review of Applied Linguistics 32: 189–205. Bell, Nancy. 2015. We are not amused: Failed humor in interaction. Berlin: De Gruyter Mouton. Carter, Ronald. 1999. Common language: Corpus, creativity and cognition. Language and Literature 8: 195–216. Carter, Ronald. 2004. Language and creativity: The art of common talk. London: Routledge Carter, Ronald & Michael McCarthy. 2004. Talking, creating: Interactional language, creativity, and context. Applied Linguistics 25(1): 62–88. Cook, Guy. 1994. Language play in advertisements: Some implications for applied linguistics. In David Graddol & Joan Swann (eds.), Evaluating language (BAAL Studies in Applied Linguistics 9), 102–116. Clevedon, England: British Association for Applied Linguistics with Multilingual Matters. Cook, Guy. 1996. Language play in English. In Janet Maybin & Neil Mercer (eds.), Using English: From conversation to canon 198–234. London, England: Routledge with Open University. Cook, Guy. 1997. Language play, language learning. ELT Journal 51(3): 224–231. Cook, Guy. 2000. Language play, language learning. Oxford: Oxford University Press. Crystal, David. 1996. Language play and linguistic intervention. Child Language Teaching and Therapy 12(3): 328–344. Crystal, David. 1998. Language play. Chicago: University of Chicago Press. Feingold, Alan & Ronald Mazzella. 1991. Psychometric Intelligence and verbal humor ability, Personality and Individual Differences 12(5): 427–35. Fraley, Barbara & Arthur Aron. 2004. The effect of a shared humorous experience on closeness in initial encounters. Personal Relationships 11: 61–78. Giora, Rachel, Ofer Fein, Nurit Kotler, & Noa Shuval. 2015. Know hope: Metaphor, optimal innovation and pleasure. In Geert Brone, Kurt Feyaerts, & Tony Veale (eds.), Cognitive linguistics and humor research, 129–146. Berlin: Walter de Gruyter. Glenn, Phillip. 2003. Laughter in interaction. Cambridge: Cambridge University Press. Glenn, Phillip & Elizabeth Holt. 2013. Studies of laughter in interaction. London: Bloomsbury. Haakana, Markku. 2010. Laughter and smiling: Notes on co-occurrences. Journal of Pragmatics 45: 1499–1512. Holmes, Janet. 2000. Politeness, power and provocation: How humour functions in the workplace. Discourse Studies 2(2): 159–185.



Introduction 

 9

Jones, Rodney. 2015. Routledge handbook of language and creativity. New York: Routledge. King, Kendall & Alison Mackey. 2016. Research methodology in second language studies: Trends, concerns, and new directions. The Modern Language Journal 100 (Supplement 2016): 209–227. Lantolf, James. 1997. The function of language play in the acquisition of L2 Spanish. In William R. Glass, & Ana Teresa Perez-Leroux (eds.), Contemporary perspectives on the acquisition of Spanish (Vol. 2: Production, processing and comprehension), 3–24. Somerville, MA: Cascadilla Press. Martin, Rod. 2007. The psychology of humor: An integrative approach. Boston: Elsevier Academic Press. Maybin, Janet & Joan Swann. 2007. The art of English: Everyday creativity. Basingstoke, UK: Palgrave Macmillan. Otsuji, Emi & Alastair Pennycook. 2010. Metrolingualism: Fixity, fluidity, and language in flux, International Journal of Multilingualism 7(3): 240–54. Pogrebin, Mark & Eric Poole. 1988. Humor in the briefing room: A study of the strategic uses of humor among police. Journal of Contemporary Ethnography 17(2): 183–210. Pomerantz, Anne & Nancy Bell. 2007. Learning to play, playing to learn: FL learners as multicompetent language users. Applied Linguistics 28(4): 556–578. Pope, Rob. 2005. Creativity: Theory, history, practice. New York: Routledge. Raskin, Victor. 1985. Semantic mechanisms of humor. Dordrecht: Reidel. Raskin, Victor. 2008 The primer of humor research. New York: Mouton de Gruyter. Swann, Joan & Janet Maybin. 2007. Special issue on language creativity in everyday contexts. Applied Linguistics 28(4). Vandergriff, Ilona & Carolin Fuchs. 2009. Does CMC promote language play? Calico Journal 27(1): 26–47. Vandergriff, Ilona & Carolin Fuchs. 2012. Humor support in synchronous computer-mediated classroom discussions. HUMOR: International Journal of Humor Research 25(4): 437–458. Wittgenstein, Ludwig. 1973. Philosophical investigations. Malden, MA: Wiley-Blackwell Publishing. Zayts, Olga & Stephanie Schnurr. 2011. Laughter as medical providers’ resource: Negotiating informed choice in prenatal genetic counseling. Research on Language and Social Interaction 44(1): 1–20.

Neal R. Norrick

1 Language play in conversation Abstract: Language play has been understood in two contrastive ways, giving rise to two separate, but partly complementary research traditions: first, play with language as its object, and second, play with language as its medium. Language is clearly the object of play in games like crosswords, anagrams and punning, where metalingual focus (Jakobson 1960) on the forms of language replaces the coherence of ordinary discourse. This sort of play with language may become serious business, for instance in the creation of concrete poetry and advertising slogans (see Crystal 1998). But language may also be the medium of play in teasing a friend or exchanging embarrassing personal anecdotes, where a play frame (Bateson 1953, Fry 1963) or a non-serious key (Hymes 1972) holds sway. Properly framed, even pointedly negative remarks can come across as playful sarcasm rather than serious aggression (see Boxer and Cortés-Conde 1997). But within the playful interaction, language may retain its literal meaning, as when friends flirtatiously pay each other compliments. The two types fall together when conversationalists non-seriously frame their interaction as play and also transform the means and routines of everyday talk, as when the flirtatious compliments become obviously exaggerated, allusive or punning (see Straehle 1993): Hence the complementary overlap between the two research paradigms on language play. My chapter will illustrate both sorts of language play and their convergence with examples from everyday conversation, and present ways of analyzing such interactions, particularly with regard to their significance for the organization of conversation, for our understanding of the forms of everyday talk, and for interpersonal relationships, especially concerning the interplay of humor and aggression.

1 Introduction Language play has been understood in two contrastive ways: first, language games with language as their object, and second, play with language as its medium. Language is clearly the object of play in games like crosswords, Scrabble and punning, where metalingual focus (Jakobson 1960) on the forms of language replaces the coherence of ordinary discourse. These language games may become serious business, for instance in the creation of concrete poetry and advertising slogans (see Crystal 1998). But language may also be the medium of play in teasing a friend or exchanging embarrassing personal anecdotes, where a play frame (Bateson 1953, Fry 1963) or a non-serious key (Hymes 1972) holds sway. DOI 10.1515/9781501503993-002

12 

 Neal R. Norrick

Tannen speaks of talk framed by the meta-message ‘this is play’, parallel to the joking manner employed to ameliorate face threatening acts (Goffman 1955; cf. framing in Goffman 1974). Properly framed, even pointedly negative remarks can come across as playful sarcasm rather than serious aggression (see Boxer and Cortés-Conde 1997). But within the playful interaction, language may retain its literal meaning, as when friends flirtatiously pay each other compliments. Language play may include language games as such when conversationalists nonseriously frame their interaction as play and also transform the means and routines of everyday talk, as when the flirtatious compliments become obviously exaggerated, allusive or punning (see Straehle 1993). Both language games and the use of language within a play frame are recurrent features of everyday interaction, and complete description of conversation should take both into account. Inasmuch as conversation is the natural home of punning and allusion, we understand these forms of humor only if we can explain their integration into everyday talk and their functioning in it. Consequently, an understanding of language games and language play in everyday conversation is a prerequisite for a complete account of verbal humor. This essay will begin with a survey of language games, then move into a consideration of language in a play frame with examples from genuine conversation to illustrate how the two coalesce and reinforce one another. Language games and humor in everyday talk are based in interaction, demanding participation by the listener/recipient. They often involve gestures, playacting, and imitations of voices and dialects. Conversational language play often works from patterns of spoken interaction, taking the form of proverbial phrases, clichés, one-liners, allusions, stock responses and puns for recurrent situations which we pick up from and weave back into conversation; instead of initiating a conversation with a simple and humorless hello, we may choose from a repository of standard formulas such as we can’t go on meeting like this under appropriate circumstances. Again in taking leave, we may pass over the uncolored goodbye in favor of jocular stock phrases like see you in the funny papers. Besides these formulas for greetings and closings, conversationalists store and recycle humorous phrases tailored to bridge an uncomfortable pause or to wrap up an old topic and to segue into a new one like cat got your tongue? We have special formulas for effecting the transition from a joke or period of non-serious talk into a new topic, namely but seriously, folks and but all kidding aside. Discourse and conversation analysis as well as linguistic pragmatics have contributed much to our understanding of how conversationalists fit jokes and puns into their ongoing talk, and how recipients react to them. Sacks (1973) discusses fortuitous puns in conversation. Sherzer (1978, 1985) extends Sacks in his investigation of both intentional, unnoticed and unintentional, purposeful puns.



Language play in conversation 

 13

Norrick (1993) provided the first full-length study of conversational joking, including punning, teasing and narrative forms. Schegloff (1987) discussed a pervasive ‘joke-first practice’ in conversation, while Straehle (1993) explored teasing in conversation (cf. Boxer and Cortés-Conde 1997), Carter (2000) approached overstatement, punning and so on in conversation in terms of their creativity, Lytra (2009) has described teasing in the classroom, and Tsakona (2013) provides a look at word play in parliamentary speech. By contrast, research on language games proper has usually been based on written data, with few exceptions such as Opie and Opie (1959), the classic folkloristic study of children’s language games, and Cook’s (2000) treatment of children’s language games as competition and collaboration. Norrick (1984) explored formulaic conversational humor pragmatically, and there is significant work on language games from literary sources (e.g. Redfern 1984), specifically riddles (Dienhart 1999), puns in the press (e.g. Bucaria 2004) and on the internet (e.g. Seewoester 2011), while the study of actual practices of language games in natural conversation is largely anecdotal, but see Norrick (1988) on binomials and McCarthy and Carter (2004) and Norrick (2009) on hyperbole. This chapter will illustrate both language games proper and creative language play and their convergence with examples from everyday conversation, and present ways of analyzing such interactions, particularly with regard to their significance for the organization of conversation, for our understanding of the forms of everyday talk, and for interpersonal relationships, especially concerning the interplay of humor and aggression. Section two just below begins with general consideration of language games and illustrates a range of games from various language communities (2.1), then moves into a discussion of individual games involving formulaic speech (2.2), overstatement (2.3), terms of direct address (2.4), pun (2.5), and question-answer sequences (2.6). Section three explores how these and related games are introduced and extended in natural conversation: riddle jokes in conversation (3.1), playing with formulaic language (3.2), playing with terms of address (3.3), overstatement (3.4), and punning (3.5). A final section (4.) concludes this essay. Notes on the corpora from which examples are cited appear at the end of the regular text.

2 Language games: play with language as its object 2.1 Language structures and language games Language games can build on various levels of linguistic structure and representation, including meanings and dialogue, words and grammar, pronunciation

14 

 Neal R. Norrick

and spelling. Spelling is manipulated in constructing palindromes (words and phrases which read the same forward and backwards, as in Madam I am Adam) and anagrams (words and phrases formed by rearranging letters, as in angel and glean), but these games generally are played outside the realm of talk in interaction, whereas our topic here is language play in conversation, so that we focus instead on a spelling-based game in English played in conversational contexts called Ghost. Ghost and the games described just below occur in conversational contexts, though the games, of course, replace topical conversation while they hold sway. In Ghost players take turns adding a letter to a growing word fragment, trying to avoid being the one to complete a valid word. Each fragment must be the beginning of an actual word, and usually some minimum is set on the length of a word that counts, such as three or four letters. The player who completes a word loses the round and earns a ‘letter’, with players being eliminated when they have been given all five letters of the word ‘ghost’. In the version I grew up with, the player who completes a word becomes a third of a ghost, then two thirds, before being eliminated the third time around, and hence the game was called Third of a Ghost. The various versions of Ghost can be played by two or more players of any age and they require no equipment, although it can be played with pencil and paper instead of being spoken aloud; for more see: http://en.wikipedia.org/wiki/Ghost_%28game%29. Tongue twisters apparently provide fun in languages all over the world. The difficulties people experience in pronouncing phrases like Toy boat and Rubber baby buggy bumper resemble problems we all occasionally have in producing speech, and tongue twisters provide a kind of test with a humorous output. Based in human articulation rather than the phonology of any particular language, tongue twisters are recognizable and difficult across languages. Consider the English She sells seashells down by the seashore, where the alternation of ‘s’ and ‘sh’ sounds presents difficulties in articulation by comparison with the Chinese Men-wai you si-shi-si-zhi shi-zi, bu zhi shi si-shi-si-zhi si-shi-zi, hai-shi si-shi-si-zhi shi-shi-zi (at the door there are forty-four lions, (I) don’t know if there are fortyfour dead lions or forty-four stone lions), which adds two more sounds, namely phonetic [z] and [dz] represented by ‘z’ and ‘zh’ respectively. Even with no further knowledge of Chinese, it should be clear that this sentence provides sufficient articulatory difficulties to work as a tongue twister. Various language communities have developed different kinds of games with language, depending in part on features of their respective language systems and spelling conventions. Thus, in Italian a conversational game is based on building chains of words, each of which must begin with the last two letters of the previous word, as in: fisico-cozza-zanzara-razzo-zolla-lascivia-iato-tossica- . . . 



Language play in conversation 

 15

Speakers with advanced skills might even produce chains  with a meaning of sorts, as in: una nave veloce cerca calamari ridicoli (a fast ship searches for ridiculous squids) Just what feature of a language system will be pressed into service for a language game certainly depends on characteristics of the language in question. In German one finds a game called Teekesselchen (tea kettle), generally based only on homophonic or polysemic nouns, played purely verbally and oriented to language (structure). In one of its typical forms, one participant says something like: ‘ich habe ein Teekesselchen und darauf kannst du sitzen’, roughly: ‘I have a tea kettle you can sit on’, where ‘tea kettle’ stands in for the word to be guessed, and perhaps a second clue ‘auf meinem Teekesselchen kannst du auch Geld wechseln’ (in my tea kettle you can also change money), looking for the solution Bank (bench) with the homophone Bank (bank). Germans also engage in a game of creating composite word chains: starting with Haus, players might generate a series like the one below, where each successive noun is formed from the second term of the foregoing composite. Hauswand, Wandschrank, Schranktür, Türschüssel, Schlüsselerlebniss, Erlebnisspark, Parkverbot, Verbotsschild, Schildkröte, . . . There does not seem to be a particular term for this language game. The German Language lends itself to this game, due to its flexibility in combining words: it would be much more difficult and hence much less interesting in English and even worse in, say, Italian. Chinese, however, features a similar language game called Word Dragons, where each new compound must be formed from the second word of the forgoing one, as in: xue-xiao, xiao-gong, gong-zuo, zuo-yong, yong-fa . . . In combining whole independent words, this parallels the German game, rather than the Italian one based on final letters. Various ‘secret languages’ like Pig Latin and Cockney Rhyming Slang represent play with language as its object to different degrees. Of course, there is an element of conspiracy and group identity in using these secret languages as well. Initially someone played with language in the sense of manipulating pronunciation and spelling for Pig Latin and phonology and semantics for Cockney Rhyming Slang to create the patterns, but succeeding generations may have simply taken over alternative lexical items in some cases. In Cockney Rhyming Slang wife was initially replaced by trouble and strife, stairs was replaced by apples and pears and so on for a number of words. Later the second, rhyming word was often

16 

 Neal R. Norrick

dropped, so that trouble alone stood for wife, apples for stairs and so on. To the extent that someone grows up hearing trouble for wife and apples for stairs, using these alternates is a matter of lexical choice but not language play as such. With Pig Latin, however, speakers apply a rule playfully to create and understand new forms, namely: remove the initial consonant (cluster) from every word and add it again at the end plus ‘-ay’; for words beginning with a vowel, simply add ‘-yay’ (in some versions, just ‘-ay’ and in others ‘-way’), hence: eway eakspay ayay ecretsay anguagelay. Creating and interpreting this sentence certainly counts as language play, and becoming proficient at communicating in Pig Latin counts as becoming a skilled player of a game. Other language communities have developed playful secret languages of their own, reflecting certain characteristics of the morphology and spelling of their everyday languages. In Italian one finds a secret language called Dialetto Farfallino or Alfabeto Farfallino (that is the Farfallino dialect or the Farfallino alphabet, where ‘farfallino’ is the diminutive form of the Italian farfalla (‘butterfly’, presumably motivated by the double ‘f’ in the word). In this mode of speaking, vowels are doubled and between them the consonant ‘f’ is inserted, so that, for example, ciao tutti (goodbye everybody) becomes cifiafaofo tufuttifi while perché ridi? (why are you laughing?) becomes peferchefé rifidifi? In Spanish-speaking countries, especially in South America, a similar widely used secret language is called L’idioma de la Pe (language of p) or Jeringonza (also known as Jeringozo or some variant of this, and played in Portuguese as Lingua do Pê in Brazil): despite their differing names all these secret languages involve adding the consonant ‘p’ between doubled vowels. Indeed, doubling vowels and inserting the consonant ‘p’ (less frequently ‘b’ and ‘v’) recur as strategies in secret languages world-wide. See: http://it.wikipedia.org/wiki/Alfabeto_farfallino and for an international list of secret languages: http://en.wikipedia.org/wiki/Language_game. The language games cited instance play with language on various levels. Ghost depends entirely on spelling. The German Teekesselchen is a game based on variable interpretations for spoken/audible realizations corresponding to separate (homophonic, polysemic) meanings, and hence it plays mainly on the semantic level. Cockney Rhyming Slang substitutes one word (or phrase) for another, thus playing primarily with lexical structure, though this play is based initially in phonology/rhyme. Alternatively, Pig Latin substitutes one spoken form for another, and thus works on the phonological level. Both Italian games described are oriented toward spelling rather than toward conversational structure. That Dialetto Farfallino depends on spelling conventions as opposed to conversational practices can be seen in the treatment of ciao: Italian ciao consists in pronunciation of just three phonemes /čao/ but four letters: only an orientation toward the written form could yield the

Language play in conversation 



 17

f­ requent playful form cifiafaofo; based on the pronunciation alone, one would expect /čafaofo/, which indeed occurs in some variants of the game. The purely spelling-based forms differ from what is found in Pig Latin, where a word like honest with a so-called silent ‘h’ yields ‘(h)onestyay’ ignoring the spurious spelled form, as opposed to ‘onest-hay’, which one would derive from reinserting the spelled ‘h’. If, as we must assume, these secret languages were invented and propagated by children, then it is understandable that older children would sometimes have had recourse to spelling as a resource less familiar to younger children, and hence more serviceable for a secret language, eventuating in the more ­spelling-oriented forms of Dialetto Farfallino in Italian. For Pig Latin in English, sound structure rather than spelling counts. Apparently for some users, at least in certain situations, Pig Latin rules apply only to primary words and not to prepositions and articles, making syntax relevant along with phonology. Thus, in the 1994 Disney cartoon film Lion King the bird Zazu says ‘ix-nay on the upid-stay’, to warn Simba and Nala to stop talking about the hyenas, subjecting only nix and stupid to the Pig Latin permutations. Perhaps more significant for present purposes is the assumption on the part of the film writers and producers that a large segment of the viewing public would understand at least this reduced form of Pig Latin. Indeed, one of the hyenas, Banzai, replies ‘Who you callin’ upid-stay?’, suggesting that Pig Latin is so generally accessible that it cannot serve as a truly secret language.

2.2 Language games with formulaic speech Play with language is evident in much of our pre-formed, formulaic speech. Set phrases often incorporate features of play with language. After all, set phrases grow out of conversation, and humor is successful in and memorable from conversation, so it is natural that our stock of formulas includes many humorous items (Röhrich and Mieder 1977, Norrick 2007). Some set phrases are playful in their own right such as ironic and punning similes: As clear as mud To swim like a stone To lie like a rug As nutty as a fruitcake Other set phrases offer opportunities for playful deployment in recurrent contexts. Thus one finds playful set phrases that recur in free conversation in

18 

 Neal R. Norrick

r­esponse to typical situations. Characteristic examples are playful greetings such as: Fancy meeting you here Look what the cat drug in Leave-taking formulas are another obvious area for language play, as the following examples witness. See you in the funny papers Don’t take any wooden nickels Further playful formulas occur in special conversational contexts such as: going through a door together: Age before beauty when some leaves the house door open: Born in a barn? when someone is blocking the light: You make a better door than window when a general silence ensues in a group: Quaker meeting (has begun) Playful pretend misunderstandings of set phrases occur as well. These pretend misunderstandings are themselves set phrases, which act as set responses to other phrases. Typically, the misunderstanding yields a pun, as in the punning misunderstanding of the conversational phrase Surely you jest with I’m not Shirley, but I do jest. Or the punning misunderstanding of a conversational phrase foreign phrase such as C’est la vie with La vie, where the responder pretends to hear French ‘c’est’ as English ‘say’. Language games are thoroughly at home in longer, sentential units like proverbs as well. Playful rhetorical strategies make proverbs more memorable and give them an additional force. Funny proverbs invoke traditional wisdom but also introduce humor with the opportunity to share play and enjoy enhanced rapport. Examples include proverbs with: surprise comparison: Fish and visitors stink after three days

Language play in conversation 



 19

animal imagery: Monkey see, monkey do stereotyping: No tickee, no washee folksy language: Them as has gits overstatement: The grass is always greener on the other side (of the fence) A watched pot never boils paradox: The more things change, the more they stay the same Expect the unexpected and tautology: It ain’t over till it’s over When you’re hot, you’re hot Funny proverbial phrases tend to be constructed around a striking image, as in: Like a cat on a hot tin roof Busy as a one-armed paper hanger Or paradox, as in: Can’t see the forest for the trees Conspicuous by its absence Whatever the source of their humor, these proverbial items can add color and a playful note to conversational interaction. To cite just one example, in the excerpt (from the London-Lund Corpus 1–2) below, two academics are discussing project funding, when one has recourse to the proverb A bird in the hand is worth two in the bush, and though he mentions just the phrases ‘in the hand’ and ‘in the bush’, it serves to metaphorize this stretch of talk. 1 B 2 A

fine. I mean it’s not that I want to. no, no, no, no, no.

20 

 Neal R. Norrick

3 4 B 5 6 7 A 8 B 9 10 11 A 12 B 13 A 14 B

but it seems absolutely fixed now. and I’d rather have it some ten million in the hand. than than than the one million in the bush. um but, yes, yes of course. I think this is highly unlikely. I I’m I’m personally. assuming that a a million in the bush is more likely to happen. yes, yes, literally. yeah. in the bush. ((laughs)) I think I know.

Notice that the second speaker comments on the phrase by saying ‘yes literally’ in line 11, and picks it up himself, repeating ‘in the bush’ in line 13, followed by laughter from the initiator. Thus, language games in the form of proverbs can structurally underpin and energize a spate of talk in interaction.

2.3 Overstatement as a language game Overstatements of all kinds involve language games in that they count as ‘doing nonliteral’. Instead of stating facts, a speaker chooses a clearly exaggerated formulation. In the previous section, we noted that proverbs and proverbial phrases may employ overstatement. Known for their apodictic mode of expression, proverbs often use images with absolute modifiers and adverbs like all, no, always and never, resulting in extreme case formulations (ECFs), as we saw above with: The grass is always greener on the other side (of the fence) A watched pot never boils We understand this last to mean in a semi-literal way that a watched pot seems to take longer to boil and the previous one to mean that distant grass tends to seem greener. According to Gibbs (1994), participants in interaction treat overstatement in proverbs as if they did not make literally true statements. Instead, they signal the speaker’s investment in a point, thus emphasizing or highlighting it, and they are taken as insisting or denying in an exaggerated way. Edwards (2000: 364) says ECFs ‘do not have automatic rhetorical effects’, but his findings in fact jibe rather nicely with the traditional rhetorical analysis of

Language play in conversation 



 21

hyperbole as a figure of amplification or attenuation by which the speaker signals emotional involvement through an exaggerated formulation. Thus, we hear It never rains but it pours not as a literal meteorological observation, but as an emotionally tinged statement about our perception of how troubles seem to multiply or how events generally seem to overwhelm us. By contrast, proverbial phrases and idioms, with their characteristic humorous potential, tend to overstatement in the form of far-fetched imagery: Work one’s fingers to the bone Throw out the baby with the bathwater or exaggerated comparisons like: As fast as lightning (lightning fast) Older than the hills These more or less idiomatic/metaphoric phrases lend themselves to deployment in conversational contexts. Consider just one example. In the excerpt below from the London-Lund Corpus (3–1), slower than a snail overstates the slowness of the pace via comparison with a snail. A . . . and Arabella poor Arabella was lame and walked sB m. A y’ know slower than a snail. so we all had to walk at Arabella’s pace. The passage continues with a further overstated reference via the composite snail’s pace: B A

good Lord. walked the whole length of this vast dining hall. at this snail’s pace, with the president going on ahead.

We see here the tendency for formulaic language to draw attention to itself and to spawn further playful language.

2.4 Language games with terms of direct address Terms of direct address offer a ready resource for conversational play with language, and an interesting area for research on language play. The serious uses of

22 

 Neal R. Norrick

direct address include any reference to a real or imagined listener with a proper or invented term of address, while the playful uses subvert primarily the conditions for proper reference (see Norrick and Bubel 2005). Terms of address or vocatives in English include: Kinship terms like ‘Mom’ and ‘Aunt Helen’; first names like ‘Samantha’ and ‘James’; familiarized first names, including short forms like ‘Sam’, nicknames like ‘Jim’ and ‘Jimmy’; terms of endearment like ‘honey’; terms of respect like ‘madam’; invectives like ‘idiot’; nonce names like ‘Ms. Know-it-all’; and title plus last name as in ‘Professor Jones’. Zwicky (1974) states that direct address serves at least two functions: firstly, calls, designed to catch the addressee’s attention and secondly, addresses, which serve to maintain or emphasize the contact between speaker and addressee. Likewise, Davies (1986) distinguishes an identifying and an expressive function. More recent work by Leech (1999) yields three pragmatic functions (1)  ­summoning attention, (2)  addressee identification, (3) maintaining and reinforcing social relationships. But both (1) summoning attention and (2) addressee identification clearly belong on the ‘attention, identification’ side vis-à-vis the ‘contact, expressive’ side. We can multiply functions on both sides of this basic divide without calling it into question. Furthermore, Wolfson and Manes (1980) contend that terms of address frequently occur in speech acts typical of intimates such as teasing. Straehle (1993) shows that first names and terms of endearment seem to proliferate and to receive extra stress in teasing sequences, and McCarthy and O’Keeffe (2003) see direct address as a characteristic feature of badinage, or what I call banter in Norrick (1993). It is important to stress that both sides are co-present, varying in prominence from one context to the next (cf. Davies 1986; Leech 1999). Terms of address pick out an addressee and they signal a relationship between the speaker and hearer, because the speaker must choose between options like kinship title, first name, nickname etc. Thus, there’s no truly neutral vocative, as Zwicky argues: whatever ‘attention, identification’ work Terms of address do, they also necessarily do ‘contact, expressive’ work in defining and maintaining relationships. To give an impression of the role direct address can assume in language play, consider a couple familiar examples. There are stock humorous conversational moves built around direct address. The basic strategy is a feigned presupposition that the addressee bears the relevant name, as in: No way, Jose Brilliant, Einstein Here ‘Jose’ seems motivated only by its rhyming with ‘way’, whereas ‘Einstein’ refers to a real historical person renowned for his intelligence, so that it

Language play in conversation 



 23

serves as ironic praise. Individual examples may involve (literary) allusions, as in: No shit, Sherlock Elementary, my dear Watson Both of these examples allude to characters in the Sherlock Holmes fiction of Arthur Conan Doyle, and both suggest that the recipient has contributed an obvious remark to the ongoing discussion. Not to forget the classic pair: See you later, alligator; After while, crocodile The initial move in this pair was presumably felt to be funny enough for someone to invent the second part and for it to become popular. Kids still love to cite these leave-taking formulas, and they provided the hook for a popular song in the 1950’s. These strategies feed on the give and take of conversation, as do the games to which we now turn.

2.5 Puns Puns frequently revolve around double meaning for some linguistic unit, either a single word like beat  with two separate senses, namely ‘excel’ and ‘stir ­vigorously’ in the construction: A boiled egg in the morning is impossible to beat or a set phrase like ups and downs with idiomatic and literal meaning potential, as in the classic joke: The elevator business has its ups and downs with either literal reference to the vertical motion of elevators or idiomatically to the oscillations of the business cycle. Many authors would include multiple use of the same material, as in the gag want ad: Lion tamer seeks tamer lion Here the elements of the initial phrase reverse to turn the ad into a jab at the inadequate lion tamer. Of course, the reversal has grammatical consequences: tamer goes from the head noun in a compound with the noun-forming suffix -er to a modifying adjective with the comparative suffix -er, while lion moves from a

24 

 Neal R. Norrick

modifier position in a compound to the head of a nominal. All this helps account for the double-take effect of this construction. Some of our examples so far have contained puns. The proverbial phrases To lie like a rug As nutty as a fruitcake both rely on puns for humor. The verb lie in the former generally imputes ­prevarication of its subject as in Judy lies like a rug, while the second term of the comparison refers to lying in the physical sense of resting in a horizontal, flat position; nutty as a fruitcake applies to subjects considered mentally unstable, but punningly compares them with the plethora of nuts found in a fruitcake. Phrases like this may briefly enliven conversation with their skewed images, but the generally do not attract enough attention to disrupt the flow of serious topical talk. Much more likely to interrupt serious talk is the sort of pretend misunderstanding mentioned above, where the conversational phrase Surely you jest elicits the punning response I’m not Shirley, but I do jest. Such integration into the fabric of everyday talk points in the direction of the sort of interactional punning to be discussed below. So far punning has been seen as a relationship between separate potential meanings for a single unit, but in the section below punning appears as a strategy for complex meaning-making, for derailing and extending topical talk, and for creating a kind of over-arching coherence, for introducing humor, and for expressing an individual sense of humor and sensitivity about (certain) linguistic items, structures and meanings.

2.6 Question and answer sequences Many language games are built on the basic dialogue feature of question and answer. Riddles are attested in Old English (before 800) and riddle jokes continue quite common in English conversation still today. Riddles are constructed on the question-answer format of adjacency pairs in natural conversation. The initiation requires a response, which normally cannot be forthcoming, so that the initial speaker must provide the answer her/himself, revealing the ruse behind the whole pretense: the teller knew the appropriate answer all along and was just testing the recipient, as in a parallel question-response sequence at school, where the teacher knows the answer in advance. In school, it is desirable for the recipient to produce the correct answer, but the riddle joke format projects a recipient who fails to answer (correctly), so that the teller can revel in delivering the answer as a punch line.



Language play in conversation 

 25

Knock-knock jokes represent a similar type of interactional language play, with an instigator and a recipient. These little routines are picked up already by pre-school children, and they continue to offer a source of language play on through adulthood. The standard format has five lines: A: Knock-knock. B: Who’s there? A: Some variable response, often including an apparent proper name. B: A repetition of the previous line followed by who? A: The punch line, which builds on the phrase set up in the previous line. as in: A: Knock-knock. B: Who’s there? A: Toby. B: Toby who? A: To be or not Toby. The whole interactional format may count as play with language as its medium, but the focus here is on the word play usually involved in the relation between the fourth and fifth turns, in the present case the realization that the name Toby can be playfully re-analyzed as ‘to be’, suggesting allusively the well-known quote ‘to be or not to be’ from the so-called second soliloquy in Shakespeare’s Hamlet (III, I, 56 ff.). Notice that, for the purposes of incongruity theories of humor from Freud to the present, there are two opposed meanings for the single stretch of talk [towbiy] – [tuwbiy], yielding ‘appropriate incongruity’ (to use Oring’s 2003 apposite term). A further example will be instructive. Consider: Knock Knock! Who’s there? Nobel. Nobel who? Nobel/no bell, that’s why I knocked! Again there are two meanings for one element: the straightforward family name Nobel, presumably pronounced as in the familiar combination ‘Nobel Prize’, versus the playful reinterpretation ‘no bell’ in the sense of ‘no doorbell’, this second reading predicated on the context of the joke where someone is knocking (as opposed to ringing) in order to obtain entrance. Riddles and knock-knock jokes are built around the question-answer structure of spoken language, and so they are already leading toward the following section, where real spoken ­interaction enters the picture.

26 

 Neal R. Norrick

Maybe even more obviously than riddles and knock-knocks, the game called Questions is based on spoken language constructions rather than lexis, grammar or spelling. Questions is played by participants maintaining a dialogue of asking questions back and forth for as long as possible, without making any declarative statements. Play begins when the first player ‘serves’ by asking a question (often ‘Would you like to play questions?’). The second player must respond with another question (e.g. ‘How do you play that?’). Each player in turn continues the dialogue using only relevant questions. Hesitation, statements and non sequiturs constitute ‘fouls’. Scoring often follows the practices of tennis (adapted from: http://en.wikipedia.org/wiki/Questions_%28game%29). The game is featured prominently in Tom Stoppard’s play (drama) Rosencrantz and Guildenstern are dead, and in an abridged form in the film version of the play, as follows: Ros: Could we play at questions? Gil: What good would that do? Ros: Practice! Gil: Statement! One - love. Ros: Cheating! Gil: How? Ros: I hadn’t started yet. Gil: Statement. Two - love. Ros: Are you counting that? Gil: What? Ros: Are you counting that? Gil: Foul! No repetitions. Three - love. First game to. . . This sort of language play takes us very close to the use of language as the medium of play in conversational interaction as such, the topic to which we turn shortly. Riddle jokes and knock-knocks are often based in punning relations between literal and metaphorical or contextually non-salient interpretations for a single element or stretch of text. We have reviewed other examples of language play based on spelling and the morphological characteristics of a language or the relations between pronunciations, words and their spellings/representations. In each case, language users exploit the opportunities offered by their language systems and spelling (representational) systems for the ­purposes of play. Finally, riddles, knock-knocks and the game of Questions clearly derive from principles of reciprocity and sequentiality inherent in ­everyday talk.



Language play in conversation 

 27

3 Language play in conversation Everyday conversation among friends thrives on banter, teasing and sarcasm. These forms of humor enliven conversation, but they also help us break the ice, fill uncomfortable pauses, negotiate requests for favors and build group solidarity. Teasing and sarcasm can be cast in literal language, thus instancing language in a play frame, but they often spill over into language games as well. In conversation the forms discussed above take on variegated forms and ­combinations. In everyday talk, serious goals and literal meaning sometimes give way to the bantering exchange of playful remarks, where interpersonal rapport holds sway over the topical consistency of communication. As McDowell (1981) says, “language play transforms verbal routines of ordinary discourse, and hence disrupts/replaces the coherence of ordinary talk.” By contrast with serious topical talk, playing with language aims to elicit amusement and good feelings; it carries little weight or consequence for the factual content or goals of talk, but displays wit and an individual personality outside serious topical talk. One effect of engaging in regular word play is to present a general self-image of someone willing to suspend the conversational business at hand for a laugh, of someone attentive to the form of talk and its potential for playful manipulation as well as for communication proper. Furthermore, word play may take on the character of a competitive game for some conversationalists, especially those who maintain a customary joking relationship with certain other individuals or groups (cf. Norrick 1993). Such speakers may keep a kind of unofficial tally of successful joking and repartee with each other, so that for any given conversation or stretch of time each knows who has scored more points in the humor competition, whether they are gaining or losing ground with respect to each other and so on. Like the telling of a narrative joke, a pun constitutes a little showcase with a single conversationalist performing for the others. Further, as Sherzer (1978) points out, puns disrupt topical talk by misconstruing and redirecting it. By contrast with personal anecdotes and jointly produced narratives, puns rank quite high on the scale of aggression and testing. We do not announce puns or preface them as we do anecdotes and canned jokes. We cannot ask if our listener has heard the one about so and so, because puns grow out of the immediate context of talk. They test our attention to this context, and our ability to reanalyze the talk within it rapidly, as well as our ability to ‘take a joke’ in some cases. Of course, we test for more than attention to context and analytic aptitude with our word play. We may play on rather marginal senses of a word, those related to arcane or abstract areas of knowledge. And we may play on more or less covert sexual and religious connotations of words and phrases. So we gather insight into the background knowledge of our listeners, but also into their ­attitudes toward

28 

 Neal R. Norrick

and tolerance about potentially embarrassing and taboo areas. Word play has no monopoly on this sort of insight, of course: Personal anecdotes and canned jokes can and often do touch on sensitive topics like politics, religion and sexuality. But the allusions in word play are camouflaged, easily denied, and hence effective tools of reconnaissance. Moreover, word play may become the primary activity during some stretches of conversation, or it may amount to an undercurrent which foams up to the surface occasionally. For some conversationalists, and especially for some pairs or larger groups in frequent contact, word play provides a recurrent cohesive element of their interactions. It may also serve to get talk started, to fill uncomfortable pauses, and to negotiate topic changes and closing. So playing with language can be an activity some conversationalists engage in within their larger interaction to render it smoother and more entertaining; it can keep them on their respective conversational toes just as it keeps them on the same shared wave length, and thereby enhance rapport between them. In the following, I would like to explore these suggestions about the interpersonal functions of word play.

3.1 Riddle jokes in conversation Riddles discussed above continue quite common in English conversation in the form of riddle jokes still today. They may appear as part of more generalized joke telling or, as in the passage below, one conversationalist may simply introduce a riddle joke, suggested by something in the context. Typically, riddle jokes are so difficult to solve that recipients decline to even guess. In the excerpt below from the Wellington Corpus (DPC014), a wife Vera, her husband Don and their son Paul are planning meals, when Don sets up a joke by saying he does not like quiche at line 3. He proceeds to deliver the riddle joke in lines 6–8, with the help of Vera, acting as the straight man. The son Paul laughs at the joke, and initiates a mock argument about light bulbs by naively asking ‘what’s the light bulb got to do with the dark?” at line 9. 1 Vera: 2 3 Don: 4 Vera: 5 Don: 6 7 Vera:

have the chicken quiche, have the chicken quiche with the rice, crust. I don’t LIKE quiche. don’t you. real men don’t eat it. (3.) how many real men does it take to change a light bulb (4.) tell me.



Language play in conversation 

 29

8 Don: none, real men aren’t afraid of the dark. 9 Paul: ((laughs)) what’s the light bulb got to do with the dark? 10 Don: well if you have a light bulb on, it’s NOT. Inasmuch as they are based on the initiator knowing in advance the proper response, while the recipient remains in the dark till s/he hears the punch line, riddle jokes fall flat when the recipient already knows the solution and supplies it immediately, as in the passage below from the SCoSE. Here college roommates Jacob and Erik are already engaged in non-serious talk, when Erik initiates a joke Jacob rejects as unoriginal, so that Erik seeks refuge in a riddle joke, which he sets up with all too much verbiage. 1 Jacob: 2 3 Erik: 4 5 6 7 8 9 10 Jacob: 11 Erik: 12 13 Jacob: 14 Erik: 15 16 17 18 Jacob: 19 Erik: 20 21 22 23 24

that that’s so unoriginal. what about something else. okay, how about ah, like there’s all this new ah, all this supreme court rulings, and pro-choice and all. I’ve got a pretty a pretty topical joke, pretty uh current joke, pretty important joke for the country to hear. just tell it. huh? you want to hear it? do you care? no. it was in my speech for sophomore year in high school. I remember it still. what do you call a cow that has had an abortion? decaffeinated. decaffeinated, that’s right, no laughter, Jacob knew it. that was the only joke I could think of, off the top of my head.

Riddles build on the question-answer format of adjacency pairs in natural ­conversation. The initiation requires a response, which normally cannot be forthcoming, so that the initial speaker must provide the answer her/himself,

30 

 Neal R. Norrick

revealing the ruse behind the whole pretense: the teller knew the appropriate answer all along and was just testing the recipient, as in a parallel questionresponse sequence at school, where the teacher knows the answer in advance. In school, it is desirable for the recipient to produce the correct answer, but the riddle joke format projects a recipient who fails to answer (correctly). In ­everyday conversation, riddle jokes grow out of topical talk and can lead into further joking as well.

3.2 Playing with formulaic language in talk We looked at playful formulaic language above, but even serious formulaic language lends itself to play, precisely due to its recognizable, hackneyed character. In the passage below (from the SCoSE), we see conversationalists piling up set phrases in exaggerated imitation or parody of a particular register, here current catch phrases about technological innovations. 1 Leona: 2 3 Sally: 4 Leona: 5 6 7 8 Sally: 9 Leona: 10 Sally: 11 Leona: 12 13 14 Sally: 15 Leona:

it’s a good tape recorder. it’s a nice one, huh? it’s a beauty. beauty. yes. top of the line. state of the art. ((laughs)) and the cutting edge. all of the above. all of the above. oh, I love it. can I have it ((laughing)). ((laughs)). it’s beautiful, oh my God.

Top of the line, state of the art, and the cutting edge serve to evoke a register worthy of parody. All of the above does not fit exactly with the others, but it’s from a formal register, and that’s close enough for the purposes of conversational language play. Any fixed expression provides a potential set up for humor, since it sets up expectations on the syntactic, semantic and discourse levels. In another passage



Language play in conversation 

 31

recorded from natural conversation (SCoSE), Jerry alters a traditional phrase to comment humorously on the current topic. Fred: I have a recipe for tofu-potato casserole. Jerry: whew that’s like the bland leading the bland. Fred: ((laughing)) yeah. Here ‘the bland leading the bland’ varies a formulaic pattern, namely ‘the blind leading the blind’, via lexical substitution, and this forces semantic reinterpretation. Fixed expressions offer a natural playground for conversationalists, who can build on them, extend them and subvert them for humorous purposes.

3.3 Playing with Terms of Address Inappropriate application of Terms of Address is word play that can generate genuine incongruence of the kind most recent humor theories recognize as the necessary condition of humor perception (see, e.g. Raskin 1985; Norrick 1993). Simply registering that direct address can occur in non-serious contexts differs from recognizing that it may occur in outright inappropriate ways to create humor (Norrick and Bubel 2009). Thus, interlocutors on a first name basis may suddenly use title and last name or kinship name for special effect, as in the next excerpt, involving two friends of roughly the same age: Anne: make sure you don’t forget your cap and scarf. Bill: yes, Mother. Bill jocularly addresses Anne as ‘Mother’ in response to her maternal-sounding reminder to wear his winter accessories. ‘Mother’ is clearly the wrong term of address for Anne: it mis-identifies the addressee and it misrepresents Bill’s relationship to her. Consider a second example of playfully inappropriate direct address for humor. In the passage below (from the London-Lund Corpus 1–12), Elsie creates a piece of dialogue in which she pretends to be a grandchild addressing her as ‘Granny’ at line 6. This is still a contextually appropriate application of direct address, though somewhat funny even to Elsie, as signaled by the following laugh. But when one of her interlocutors addresses her as ‘Granny’ in the next turn (line 7), it becomes an instance of inappropriate direct address, since he is not her grandchild. Indeed, Calvin’s following statement ‘you`ll have millions of

32 

 Neal R. Norrick

grandchildren’ in line 9 gives the impression that Elsie is not a grandmother yet at all, and this renders the term of address ‘Granny’ all the more incongruous. that they’ll be company for you. 1 Elsie: 2 and whereas other old women might be terribly lonely, 3 you will never be lonely, 4 cos your family will come along, 5 and pe- pay you visits. 6 and say ‘let me do the shopping for you Granny.”(laughs) 7 and suchlike. 8 Calvin: well Granny. 9 you’ll have millions of grandchildren. To this point then, we have reviewed various cases of direct address and terms of address as resources for play in conversation, and seen that direct address may accompany teasing and banter with various effects, and that it may also generate humor through playful mis-application of terms of address in various ways. Let’s inspect one final example of playfully inappropriate direct address, this time one involving a nonce epithet. In the excerpt below from the SCoSE, Ted has expressed reservations about Justine’s collocation ‘make publicity for’, and she agrees that it sounds wrong to her, too. When Ted goes on to voice a particular preference, however, Justine jocularly addresses him as ‘sir’ and appends the comic nonce epithet ‘Mr Prescriptive’. 1 Justine: I say that, 2 and I don’t think it’s good English. I would prefer that you said ‘did’. 3 Ted: 4 Justine: yes SIR, Mister Prescriptive. ((laughing)) ((laughing)) I said I would preFER. 5 Ted: you can have preferences, 6 7 even if you’re Mister DEscriptive. In addressing Ted inappropriately as ‘sir’ in line 4, Justine signals mock respect in a context where we would expect her to reject his prescriptive tendencies (since both are linguists). Just like the kinship name ‘Mother’ in place of the usual first name ‘Anne’ in the previous example, the title ‘sir’ in place of the first name ‘Ted’ playfully mis-identifies the interlocutor, and Ted himself extends the play with names by creating the parallel moniker ‘Mister Descriptive’. Like formulaic language more generally, terms of address present formats easy to twist into playful constructions and ready for parody in everyday ­conversation.

Language play in conversation 



 33

3.4 Overstatement as play in conversation Instead of literal description, a speaker may playfully exaggerate using imagery and extreme vocabulary. In the passage below from the London-Lund Corpus (3–4), the speaker mixes the metaphoric hyperbole of ‘fry’ with the extreme case formulation of ‘never hear a word’. . .. I’ve had exactly the same complaint from anthropology on Rayon Street. they say the noise there is intolerable, and that in the summer of course, it’s a question of fry and and be audible, or um open your windows and never hear a word. ‘Never hear a word’ greatly overstates the difficulty of hearing and comprehending, while ‘fry’ both overstates on the temperature scale and shifts metaphorically from overheated rooms to pans on fires or cooking ranges. Simple exaggerations with extreme expressions like all and never are not necessarily perceived as playful or funny, though a series of extreme formulations can combine to create a humorous scene again from the London-Lund Corpus (3–1), as in: Anne: started with Sherry in the senior commonroom. with the president sitting there like God, you know talking absolute inanity on a very profound level. ((laughs)) and everybody hanging on her every word. Helen: ((laughs)) The comparison like God is intrinsically extreme (as I argue in Norrick 2004). Along with the lexically extreme absolute, everybody and every, and the nice ­juxtaposition of ‘absolute inanity on a very profound level’, it renders the description impossible to take seriously. Again in the passage below, this time from the SCoSE, the simple extreme negative expression ‘You never check the oil’ in line 7 generates no laughter, but the overstatement ‘I check the oil about every say fifteen or twenty MINutes’ in lines 13–14 does elicit laughter. Under any normal circumstances, and especially in this context, the question ‘How’s your oil’ clearly relates to its level rather than its quality, which sets the stage for Ned’s reply in terms of the latter. 1 Frank: 2 3 Ned: 4 Lydia:

how’s your oil? have you checked your oil. my oil is great. ((laughs))

34 

 Neal R. Norrick

5 Frank: 6 Ned: 7 Claire: 8 Ned: 9 Claire: 10 11 12 13 Ned: 14 15 Claire: 16 Brandon: 17 Ned: 18 19 Brandon: 20 Ned: 21 Frank: 22

how is your coolant. ((laughing)) my coolant is great too. you NEver CHECK it. I check them SO often because I’m on his back. ‘don’t you THINK we hafta check the oil and the coolant?’ ((laughing)) ‘NAHneh NAHneh’. ((laughing)) I check the oil, about every say fifteen or twenty MINutes. ((laughing)) I wish you’d show me how to do it. ((laughs)) ((laughs)) ((laughing)) I just checkevery day. ((laughs)) now let me tell you. as recently as four weeks ago . . .

Ned’s punning response leads to laughter, and establishes the sort of play frame in the sense of Bateson (1953) and Fry (1963) that we turn to in earnest in the next section, though it is important to note here already that punning and laughter characteristically signal a potential play frame. Far from taking offense at Ned’s blatant refusal to answer the question he intended, Frank seems willing to act the part of the stooge, and feeds Ned another straight line in line 5 to extend the pun. Then Claire enters the fray with hyperbole, which sets off histrionics between her and Ned built around their differing attitudes on the care and maintenance of the family car. Their parody presentation of a squabbling couple works as a ‘team presentation’ in the sense of Goffman (1959; see Norrick 2004). Once Ned and Brandon run the hyperbolic role play into the ground, Frank introduces a story suggested by the discussion of fluid levels in automobiles in lines 21–22, rather than leaving for the gas station as he claimed he was. Furthermore, the combination of hyperbole with irony often creates humor as well. Consider the following passage from the SCoSE, in which the brothers Brandon and Ned are discussing movies, when Ned invokes irony in the narrow sense of mentioning a proposition opposite what he believes and hopes to convey, as in the analysis by Sperber and Wilson (1981). Brandon: I watched The Fountainhead just a couple weeks ago. with Gary Cooper and Patricia Neal-



Language play in conversation 

 35

Ned: boy I’ll bet THAT’S a great movie. Brandon: ((laughing)) it’s a terrible movie. ((laughs)) Ned: Ned’s straightforwardly ironic ‘I’ll bet that’s a great movie’ elicits more laughter than it seems to deserve, but it has little effect on the conversation otherwise. Brandon reverts immediately to the literal ‘terrible movie’, rather than joining in the ironic approach, and proceeds to his description of the film. The perception of incongruity which forms the basis for humor and laughter according to Bateson (1953), Koestler (1964: bisociation) and others derives from the discrepancy between what is said and what we perceive to be the case, between calling a movie great and believing it was terrible, and hence the simultaneous perception of an object within two contrasting frames of reference or the compatibility with opposed semantic scripts as in Raskin (1985), Attardo and Raskin (1991), Raskin and Attardo (1994). Hyperbole through metaphor is also a form of language play and a way of creating humor, as with the aircraft terms payload and fuselage applied to an insect in the excerpt below from the SCoSE. In this passage, Frank establishes a humorous key with hyperbole, first in his choice of aeronautical vocabulary like take off and payload, then in his grossly exaggerated ‘twenty tons’ in line 4, though no laughter ensues until he commences his claim to have ‘never seen an insect that big, ever’ in lines 7–8. The play frame takes firm hold when Ned and Brandon begin suggesting inappropriate names for the insect. Frank enlists Brandon as a witness to his hyperbole, then extends his aeronautical metaphor, using the specifically aircraft term fuselage twice and wingspan once. Finally, he puts an end to his own extended metaphor and hyperbole in offering an objectively appropriate comparison with a hummingbird. 1 Ned: 2 Frank: 3 4 5 Ned: 6 Frank: 7 8 9 Ned: 10 Frank: 11 Ned: 12

I keep hearing people call them things like hornets. let me tell you. that dude was big enough, to take off with a payload of about twenty tons. well what do you call it? I didn’t know what to call it. I had never seen an insect that big. ever. ((laughs)) the only thing I could think to call it((laughing)) call it, ‘get thee hence.’

36 

 Neal R. Norrick

13 Brandon: 14 Ned: 15 Frank: 16 17 18 19 Ned: 20 Frank: 21 22 Brandon: 23 24 25 Frank: 26 Ned: 27 Frank: 28 29 30 Ned: 31 Frank:

call it sir. ((laughs)) let me tell you what I call it. ‘my God look at that big bug.’ it had a fuselage THAT big. ((holds up fingers)) ((laughs)) yeah. Brandon, I’m not exaggerating, am I? oh no. no. easy. it had a fuselage like THAT. ((laughs)) and a wingspan like THAT. oh man. never seen one like that. so we’re talking primordial here. it was just slightly smaller than a hummingbird.

Of particular interest here is the way participants take turns in making contributions to the humorous framework once established. All three men take a shot at naming the bug and commenting hyperbolically on its size. The conspiracy reaches its high point when Frank appeals to Brandon for testimony that he is not exaggerating, and Brandon goes even one step further in saying ‘oh no. no. easy’. At the end, even Ned kicks in primordial as a show of solidarity. Here the three participants succeed, as a group, in humorously describing a past event only two of them experienced first-hand via extended metaphor, hyperbole and allusion. Language play looms large in the sense of play focused on language, but also in the sense of language being used non-seriously within a play frame. Notice also that Brandon’s ‘call it sir’ echoes a line from an old riddle joke, one version of which goes as follows: Question: What do you call a seven-foot, three-hundred-pound bully armed to the teeth? Answer: Sir. The allusion works on several levels at once in conversational humor. First of all, conversationalists gain prestige any time they can successfully weave an allusion into the fabric of spontaneous conversation. According to Freud (1905), we



Language play in conversation 

 37

derive a childlike pleasure from the serendipity of finding old acquaintances in new environments. So even unfunny allusion can excite a laugh of recognition and a moment of rapport between participants in a conversation, because they can bask in their shared ability to identify the relevant piece of pre-existing text, and allusion counts as word play in any case. Further, reference to a joke makes Brandon’s line a special type of allusion for purposes of conversational humor. Allusion to a text funny in itself has an obvious double humorous potential, first in its actual contribution to the current text, and second by recalling the original text for listeners in the know. Moreover, in the present case, the original joke revolves around a pun. In the question, ‘what do you call’ has the force of ‘how do you designate’, whereas sir in the answer reanalyzes the question as something like ‘how do you address’. So Brandon’s turn also works as a pun itself along with the allusion and word play proper based on the inappropriateness of sir as a class name. Finally, the allusion is especially apt in its reference to a rather large member of the species as well, so that it works on several levels of language play simultaneously. Any unannounced intertextual reference or allusion poses an understanding test for the recipients, and thus it can elicit laughter and enhance rapport in its own right. Further, Brandon’s turn combines allusion with punning and word play, so it should pose a compound test. Interestingly, Ned responds to the test immediately and appreciatively, while Frank fails to react to it, perhaps because he was intent on delivering his own line, though he may simply have been unfamiliar with the joke in question as well. This appreciation for a witty allusion and the differential reaction to it are the sorts of data participants take more or less conscious note of, and they ultimately accrue to the personalities conveyed in humorous conversational language play interaction. So far in this section we have begun to see how some of the resources for language play in the sense of language games enter into conversational contexts where participants in a play frame use language as their medium, we have seen repeatedly that it is difficult to keep the two separate, especially because the playful focus on language tends to open up a play frame, and to lead into more general playful activity.

3. 5 Punning in conversation Playful activity in conversation often begins with a pun or another nonliteral response to foregoing talk. In interactional terms, puns are semantically/pragmatically skewed responses to foregoing talk. The punster constructs an ambivalent

38 

 Neal R. Norrick

utterance with one meaning oriented toward understanding the preceding utterance and a second meaning also fitted to that utterance but based on a contextually inappropriate analysis of it—which is roughly the definition of the pun Sacks (1973: 139) offers. The punning turn consequently clashes with the topic and/or tenor of current conversation, while some linguistic element establishes its claim to a rather tenuous formal relevance. Thus, in the passage below from the SCoSE, involving three undergraduate student assistants in the departmental copying room, talk shifts abruptly from the concrete activity of cutting paper to Arnold’s mental condition via the fortuitous connection between the concrete and mental senses of the phrase off center. Arnold: Judy: Beth: Arnold:

an exact cut. (2.5) oh no. this one is a little off center. that’s because you’re a little off center. ((laughs)) no it’s Tom’s print.

Judy expropriates a phrase Arnold initially introduced with literal reference to some papers he is cutting, then applies it to Arnold himself, so that it takes on its figurative sense. She lets the dual meaning potential of Arnold’s phrase off center entice her into a punning attack on him. This points up one sort of verbal ­aggression often associated with puns: The punster moves into an antagonistic relationship with one or more listeners, thus realigning the participants in the conversation. Judy puts Arnold to a kind of test to see if he can retain his composure and to reply in kind. Judy may routinely launch verbal assaults on her interlocutors or she may limit her attacks to Arnold and a few others with whom she enjoys what we are calling a customary joking relationship. Either way, verbal aggression of this sort reveals something of the personality Judy chooses to express and her relationship with Arnold; at the same time, Arnold’s failure to respond at all provides relevant, though initially ambiguous, social data about him, as discussed above. In interpersonal terms, a punning turn at talk represents a refusal to respond appropriately to the serious, literal sense of the foregoing turn, and an attempt to change the topic through the introduction of potentially objectionable subject matter, over and above whatever aggression the pun itself involves. All this should threaten the face of the first speaker and violate the principles of politeness as described by Lakoff (1973, 1982) and Brown and Levinson (1978), according to which failure to respond appropriately is an imposition on the other participants in talk. Nevertheless, by introducing a play frame and using humor, the punster



Language play in conversation 

 39

goes ‘off record’, so that the first speaker need register neither the imposition nor the face threat. Furthermore, if humor allows us to talk off record, it provides us with a way of accomplishing certain conversation aims without strict accountability. And this helps explain the presence of something as disruptive and potentially uncomfortable as punning in conversation. Punning has redeeming social value to counterbalance its disruption of talk, its near imposition and impoliteness. It provides a way of talking off record, so that we can manipulate the flow of topics, test for ‘relevant social data’ and realign participants in non-confrontational ways. Nor should we overlook the entertainment we derive from punning with its consequent enhancement of rapport from demonstrating shared background knowledge and laughing together, with the competitive game environment it creates, and, hence, the opportunity it affords us for the presentation of a playful witty personality. We should keep an eye out for these functions of punning and word play in the examples to come. The next excerpt (from the LSWEC) exemplifies the trajectory of a pun in the response slot to an ongoing explanation. It shows how the skewed relevance of punning suggests a new direction for talk at odds with what has gone before for only a short stretch of interaction, after which the primary speaker goes on with his serious discussion of the technical term ‘core dump’. 1 Hank: 2 3 4 5 6 7 8 Marv: 9 10 Milt: 11 Marv: 12 13 Larry: 14 Milt: 15 16 17 Hank:

I think it’s like five hundred bites. If you try to access the five hundred and tenth bite, of the five hundred bite array, you’re walking off the end of the array, into what is probably protected domain. and so there is something wrong with your program. so it gets killed. hm, so she’s kind of like the terminator? ((laughs)) [the splatter lady.] [well I don’t know if you’ve ever heard the phrase] core dump? ((laughing)) yeah it sounds incredibly unpleasant. it sounds like what happens if you eat whole apples y’know. ((laughs)). well when you get a segmentation violation, it causes a core dump. so in my little organization, it doesn’t cause a core dump. …

Marv is already in a playful mood before the term ‘core dump’ appears. He’s already suggesting funny comparisons with ‘terminator’ and ‘splatter lady’ in

40 

 Neal R. Norrick

lines 9–10, so there’s a play frame in force for him, though Hank and Milt remain unmoved and serious throughout. Indeed, from their perspective Marv’s levity appears to be constitute an unwelcome intrusion. Still for Marv the term ‘core dump’ presents possibilities for analysis suggestive of something ‘incredibly unpleasant’, namely involving the cores of apples and what happens as they pass through the human body, if eaten, since dump is a coarse word for ‘bowel movement’, and Larry reacts to his suggestion with laughter. This excerpt then points up the limited power puns have to redirect the flow of talk and action. The pun suggests a new topic, and seeks to introduce a play frame which orients participants toward further language play, but to no avail. The cumulative effect of these two factors often alters the alignment of participants as well as changing the topic, but not in this particular case, particularly because Hank and Milt seem intent on serious problem-oriented discussion. Let us next investigate a passage (from the SCoSE) where an initial pun provides the basis for several turns of word play. Here Jason and his wife Margaret are at the home of Trudy and Roger for dinner. The four are seated at the table over dessert and coffee, and Jason is describing a painting. 1 Jason: 2 3 Margaret: 4 Jason: 5 Roger: 6 Jason: 7 Margaret: 8 Roger: 9 Margaret: 10 Roger: 11 Trudy: 12 Margaret: 13 14 15 Roger: 16 17 Jason:

that painting in our livingroom of the BOAT, in theyawl in the channel? Maine? there’s a little boat and an island. y’all in the channel? ((laughs)) yawl. [yawl.] [it’s a] boat, y’all. what are y’all doing in the channel. ((laughs)) I need a little port. ((laughs)) ((laughs)) y’all in that channel ((laughing)). what are y’all in that channel for. I know. sorry. who painted yawl in the channel? it’s a painting by a painter named . . .

This pun picks out a word in the preceding turn for humorous comment, rather than filling the second slot of an adjacency pair. Consequently, it violates our expectations for sequential relevance by forcibly yoking the noun yawl with



Language play in conversation 

 41

the Southern United States personal pronoun y’all based on their fortuitous ­phonological identity. Roger marks his pun with a final laugh in line 5, but Jason treats it as a legitimate failure to understand, repeating the crucial word by way of clarification. Margaret picks up on the pun in line 7, and produces a parallel utterance. Note, however, that Margaret’s ‘it’s a boat, y’all’ cannot count as a pun itself, since it maintains the sense of the previous turn, though it surely illustrates word play and even role play in its attempt to imitate Southern speech patterns. By the same reasoning, we must consider Roger’s repetition of y’all in the next turn word play and role play, but not a pun as such. Then Margaret delivers ‘I  need a little port’ in line 9 in such a manner that it sounds more like a request for a drink than an explanation of someone’s presence in a channel, and that does amount to a pun, before she recycles ‘y’all in the channel’. Margaret and Roger spar verbally as a basic part of their interactional routine, and Margaret seems here to take advantage of the first opportunity she spies to redress the perceived imbalance with a pun of her own. One pun leads to another, not only because it establishes a play frame and prompts participants to work out secondary meanings for apparently appropriate utterances, but also because punning is a competitive game for those conversationalists who engage in it regularly. It is worthy of note that Roger, who instigated the word play, also returns to topical talk with a question directed to the interrupted speaker in line 15. In saying ‘sorry’ Roger apologizes either for the poor quality of the initial pun, for the interruption or perhaps for both. Jason’s silence may have prompted this reaction; after all, he neither laughed nor took a speaking turn after his attempted clarification. Just because word play and humor are technically off record for serious purposes, this does not prevent them from being an unwelcome interruption to the speaker who lost the floor or a shock to the sensibilities of participants who feel the word play and laughter intrude frivolously into sacrosanct matters. And precisely such reactions provide examples of the ‘relevant social data’ discussed above. In any case, the turn segues back into Jason’s description of a painting. So we see that participants in talk may themselves recognize the aggressive and disruptive effect of punning, and attempt to ameliorate it and return to the flow of topical conversation which occasioned the initial pun. Here the painting being described, and, in the previous example, technical considerations involving computers, ultimately override the invitation to language play. To summarize briefly, in punning we analyze a stretch of foregoing speech in two ways, and respond to the interpretation which is incongruent with the current context. This doubly disrupts topical talk, first in suggesting a new topic with spurious claim to relevance, and second in introducing a play frame,

42 

 Neal R. Norrick

which makes more punning and other sorts of word play appropriate. Other ­participants may construct puns of their own, they may manipulate the phrasing of the original pun and they may move into alternate sorts of word play, any or all of which at least postpones and potentially obliterates further topical talk. Punning serves to display wit and test for shared knowledge and attitudes; and they awake a competitive urge in those individuals who make word play a part of their personal conversational style. Besides the relatively mild aggression associated with puns as understanding tests, we observed a more threatening potential for aggression directed at a particular person or group in punning. In addition, interlocutors may take offense at the interruption and/or frivolity of punning in some cases. Still, as a form of word play, punning generally contributes to the enjoyment of the talk exchange and enhances rapport, especially for speakers who actively participate in the ritual antagonist strategy of conversational interaction.

4 Conclusions This essay has investigated language play in conversation from two distinguishable perspectives: first, language games with language as their object, and second, play with language as its medium. We have seen that language is clearly the object of play in conversational games like Ghost and riddles, where metalingual focus on the forms of language replaces the coherence of ordinary discourse. But language may also be the medium of play when friends engage in teasing or conspiring to produce a fanciful image, where a play frame or a nonserious key holds sway. A major focus of this non-serious interaction may be language forms as such, so that language games become an integral part of playing in language, as in several excerpts analyzed here, where the two activities mesh. In real conversation when language becomes the medium of play, participants rarely respect the categories established here for forms of play with ­language: instead the categories may disintegrate, expand and melt into one another. Language games and humorous texts, including canned jokes, serve as templates for parody and playful extension. At the same time, formulaic humor often triggers additional play in language: an overstatement, a simile, even a term of address, an ambiguous construction or a slip of the tongue can provide the basis for teasing, and it may suggest a pun, the pun may lead into teasing, joke telling and so on. Everyday conversation is the home of all varieties of verbal humor and it seethes with ever new opportunities for imaginative play.



Language play in conversation 

 43

Notes on Corpora The data cited in this investigation derive from four different corpora of transcribed English conversation. First, the Saarbrücken Corpus of Spoken English (SCoSE), an extensive collection of audio and video recordings of free conversation and conversational interviews. Notes on the transcription conventions and on participants in the recordings, along with transcribed excerpts from the SCoSE are available online at: http://www.uni-saarland.de/fak4/norrick/sbccn.htm. Second, the London-Lund Corpus of Spoken English (LLC) http://khnt.hit.uib.no/icame/manuals/LONDLUND/INDEX.HTM Third, the Wellington Corpus of Spoken New Zealand English (WSC) http://khnt.hit.uib.no/icame/manuals/wsc/INDEX.HTM Fourth, the Longman Spoken and Written English Corpus (LSWEC) developed for the Longman grammar of spoken and written English, by Douglas Biber, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan (Pearson Education, Harlow, 1999), and the Longman student grammar of spoken and written English, by Douglas Biber, Susan Conrad, and Geoffrey Leech (Pearson Education, Harlow, 2002). I accessed the LSWEC at the Corpus Linguistics Research Program administered by Doug Biber at Northern Arizona University in Flagstaff in the spring of 2007 and again in 2008. I gratefully thank Doug for his help and the opportunity to take advantage of this rich data source. Excerpts from these other corpora have in some cases been partially adapted to the transcriptions conventions of the SCoSE described on the website identified above.

References Attardo, Salvatore. 1994. Linguistic theories of humor. Berlin: Mouton de Gruyter. Attardo, Salvatore & Victor Raskin. 1991. Script theory revis(it)ed: joke similarity and joke representation model. Humor 4. 293–347. Bateson, Gregory. 1953. The position of humor in human communication. Cybernetics, 9th Conference. H. von Foerster (ed.), 1–47. New York: Josiah Macy, Jr Foundation. Bergson, Henri. 1899 [1911]. Laughter: An essay on the meaning of the comic. New York: Macmillan. Boxer, Diana & Florencia Cortés-Conde. 1997. From bonding to biting: Conversational joking and identity display. Journal of Pragmatics 27. 275–294. Brown, Penelope & Stephen Levinson. 1987. Politeness: Some universals in language usage. Cambridge: Cambridge University Press.

44 

 Neal R. Norrick

Bucaria, Chiara. 2004. Lexical and syntactic ambiguity as a source of humor: The case of newspaper headlines. Humor 17. 279–309. Cook, Guy. 2000. Language play, language learning. Oxford: Oxford University Press. Crystal, David. 1998. Language play. London: Penguin. Davies, Eirlys E. 1986. English vocatives: a look at their function and form. Studia Anglistica Posnaniensia 19. 91–106. Dienhart, John M. 1999. A linguistic look at riddles. Journal of Pragmatics 31. 95–125. Dunkling, Leslie. 1990. A dictionary of epithets and terms of address. London: Routledge. Edwards, Derek. 2000. Extreme case formulations: Softeners, investment, and doing nonliteral. Research on Language and Social Interaction 33. 347–373. Freud, Sigmund. 1905 [1960]. Jokes and their relation to the unconscious. New York: Norton. Fry, William F., Jr. 1963. Sweet madness: a study of humor. Palo Alto: Pacific Books. Gibbs, Ray Jr. 1994. The poetics of mind. New York: Cambridge University Press. Goffman, Erving. 1955. On face-work. Psychiatry 18. 213–31. [Reprinted in: E. Goffman. 1967. Interaction ritual. Chicago: Aldine, 5–45.] Goffman, Erving. 1959. The presentation of self in everyday life. Garden City: Anchor Books. Goffman, Erving. 1974. Frame analysis. New York: Harper & Row. Hobbes, Thomas. 1651 [1909]. Leviathan. London: Oxford University Press. Hymes, Dell H. 1972. Models of the interaction of language and social life. In John J. Gumperz and Dell Hymes (eds.), Directions in sociolinguistics, 35–71. New York: Holt, Rinehart and Winston. Jakobson, Roman. 1960. Closing statement: Linguistics and poetics. In Thomas Sebeok (ed.), Style in language, 350–377. Cambridge, MA: MIT Press. Koestler, Arthur. 1964. The act of creation. New York: Macmillan. Lakoff, Robin. 1973. The logic of politeness; or, minding your p’s and q’s. In C. Corum, T. Cedric Smith-Stark & A. Weiser (eds.), Papers from the ninth regional meeting of the Chicago Linguistics Society, 292–305. Chicago: Chicago Linguistics Society. Lakoff, Robin. 1982. Persuasive discourse and ordinary conversation, with examples from advertising. In Deborah Tannen (ed.), Analyzing discourse: Text and talk (Georgetown University Roundtable on Languages and Linguistics 1981), 25–42. Washington, DC: Georgetown University Press. Leech, Geoffrey. 1999. The distribution and function of vocatives in American and British English conversation. In H. Hasselgård & S. Oksefjell (eds.), Out of corpora: Studies in humour of Stig Johansson, 107–118. Amsterdam: Rodopi. Lytra, Vally. 2009. Constructing academic hierarchies: Teasing and identity work among peers at school. Pragmatics 19. 451–468. McCarthy, Michael & Ronald Carter. 2004. ‘There’s millions of them’: Hyperbole in everday conversation. Journal of Pragmatics 36. 149–184. McCarthy, Michael J. & Anne O’Keeffe. 2003. ‘What’s in a name?’: vocatives in casual conversations and radio phone-in calls. In P. Leistyna and C. F. Meyer (eds.), Corpus analysis: Language structure and language use, 153–185. Amsterdam: Rodopi. McDowell, John H. 1981. Speech play. Children’s Literature Association Quarterly 6. 26–28. Mindess, Harvey. 1971. Laughter and liberation. Los Angeles: Nash. Nash, Walter, 1985. The language of humor: style and technique in comic discourse. London: Longman. Norrick, Neal R. 1984. Stock conversational witticisms. Journal of Pragmatics 8. 195–209. Norrick, Neal R. 1988. Binomial meaning in texts. Journal of English Linguistics 21. 72–87.



Language play in conversation 

 45

Norrick, Neal R. 1993. Conversational joking. Bloomington: Indiana University Press. Norrick, Neal R. 2004. Humor, tellability and conarration in conversation. Text 24. 79–111. Norrick, Neal R. 2007. Set phrases and humor. In H. Burger et al. (eds.), Phraseology: An international handbook of contemporary research, vol. 28(1) 302–308. Berlin: de Gruyter. Norrick, Neal R. & Claudia Bubel. 2005. On the pragmatics of direct address in conversation. Lodz Papers in Pragmatics 1. 159–178. Norrick, Neal R. & Claudia Bubel. 2009. Direct address as a resource for humor. In Neal R. Norrick & Delia Chiaro (eds.), Humor in interaction, 29–47. Amsterdam: Benjamins. Opie, Iona & Peter Opie. 1959. The lore and language of schoolchildren. Oxford: Clarendon Press. Oring, Elliott. 2003. Engaging humor. Urbana: University of Illinois Press. Raskin, Victor. 1985. Semantic mechanisms of humor. Dordrecht: Reidel. Raskin, Victor & Salvatore Attardo. 1994. Non-literalness and non-bona-fide in language. Pragmatics & Cognition 2. 31–69. Redfern, Walter. 1984. Puns. Oxford: Blackwell. Röhrich, Lutz & Wolfgang Mieder. 1977. Sprichwort. Stuttgart: Metzler. Sacks, Harvey. 1973. On some puns with some intimations. In Roger W. Shuy (ed.), Report of the twenty-third annual roundtable meeting in linguistics and language studies, 135–144. Washington, DC: Georgetown University Press. Schegloff, Emanuel A. 1987. Some sources of misunderstanding in talk-in-interaction. Linguistics 25. 201–218. Seewoester, Sarah. 2011. The role of syllables and morphemes as mechanisms in humorous pun formation. In Marta Dynel (ed.), The pragmatics of humour across discourse domains, 71–104. Amsterdam: John Benjamins. Sherzer, Joel. 1978. Oh! That’s a pun and I didn’t mean it. Semiotica 22. 335–50. Sherzer, Joel. 1985. Puns and jokes. In Teun A. van Dijk (ed.), Handbook of discourse analysis, vol. 3: Discourse and dialogue, 213–221. Academic Press: London. Sperber, Dan & Deirdre Wilson. 1981. Irony and the use/mention distinction. In Peter Cole (ed.), Radical pragmatics, 295–318. New York: Academic Press. Straehle, Carolyn A. 1993. ‘Samuel?’ ‘Yes, dear?’ Teasing and conversational rapport. In Deborah Tannen (ed.), Framing in discourse, 210–230. Oxford: Oxford University Press. Tannen, Deborah. 1989. Talking voices: Repetition, dialogue, and imagery in conversational discourse. Cambridge: Cambridge University Press. Tsakona, Villy. 2013. Parliamentary punning: Is the opposition more humorous than the ruling party? European Journal of Humor Research 1. 101–111. Wolfson, Nessa & Joan Manes. 1980. ‘Don’t dear me!’ S. McConell-Ginet & R. Boker (eds.), Women and language in literature and society, 79–92. New York: Praeger. Zwicky, Arnold. 1974. Hey, whatsyourname! In M. W. LaGaly, R. A. Fox & A. Bruck (eds.), Papers from the tenth regional meeting of the Chicago Linguistic Society, 787–801. Chicago: Chicago Linguistics Society.

Thorsten Huth

2 Playing with turns, playing with action? A social-interactionist perspective Abstract: This study examines L2 interaction in order to explore the possible scope of a social-interactionist perspective on the notion of language play. By analyzing how second language learners deploy typed turns, sequences thereof, and their underlying preference structure, this study shows that language learners are able to recognize turns and the orderliness of turn-taking on a meta-level by consciously, systematically, and collaboratively manipulating them as objects and patterns. When conceptualizing language as both action and game from the outset, a two-fold notion of play arises from the data, namely play as a marked activity versus play as a basic condition of social interaction.

1 Introduction Humans play, as children and during adulthood. Since play constitutes a patterned, systematic activity, we can play and recognize play when we see it. Language learners, their peers, or their teachers are no exception. As Cook (1997, 2000) observes, language learners routinely engage in language play either directed at discrete linguistic items or through the use of them; clearly, language can be both object and means of language play. Emerging research in applied linguistics indicates that language play is implicated in the social and cognitive processes relevant for language development over time, and that play provides a resource for learners and teachers alike in the process of navigating social spaces in the classroom (Kramsch and Sullivan 1996; Lantolf 1997; Tarone 2000; Broner and Tarone 2001; Belz 2002; Kramsch 2002; Cekaite and Aaronson 2004, 2005; Bell 2005; Kim and Kellogg 2007; Pomerantz and Bell 2007; Bell 2011). This study primarily seeks to illustrate that and how L2 language play extends beyond lexis and morpho-syntax into interactional mechanisms inasmuch as L2 learners may not just treat words or morpho-syntactic structures as manipulable objects from a meta-linguistic perspective, but that they can do the same with the turns they take in interaction. If discrete linguistic units such as lexical items (Bell 2011) can be objects of play in L2 language learning contexts, can learners indeed “play” with the turns they take in conversation? With this question in mind, this project approaches the study of language play by theorizing language as action, since words, phrases, and utterances do things (i.e. achieve, perform, DOI 10.1515/9781501503993-003

48 

 Thorsten Huth

accomplish) once launched by speakers into the social and interactional spaces surrounding them (e.g. Austin 1962; Leech 1983; Levinson 1983; Mey 2001; Searle 1969). Specifically, I apply a social-interactionist framework as put forth by conversation analysis (Atkinson and Heritage 1984; Sacks, Schegloff, and Jefferson 1974; Schegloff 2007; Sidnell and Stivers 2013), seeking to sketch out what such a perspective on language play may offer. Conversation analytic research as a gateway to understanding language play has already been conducted in instructed language learning settings and beyond. Such research suggests that and how the situated construction of identity is implicated in playful behavior of language learners in class (Cekaite and Aronsson 2004); that and how even low proficiency L2 learners regularly initiate playful interaction in class to explore specific language forms and their meaning (Cekaite and Aronsson 2005); that and how the regularity of specific institutional interaction formats can be utilized as a systematic resource by participants to playfully invoke contexts and meanings otherwise not available to the verbal transaction(s) in progress (Haakana and Sorjonen 2011); that and how specific turn-taking mechanisms (such as repair) provide systematic resources for playful pragmatic moves in online communication (Lazaraton 2014); that and how language play in interaction may affect sequencing in interaction (Bushnell 2008); finally, that and how the doing of language play itself is achieved in and through talk-in-interaction (Waring, 2013). This study differs from the above in that it is less interested in how the activity of “doing language play” is socially organized or how interactional mechanisms such as turn-taking or repair provide the means for “doing language play.” Rather, this study is interested in interactional mechanisms as objects of language play, conceptualized here as conscious and joint pattern manipulation on the action level. Can and do L2 learners consciously and collaboratively play with (i.e. manipulate) turn-design, paired action, or preference structure in clearly defined contexts (Sacks [1973] 1987; Bilmes 1988; Schegloff, Jefferson, and Sacks 1977; Pomerantz 1984; Kotthoff 1993; Lazaraton 1997; Pomerantz and Heritage 2013; Bilmes 2014)? Can we in fact isolate instances of L2 learners’ conscious and collaborative pattern manipulation on the action level and the turns that house these actions? In other words, can we observe how L2 learners play with the orderliness of turn-taking itself? The analysis below illustrates that this is indeed the case if we understand the notion of play itself as participants’ joint metalinguistic awareness of conversational mechanisms and principles that reach beyond and across turns. If part of the purpose of this volume is delineating the notion of language play itself, then this social-interactionist perspective on L2 learner conversation suggests that, whatever scope the notion of play (i.e. play in general and language play in particular) may hold in store, it does demonstrably



Playing with turns, playing with action? A social-interactionist perspective 

 49

entail participants’ a) meta-linguistic awareness, b) joint attention to doing interaction, and c) negotiating the relative social accountability of actions currently in progress. As I will argue up front, the notion of language play arising from my data can either be viewed as a marked activity or, alternatively, as the very condition of interaction itself. In order to develop this line of argument, I first describe a view of language (language = action) specifically consistent with the epistemological and methodological framework of Conversation Analysis (CA hereafter). I then review the notion of turns, paired action, and preference structure and discuss their implications for studying verbal behavior consistent with a general notion of language play. I proceed to analyze two data examples of L2 interaction to illustrate how playing with actions on the utterance level may constitute a marked activity while at the same time falling within the parameters of basic principles of talk-in-interaction to which speakers of a given language are necessarily bound.

2 Interaction as a patterned system Once we launch an utterance into the social and interactional space surrounding us, others will read what we say in terms of what it is doing. That is, interlocutors will ascribe some action to that which has been said (“action ascription”, see Levinson 2013). This ascription is how an utterance as mundane as “Good morning.” can be understood not as an assessment of the relative quality of a given morning, but rather as an action (a greeting). Similarly, “I’m sorry” can be understood not exclusively as a referential description of one’s emotional state, uttered, as it were, to the world in general, but rather as a specific action (an apology), directed at some person or state of affairs. When communicating with one another, humans draw from culturally shared knowledge to make sense of what anyone is saying based on a multilayered social, interactional, and ultimately temporally-situated analysis. A major resource in this analysis is considering what that which is said is accomplishing in terms of some discernable action. The field of linguistic pragmatics (Leech 1983; Levinson 1983; Mey 2001) continues to investigate the relevant social and interactional contexts that may affect this process from a wide variety of methodological angles. Prominent frameworks include politeness theory (Brown and Levinson 1987), the notion of “face” across cultures (Scollon and Scollon 2001), maxims of cooperation and politeness (Grice 1975), and sociocultural theory (Lantolf 2000). Differences in relative emphasis and disciplinary utility notwithstanding, all of the above proceed from a number of basic assumptions about how language use in social and cultural contexts works, including

50 

 Thorsten Huth

(1) language use is an orderly, patterned activity; (2) language use is action, language does things; (3) these things are specific primarily social actions rooted in patterned contexts in and beyond sentence-level grammar; (4) such actions or action types may be universal and/or culturally variable; (5) systematic inquiry can reveal what these actions are and how they work within and across language communities. This study applies such a perspective through the specific analytic lens of CA. It focuses on the way turns-at-talk are designed by speakers (“turn design”), the way in which turns may be specifically typed and form discernable pairs (“paired action”), and the way the pairing of turns may display a mutually shared, qualitative relationship (“preference structure/preference organization”). I review these concepts in this order below in order to ascertain the extent to which they provide L1 or L2 speakers with a systematic resource for doing what may arguably amount to a kind of language play grounded in manipulating specific actions via manipulating the turns that house them. One basic insight from CA-informed research is that speakers do not start speaking at random, but rather organize the back and forth of talk through the orderly succession of turns (Sacks, Schegloff, and Jefferson 1974). Speakers take turns so that participants in an interaction may in fact hear and understand what the other is saying (and doing by saying). That is, part of having a conversation includes managing the act of turn-taking itself as speakers gauge whose turn it is now and whose it may be next (“turn allocation,” see Hayashi 2013). Speakers are parsing out not only the lexical and morpho-syntactic contingencies of the utterance-in-progress, but are also appraising precisely when a turn–at–talk may have achieved its pragmatic purposes, i.e. either when it is “done doing”, or when the completion of the current “doing” can relevantly be projected (“transition relevant places,” see Clayman 2013). In this process, speakers rely on turndesign, the systematic construction of a turn and its individual components. Drew (2013: 132) describes the vast scope of linguistic and paralinguistic resources relevant for designing a turn: “…lexis (or words), phonetic and prosodic resources, syntactic, morphological and other grammatical forms, timing (e.g. very slightly delaying a response), laughter and aspiration, gesture and other bodily movements and positions (including eye gaze).” Hence, shaping and constructing a turn is a complex process drawing from the full array of linguistic and paralinguistic resources available to speakers in a given language community. Co-participants are thus by no means passive listeners. On the contrary, each party to an interaction is actively monitoring what is being said and done by others in order to manage whose turn it may be next and to manage



Playing with turns, playing with action? A social-interactionist perspective 

 51

speaker-transitions. As interactants grapple with making their own contributions and monitoring those of their co-participants; as interactants search for words, re-edit, and repair what they are saying in mid-sentence or in mid-turn; and as they overlap in their talk, finish each others’ utterances or interrupt one another, turn-taking in naturally-occurring talk is not the tidy and controlled succession of neatly defined turns we so often encounter in fiction writing or in movie scripts. Rather, turn-taking in naturally occurring conversation looks something like the example below (see Appendix for data transcription key): (1)  (adopted from Drew 2013: 133) 1 Hal: …we stopped at a place called Chil’m 2 Les: .t.hh Oh yes it’s beautiful the:re [isn’ tit= [Isn’t it lovely= 3 Hal: 4 Les: =At leas’ it wa:s, he[he heh uh [A h : we only: we only (.) k- uh 6 Hal: 7 wuh eh w- e- stopped there purely by chance. We were 8 doin’ bed b[breakfasts, [but 9 Les: [.hhhh [.hhh[We you see the fam- the 10 funny thing is my: family come partly from Chil’m We see that real-life, real-time talk is trickier than it’s idealized version we find in fiction novels. Ordinary conversation routinely features overlaps (e.g. lines 02/03, 04/06, 09/10) both terminal and otherwise as speakers negotiate whose turn it is both currently and next. Attempts to take the conversational floor may be successful (line 02), short-lived (line 03), or a turn-in-progress may be abandoned in favor of a speaker transition (lines 08/09). It is central to realize that turn-taking a) is sensitive to temporality in real-time, b) proceeds on the general assumption that one speaker has the conversational floor at a time, and c) provides specific resources for co-participants that aide in negotiating whose turn it is at a given time. All of these affect how interactants understand that which is said and that which is thereby done on a moment-by-moment basis (Sacks, Schegloff, and Jefferson 1974). The turn-taking process usually works out in an orderly fashion because there are systematic structures between and across turns. For example, some turns that we construct are typed. This means that the action (or possible range of actions) that a given turn may accomplish may be specific in the sense that it requires other typed turns to follow, thus forming a sequence of typed turns, a pairing. Sequences of turns (Schegloff 2007) and the actions they may perform occur (minimally) in pairs, with one speaker producing a first pair-part

52 

 Thorsten Huth

(e.g. a question) and a second speaker furnishing a suitable, fitting second pairpart (e.g. an answer), as the example below illustrates: (2) (adopted from Levinson 1983: 312) 01 (ring) 02 A: hello 03 B: hello rob. this is laurie. 04 how’s everything 05 A: pretty good. 06 how ‘bout you. 07 B: jus’ fine. the reason I called was . . . Lines 04/05 and 06/07 contain two question-answer pairs that can be mapped as in Fig. 1 below. 04 05 06 07

A: B: B: A:

Question Answer Question Answer

how’s everything pretty good how ‘bout you jus’ fine

First pair part (FPP) Second pair part (SPP) First pair part (FPP) Second pair part (SPP)

Fig. 1: First and second pair parts in question-answer sequences

The pairing of such two related actions is not just possible, but also often imminently relevant to an interaction as a structural unit; the presence or absence of pairing is socially accountable and thus consequential for the further proceedings of the conversation. This relevance is illustrated in the following example: (3) (adopted from Atkinson and Drew 1979: 52) 01 A: Is there something bothering you or not? 02 (1.0) 03 Yes or no 04 (1.5) 05 Eh? 06 B: No. Speaker A in this example clearly considers an answer as “due,” making the initial absence of an answer accountable in subsequently shorter promptings; these promptings follow the pauses in which speaker B is expected to provide a relevant next action, namely an answer. As we know, the absence of talk is not



Playing with turns, playing with action? A social-interactionist perspective 

 53

tantamount to not communicating; absence can and regularly is understood in terms of providing or withholding relevant “next actions.” The presence or absence of particular turns may thus be accountable and contribute to how interactants understand one another. “Not talking” clearly does not amount to “not doing” and must be understood as an action in and of itself. Therefore, even in response to silence, speakers never launch their utterances into a contextual void. In interaction, there is always some action that precedes and informs a given speaking turn, and that turn will serve as a frame of reference for the next turn to follow (“conditional relevance,” see Schegloff, 1968). Speakers constantly monitor what was just said and thereby done, and they constantly decide what is to come next, suitably, relevantly, and accountably (Sacks 1987). In short, when we study transcripts of real-time, real-life talk, we see that and how individual turns follow each other, that they connect to one another, often in terms of specifically typed turns (turns-at-talk,” see Goffman 1981: 23) and discernable sequences thereof that form distinctive pairs (Schegloff 1968, 2007). The forward trajectory of some first actions (such as questions) can be specific in that a given second action (or relevant range of second actions) may be accountably “due” (such as an answer). However, some pairs in interaction are less intuitive than question-answer. Pairs may provide several alternatives of relevant next actions. Consider invitations, which can either be accepted or refused (Davidson 1984; Drew 1984). Interestingly, in the face of such a choice, interactants will frequently not perceive the possible alternative responses to a given first pair-part to be equivalent. Rather, one second pair-part may be preferred over others. For example, in English, in some contexts, a preferred response to an offer is an acceptance, while a dispreferred response is a rejection (Heritage 1984; Lerner 1996; Schegloff 1990, 2007; Taleghani-Nikazm 2006). How can we tell? A look at how turns are designed reveals that preferred and dispreferred turns display distinct and systematic features. For example, an acceptance is generally a quick and unproblematic matter. Next turns containing an acceptance generally display (1) simple, short acceptances, and (2) no delay in their delivery; in fact, acceptances are often produced latched, or in overlap with, the first pair-part. Consider the following example, adopted from Heritage: (4) (adopted from Heritage 1984: 258) 01 B: Why don’t you come over and see me some[times [I would like to 02 A: The acceptance in (4) is so immediate that it occurs in overlap (line 02) with the invitation in the prior turn (line 01). Rejections, on the other hand, are

54 

 Thorsten Huth

characterized by a distinctly different turn design. In the following example, equally adopted from Heritage (1984), an invitation is refused: (5) (adopted from Heritage 1984: 266) 01 B: uh if you’d care to come over and visit a little while this morning 02 I’ll give you a cup of coffee 03 A: hehh well that’s awfully sweet of you, I don’t think I can make it 04 this morning .hh uhm I’m running an ad in the paper and-and 05 uh I have to stay near the phone Unlike in (4) above, the turn containing the rejection in (5) shows different yet distinct features. First, the response is delayed by an out-breath and furthermore prefaced with “well,” both constituting markers that perceptibly delay the proceedings of the turn. These markers are followed by an appreciation (that’s awfully sweet of you), which has a mitigating function. Only then does A deliver the refusal, which is, however, formulated in an indirect way (I don’t think I can make it this morning). Furthermore, A adds an elaborate account of a reason for  the rejection (.hh uhm I’m running an ad in the paper and-and uh I have to stay near the phone), thus framing the rejection as an inevitable result from circumstances outside of the speakers’ control, arguably lessening the speaker’s accountability. In sum, preferred and dispreferred next turns can be identified by their turn design. Preferred actions tend to be of short duration, are often produced in overlap with preceding or following turns, and are produced seamlessly and quickly. Dispreferred actions, in contrast, are frequently produced with some sort of hesitation, indirectness, mitigation, or with otherwise more elaborate turn structure, where the actual disagreement or rejection is often delayed until the last part of the turn (Davidson 1984; Heritage 1984; Schegloff 1990). The notion of preference structure thus encapsulates speakers’ knowledge of what kind of “next actions” are available, and to which extent they are preferred or dispreferred (Atkinson and Heritage 1984; Bilmes 1988, 2014; Pomerantz and Heritage 2013). However, preference organization does not only apply to responses to actions (i.e. second pair-parts). First actions (i.e. actions that initiate their own forward trajectory, possibly typed, possibly paired) may also be preferred or dispreferred. Empirical evidence shows that and how speakers routinely organize their actions in such a way as to enhance the chance of preferred actions occurring, and to inhibit the occurrence of dispreferred actions and/or responses (Davidson 1984; Heritage 1984; Pomerantz 1984; Taleghani-Nikazm 2006). For example, in English, requests may constitute dispreferred actions in and of themselves, and offers appear to be preferred activities over requests in some



Playing with turns, playing with action? A social-interactionist perspective 

 55

contexts (Heritage 1984; Lerner 1996; Schegloff 1990, 2007; Taleghani-Nikazm 2006). Hence, interactants may well and often do structure their talk and the action(s) the talk performs in such a way as to elicit offers rather than having to issue an outright request. In summary of this brief and necessarily basic review of some fundamental principles of talk-in-interaction, we note the following. If we conceptualize language as action, we see that when speakers say something, they always also do something – that is, speakers accomplish actions. From utterance to utterance and from turn to turn, some actions are typed, some actions project particular next actions (or a range of next actions) to occur. Speakers use the properties of specifically typed turns (e.g. pair-parts), their relevant forward trajectories, the sequences that they may form and the qualitative relationship between the parts of such sequences to one another (i.e. preference structure) as systematic resources to make sense of each other’s conduct. These mechanisms are a matter of intricate timing and specific placement of particular turns-at-talk; they involve speakers’ competencies to design their utterances and turns, and thus rely on fundamental, patterned resources for making sense of talk-in-interaction which rest on social convention in a given language community. The question I pursue below is whether these aspects of interaction can be the object of play by L2 learners, that is, if L2 learners may be able to recognize and manipulate these interactional principles and mechanisms on a meta-level by means of conscious, deliberate, and joint manipulation of specific actions, their placement, and their patterning.

3 Playing with actions in L2 To analyze how L2 learners may engage in joint play with turns, sequences, and preference structure, it is useful to consider what we know about CA-based research within a language, across languages, and in interaction that reaches across cultural and linguistic boundaries. It is well established that some of the interactional mechanisms reviewed above can be both universal and culturally specific. While turns, pairs thereof, and preference structure and its attendant turn-design features are manifest across languages, the specifics of what kinds of actions are dis/preferred in different languages may vary. For example, while in American English, the preferred response to an offer may well be an acceptance, in Farsi the preferred response to an offer is (repeated) rejection, regardless of what the speaker may desire personally (Taleghani-Nikazm 1998). Hence, knowing specifically typed turns (such as offers), their relevant pairing, and the preference structure organizing that pairing (offer-acceptance/rejections

56 

 Thorsten Huth

vs.  offer-repeated rejection prior to acceptance/rejection), in the context of a given language/culture is central for speakers to be successful communicators. For L2 learners, such interactional mechanisms can be tangible learning targets. As has been amply documented, L2 learners routinely (though tacitly) transfer the pragmatics from their L1 when speaking an L2 (pragmatic transfer, see Kasper 1992). This predictably leads to the kind of miscommunication that occurs when speakers of different linguistic and cultural backgrounds talk to one another, irrespective of the common language(s) to which they resort in their talk (Golato 2002; Taleghani-Nikazm 2002). Not surprisingly, language teachers are exploring not only L2 pragmatics in general as a learning target in the L2 classroom (e.g. Rose and Kasper 2001), but also specific mechanisms underlying interaction such as turn-taking, paired action, and preference structure (BarrajaRohan 2011; Betz and Huth 2014). Available studies in second language acquisition demonstrate that even novice language learners are able to learn how interactional resources such as preference structure organize the back-and-forth of talk in clearly defined sequential contexts in the L2, such as in the specific contexts of compliment-responses (Huth 2006; Huth 2007), telephone openings (Huth and Taleghani-Nikazm 2006), or requests (Taleghani-Nikazm and Huth 2010). By now, a host of instructional materials for elementary-level second language teaching has been made available in this vein (Barraja-Rohan and Pritchard 1997; Betz and Huth 2014) and language teachers are recognizing that interactional learning targets are a relevant part of teaching language in the service of increasing crosscultural understanding (Dervin and Liddicoat 2013; Liddicoat and Scarino 2013). However, beyond the finding that L2 learners can learn L2 interactional structures such as typed turns, paired action and preference structure, do we have evidence that these are salient metalinguistic objects for L2 learners available for systematic, joint manipulation? “Playing” in this sense would entail conscious, systematic, and collaborative pattern manipulation by L2 learners. This kind of language play would not be directed at sentence level lexis or grammar, but would rather target specific actions and the turns that house them. This kind of play would not simply reveal language playing directed at lexical or morpho-­ syntactic structures as it happens in the overall matrix of interaction, but rather the act of playing with action itself. In the following analysis, I explore two data examples. Both feature dyadic interaction of American learners of German who speak the target language on the telephone with one another. Both data examples result from the same class-related assignment, which prompted learners to have a conversation on the telephone in German for about 7 minutes (see Huth 2006; Huth 2010). This assignment furthermore provided a secret prompt to one of the participants only, encouraging them to pay the other participant a compliment if and when there



Playing with turns, playing with action? A social-interactionist perspective 

 57

may be a suitable opening in their talk. The first data example features learners of intermediate to advanced proficiency level (enrolled in third and fourth year college-level language classes), whereas the second data example features beginners (enrolled in second semester language classes). In both examples, interactants produce compliments and compliment-responses while speaking German, and they do so after having learned about and practiced such sequences in terms of turn taking, paired action, and preference organization in the L2 (see Huth 2006; Huth, 2010). Interestingly, the intermediate learners appear to transfer the L1 interactional organization of compliments in their talk, though, as I will show, also quite clearly engage in language play in the sense that they manipulate compliments and responses to them to suit their needs in situ. The beginners, on the other hand, engage in both L1 and L2 interactional patterns, and manage the transition from one to the other and back collaboratively, which may be viewed as play as well. In both examples, we will see that participants actively display their meta-pragmatic awareness of what they are doing interactionally. In other words, the data analysis shows that co-participants are clearly conscious of engaging in a verbal activity that has its own interactional organization complete with typed turns, the pairing of specific kinds of actions relevant to the local context, and preference organization specific to the local context (i.e. producing compliments and compliment responses). Consider example (6) below. In this example, we can see the principle of L1 pragmatic transfer at work: while speaking in the L2, these language learners are engaging in a 1-1 transfer of the specific structures (organizing compliments and compliment-responses) in their L1, even though they had previously learned about German ways of doing compliment-responses in interaction. In other words, even though both are using the L2 (German) with one another, they are organizing the back-and-forth of their talk by producing typed turns and pairs thereof, and by orienting to the specific preference structure relevant in their L1 (American English). Participants A and B are both college-aged males in their early 20s. Both belong to the same intramural basketball team and clearly have good rapport outside of class. As they talk on the phone, they are each located in two different university offices and cannot see one another, fulfilling a classrelated oral assignment. Part of this assignment is a secret prompt to one of the two interlocutors to compliment the co-participant if and when possible during their conversation (see Huth 2006). In the overall context of going shopping, the following exchange develops in their interaction: (6) (adopted from Huth 2010: 547) 01 A ((smile voice)) ja ich brauche ein neues kleider ((smile voice)) yes I need a new dresses

58 

 Thorsten Huth

02 B ((smile voice)) oh ja? ((smile voice)) oh yes? 03 A aber du ha (.) du hast but you ha (.) you have 04 ein gute augen fuer die kleider = a good eyes for the dresses = 05 B = oh ja meine augen schaut sehr gut aus = oh yes my eyes look very good 06 mit blaue kleider = with blue dresses = 07 A = ja und du und du hat ein schönes = yes and you and you has a beautiful 08 augen[heheh eyes [heheh 09 B [ohoho oh danke glenn? [ohoho oh thanks glenn? 10 A ja [ja yes [yes [ich liebe dich he[he 11 B [I love you he[he 12 A [hehehe[he [hehehe[he 13 B ((smile voice ends)) [und uh: ja. ((smile voice ends)) [and uh: yes. 14 kennst du das nächstes woche haben wir do you know that next week we have 15 zwei intramural basketballspiele? two intramural basketball games? In line (01), A initiates the topic of needing new clothes, and he does so in what is termed a smile voice, which is immediately picked up by his interlocutor in the next turn (oh ja?, line 02), carried through the ensuing compliment-response sequence, and discontinued only when a topic shift is initiated (line 13). The word “Kleider” in German may have a role to play in this example as it specifically relates to a plural of dresses (clothing for females only) and can easily be confused with the word “Kleidung” (clothing in general) by learners. We cannot determine what the interactants may understand “Kleider” to mean. However, what we can observe is a quick exchange of compliments and responses to them: A compliments B (aber du ha (.) du hast, ein gute augen fuer die kleider, lines 03/04), B agrees with the compliment by producing a compliment to himself,



Playing with turns, playing with action? A social-interactionist perspective 

 59

thus engaging in overt self-praise (= oh ja meine augen schaut sehr gut aus mit blaue kleider, lines 05/06), which is followed by A complimenting B again, this time on the beauty of B’s eyes (= ja und du und du hat ein schönes augen, lines 07/08). All compliments, including the instance of overt self-praise, are accepted (oh ja, line 05; ja, line 07), and the last compliment is followed by a confession of love (ich liebe dich, line 11), directly followed by shared laughter (lines 11/12). Line (13) now features a topic shift, the laughter subsides, and the smile voice is discontinued. The mutual complimenting is over; some other activity with its own forward trajectory has started. This is a clear example of 1-1 pragmatic transfer as the participants make use of the interactional patterns surrounding compliments and compliment-responses that they both share in their L1 (American English). Even though both participants had previously learned about specific German ways of responding to a compliment, i.e. ways of responding to a compliment not found in American English, none of these compliment-response strategies surface in this example. Whatever both interactants are doing as they are moving through this exchange, they are doing it in terms of the preference structure operant for complimentresponses in American English while using German words and grammar talking to one another. Central for the purposes of locating joint and systematic pattern manipulation on the action level in this example is pointing out that participants are engaged in doing something special; this exchange clearly stands out from the surrounding talk, is overtly marked as standing out, and shows that and how participants are in a way bending the preference structure underlying compliment-response sequences in their L1 for the duration of this exchange. How do we know? Compliments can be understood in terms of doing positive assessments (Pomerantz 1978; Golato 2002). Empirical studies in American English show that compliment-response turns display either acceptances (routinely) or rejections (very rarely), and a number of actions that do both – that is, actions that accept and reject the prior compliment at the same time (Golato, 2002; 2005). To explain the latter kind, it has been suggested that they concurrently satisfy two kinds of preference that would appear to collide in one place, namely, the preference for agreeing with a positive assessment and the preference for avoiding self-praise. How does this information affect the analysis of example (6)? We see a quick succession of mutual compliments which are all enthusiastically agreed with; we see an instance of self-praise that is agreed with as well; and finally, we see that the exchange of progressively escalating compliments culminates in a confession of love. All of this is accompanied by a smile voice and ends in shared laughter. Not until a topic shift occurs do the laughter and the smile voice subside. If the preference for avoiding self-praise is operant in American English, then here this

60 

 Thorsten Huth

preference is not only systematically but also consciously and collaboratively suspended for the duration of the smile voice and laughter in this data example. Hence, we see in this example that actions commonly dispreferred in participants’ L1 (agreeing exuberantly with compliments; not sanctioning self-praise) are designed, performed, and treated here as preferred, or at the very least treated as not dispreferred. The effect is certainly laughable; even the most superficial readings of example (6) suggest that two young males are not taking that which they are saying (and thereby doing) at face value, but rather engaging in mock behavior directed at the act of complimenting itself. The objects and virtues on which they compliment each other in concert with the progressive escalation towards physical and sexual innuendo amount to a performance that is treated by the participants as not socially accountable at this point in the interaction. Nonetheless, the laughter, the smile voice, and the placement of turns and the actions that they perform are serious and rather precise business for both interlocutors insofar as the success of this performance as a performance (and not as serious innuendo) crucially depends on mutual alignment of the speakers and on the precise placement of specifically typed turns that engender specific actions in a succession that does not conform to the kind of preference markers one might normatively expect. In other words, both interlocutors have to make a joint switch from generally sanctionable and dispreferred behavior to preferred/not dispreferred and not socially sanctionable behavior by systematically, consciously, and collaboratively marking a shift from serious to mock behavior. Interestingly, this performance is initiated and accompanied by smile voice and laughter; the discussion features the subsequent use and placement of turns that perform actions otherwise not to be expected in these positions (and certainly not in line with the preference structure relevant for the sequence-in-progress), and is closed with a topic shift and the sudden discontinuation of all the features that made this episode stand out from the surrounding talk. Example (6), I argue, is not an instance of language play in which interaction merely surrounds or produces the activity of doing play with language. Rather, we see that interactional mechanisms (i.e. turns, sequences, and preference structure) become the metalinguistic object of play themselves as participants manipulate them to suit their social and interactional needs. We might view their actions as bending the preference structure of compliments and compliment-responses for their immediate local purposes, namely, making a mockery of the act of complimenting one another. In order to do that, both interactants consciously, systematically, and collaboratively produce a succession of specifically typed turns that, in this particular constellation, mark the activity in progress as non-serious and as not socially sanctionable. The notion of play operant in this example, then, arises from the systematic properties of the interaction in progress and L2



Playing with turns, playing with action? A social-interactionist perspective 

 61

learners’ conscious and collaborative departure from otherwise established normative orientations in a specific local and sequential context. The production and placement of compliment and compliment-response turns as well as navigating their underlying preference structure has become an object of conscious, systematic, and collaborative pattern manipulation. The previous example showed how participants organize their talk in terms of the interaction mechanisms operant in their L1 even though they speak in an L2. In contrast, the next example and its analysis move beyond the interaction patterns of language learners’ L1 and include instances of L2 learners engaging in a compliment-response sequence that is specifically German. That is, interactants in (7) below collaboratively engage in a compliment-response exchange that is not found in American English, and one that was previously taught and practice in class as specifically German. It may be noted that, whatever participants may have learned about L1 or L2 compliment sequences and preference structure in class previously, when they compliment one another on the phone, they now have a choice: even though they may speak in the L2 on the phone, they may treat the activity in progress (i.e. complimenting and responding to a compliment) in terms of their L1 or in terms of their L2 preference organization. They have learned and practiced both. The question is now one of spontaneous choice: how do participants negotiate the fact that they have two alternate ways (L1 and L2) of “doing compliments” at their disposal on the level of how turns and the actions they house follow one another systematically. Jeremy and Lauren are second-semester novice learners of German about 20 years of age. They produce the talk from which (7) below is excerpted as part of the same assignment as participants A and B in (6) above: they are located in different university offices and are talking on the phone to fulfill a class-related oral assignment. Jeremy received a secret prompt to compliment Lauren if and where suitable during their telephone conversation (see Huth, 2006). Similar to participants A and B in example (6) above, Jeremy and Lauren mark the activity of complimenting one another as something special, as something that stands out of the surrounding talk, both at the beginning and at the end. (7) (adopted from Huth 2006: 2037) 01 jer: .hhh uhm du habst uh eine uh schönes hemd .hhh uhm you have uh a uh beautiful shirt ( . ) diese ( . ) heute 02 ( . ) this (.) today 03 (1.0) 04 lau: uh uh-

62 

 Thorsten Huth

05 jer: or uh du du habst uh eine ( . ) schönes hemd heute. or uh you you have uh a ( . ) beautiful shirt today. 06 lau:→ oh danke or uhm! uh schone ja? he [he oh thank you or uhm! uh beautiful yes? he [he [ja sehr 07 jer: [yes very 08 scho- ja sehr s- schon beau- yes very b- beautiful 09 lau: uh= uh= 10 jer: =bitte. =you are welcome. 11 lau: es freut mich (.) uhm that’s nice ( . ) uhm 12 (2.0) uh wie? uhm wie machst du an am an die teste? 13 uh how? uhm how do you do on on on the test? In lines 01/02, Jeremy produces the initial compliment turn (.hhh uhm du habst uh eine uh schönes hemd), which may pose trouble for Lauren as there is a long pause (1 second, line 03) and Lauren produces uh- (line 04). Jeremy repeats the compliment (or uh du du habst uh eine ( . ) schönes hemd heute., line 05). In line 06, Lauren now does something very interesting; she a) provides a complimentresponse (oh danke), b) initiates repair as she code-switches into English (or uhm!), c) issues yet another though very different compliment-response (uh schone ja?), and then d) produces laughter (hehe). Hence, Lauren replaces her initial compliment-response (the common American English appreciation token ‘thank you’ translated into German) in real time with a different one. The replacement features a same-strength adjective (schone) and what is called response-pursuit by means of a tag question (ja?), a move that seeks further response to that which has been said. In line with Lauren’s response-pursuit, Jeremy agrees (scho- ja sehr s- schon, line 08) in the next turn. The sequence then is closed (lines 09-11) and is bounded by a long pause (2 seconds, line 12) and an unmistakable topic shift (line 13). We see how a compliment is met with initial L1 pragmatic transfer as Lauren, after some difficulties, provides an appropriate and highly frequent complimentresponse in American English, translating as she does the American English appreciation token “thank you” into German (danke). However, she then replaces that action with another: she provides a second compliment-response which is not found in American English interactional data, but frequently found in German interactional data. This kind of sequence looks like the following



Playing with turns, playing with action? A social-interactionist perspective 

 63

(as mapped out in Fig. 2 below): A compliments B, B produces a same-strength adjective + tag (=response pursuit), A compliments again, and B accepts the compliment. A: übrigens (.) das fleisch exzel[lent by the way ( . ) the meat exce[llent B: [super ne? [super, right? A: exzellent excellent B: joa Yeah

A compliments B B same-strength adjective + tag A compliments again B accepts compliment

Fig. 2: German compliment-response with response-pursuit (Golato 2002: 558)

Data example (7), then, illustrates how Lauren and Jeremy, both native speakers of English, both intermediate learners of German, are producing and reacting to a compliment in ways consistent with a distinct, German interaction pattern due to prior classroom intervention (see Huth 2006 for details). In order to do this successfully, both interlocutors have to make a conscious, systematic, and most importantly, collaborative effort to “embed” the German structure in their talk. Lauren’s repair initiation (or uhm!) in English and her repair operation (uh schone ja?) clearly mark and thus signal to Jeremy that she is doing a pragmatic repair as she is replacing the immediately preceding action (one compliment-response) with another (a second and different compliment-response). Hence, both participants are successfully manipulating the sequential organization of their talk on a meta-pragmatic level. Both Lauren’s repair operation and her laughter can be seen in exactly that light, namely as meta-pragmatic markers. Every repair operation requires a repairable (Schegloff, Jefferson, and Sacks 1977), and here it is clearly the type of action performed in the immediately preceding context – the type of compliment-response turn provided. Furthermore, each instance of laughter requires a potential laughable, i.e. that which is the object of, or point of reference for, the manifest laughter (Glenn 2003). Here, the “laughable” is likely the repair operation on the action level (replacing one typed turn with another). Hence, the manifest laughter here is unlike the laughter we encounter in (6) above. In (6), one might say participants are merely (though jointly) bending and/or breaking the known and shared normative expectations of how compliments are interactionally organized in their L1, and the laughter (as well as the smile voice) provide the requisite markers to suspend the activity’s social accountability for its duration. In example (7), in contrast, Lauren uses laughter to mark her repair

64 

 Thorsten Huth

operation as laughable, hence her laughter marks the subsequent reissuing of a different kind of compliment-response (a different type of action) in the specific position into which it is produced, complete with its own particular forward trajectory of next turns in German. Interestingly, both Jeremy and Lauren proceed to produce exactly these, carry them through, and bring them to a sequential close with ease and in joint orientation. Therefore, example (7) illustrates that and how L2 learners are in a way “embedding” an L2 interactional structure in their talk, and they are doing so consciously, systematically, and collaboratively. The participants realign what they say and do in specific sequential positions, calibrate specifically typed turns and what they are doing in a given sequential position, and as such, (re-)arrange their interaction in systematic patterns that are anchored in a set of structured verbal behaviors that is not part of the canonical repertoire of interactional sequencing found in their L1. In short, Jeremy and Lauren are consciously, systematically, and collaboratively patterning their interaction to match a type of interaction structure specific to the German language; they are “doing being German” as compliment givers and recipients for the duration of the compliment sequence. In sum, both data examples illustrate how interactants are manipulating the patterning and sequencing of the turns they produce and the actions these accomplish in slightly different ways for the duration of a given, though clearly bounded and marked, verbal activity. This process is initiated, accompanied, and concluded by specific markers, including laughter, smile voice, repair on the action level, and topic shifts. In both examples, these characteristics accompany the deliberate and systematic deployment of actions that lie outside of the otherwise normative expectations operant in these sequential contexts and may thus be seen as a sort of game; as long as a departure from normative expectations on the action level is sufficiently marked as a performance (i.e. as the doing of something special or unexpected), the rules of the game can be playfully bent or broken without changing them. For ultimately, the ways in which this kind of pattern manipulation happens are by no means random, but rather firmly anchored in the principles and mechanisms of talk-in-interaction itself; speakers are bound to engage in turn-taking, speakers are bound to furnish specific actions in temporal succession, and these will be interpreted in light of the overall pragmatic import of the immediately preceding constellation of turns and sequences of turns and what they may do. I posit that this is precisely what a social-interactionist perspective on language use would offer for characterizing a notion of language play when typed turns, their sequencing, and the attendant preference structure become the object of conscious pattern manipulation, both in native speaker interaction and in interaction of second language learners.



Playing with turns, playing with action? A social-interactionist perspective 

 65

4 Joint pattern manipulation as play Humans recognize a game or the activity of play when they see it. Minimally, all that is required to do so is discerning and recognizing the conventional (i.e. shared) structures and patterns of a game or a playful activity to which its participants must jointly orient if they are to succeed in playing it. The metaphor LANGUAGE = GAME (see Wittgenstein 1973) is not new and underlies any basic conception of patterned social convention. Clearly, without a notion of orderliness in the sense of a rule-based system (and manifest ambiguity in contextual language use notwithstanding), we could not engage in (language) play collaboratively, nor could we recognize, bend, or break the systematic patterns, principles, or mechanisms underlying a given game that involves a given measure of social accord. In any game, bending or breaking the rules is possible and in fact routinely done by participants, and such transgressions have a way of making the normative expectations visible. Whether we flaunt social convention by cheating at a game of cards or by furnishing overly generous interpretations of traffic laws, such departures from normative expectations are usually sanctioned by coparticipants if they are noticed (e.g. dismissing the player, honking), thus ostensibly reaffirming the validity of a given convention. As such, the verbal behavior visible in both data examples constitutes conscious, systematic, and collaborative pattern manipulation that reaches beyond normative expectations of patterned behavior surrounding a clearly defined verbal activity. In light of the fundamental CA concepts reviewed above, we see that this perspective on language play is compatible with how CA conceptualizes the fundamentals of language itself. Both talk-in-interaction and play can be viewed as orderly activities inasmuch as they rely on social convention; the orderly patterns of both talk-in-interaction and play extend beyond individual interactants or players into the realm of shared cognition; both provide often ambiguous though clearly rule-based behavioral patterns to which participants normatively orient, both provide orderly structures that participants may engage in or choose not to engage in, and either course of action is socially accountable and sanctionable. In other words, in both playing a game and in talk-in-interaction, we may conform to, bend, or break the principles of engagement at our own peril, if and when we see fit, but always to consequential social effect; there is no talk that is inconsequential in the sense that it does not engender socially accountable meaning (Huth 2010). The basic forces underlying typed turns, pairs and sequences thereof, and preference organization are built on social and cultural convention, otherwise we would not find variability within or across languages (Tannen 1981; Tannen 2013), nor would we be able to recognize or analyze them systematically. Engaging in talk-in-interaction itself can thus be seen as playing

66 

 Thorsten Huth

a game in terms of either following, bending, or breaking patterned behaviors anchored in social convention. The similarities of such a basic notion of play and using language in social and interactional context are striking. If Wittgenstein (1973) as a language philosopher formulated an intriguing hypothesis for how “language” as a semiotic resource might well be conceptualized, then I would argue that CA-research has independently furnished an entire conceptual inventory built from hard, empirical evidence that happens to match Wittgenstein’s a priori theorizing to a striking degree of detail. Play can be aptly described as a patterned human activity in which its participants jointly and collaboratively orient to behavioral patterns which are anchored in the matrix of social convention within a given group of speakers; doing or not doing so is a matter of choice for each group member though always socially accountable and sanctionable. Similarly, nego­ tiating the normative patterns of a game will affect the relative success or failure of playing it. These are the parameters allowing us to recognize a game when we see it and thus constitute the requisite structural knowledge allowing us to play to begin with. Ultimately, if we consider the similarities between some of the fundamental principles of talk-in-interaction reviewed above in light of the metaphor LANGUAGE = GAME inasmuch as both provide patterned resources to which a group of participants normatively orients, the notion of playing as an activity that stands out of other activities both comes into sharp focus and dissolves at the same time. For if issuing turns-at-talk is consistent with the notion of (language) play as such, then engaging in interaction itself becomes the kind of language game that Wittgenstein may have imagined; taking turns as a patterned activity becomes play in and of itself. Just as humans cannot not communicate, the very nature  of talk-in-interaction and its patterning in a given language community would render the very activity of engaging in interaction as play. For when humans engage in interaction, they cannot help but issue turns-attalk and organize them in temporal succession. At the same time, and as the analysis above suggests, the mechanisms and principles of talk-in-interaction allow interactants to mark specific turns and sequences thereof as standing out from its surroundings, to mark the doing of something special. Nonetheless, thus bending the rules deliberately, systematically, and jointly to engage in play with action on the utterance level relies on the very principles that are being manipulated. Last, a note on methodology. While the data analysis of (6) and (7) above would withstand methodological scrutiny in terms of action ascription, next-turn proof procedure, and the empirical basis for the specific L1/L2 language learning and language usage contexts relevant for these data examples, it may be noted



Playing with turns, playing with action? A social-interactionist perspective 

 67

that the specific conclusions I draw from that sequential analysis may at first glance be seen as less methodologically orthodox than desirable. Conversation analysis’ emic dictum clearly requires that the analytic categories relevant for an analysis (here: a given notion of “play”) can be shown to be demonstrably relevant for the very interactants participating in the interaction analyzed. As such, one might expect a detailed CA of (6) and (7) to illuminate that and how participants are “doing being playful”. However, and to assuage such concerns, my course of argument here does not aim at such an analysis nor does it produce such assertions; I do emphatically not claim that the participants in (6) and (7) are “doing being playful” in conversation analytic terms. I simply demonstrate, in detail, that and how participants mark a given constellation of actions as standing out from the surrounding talk and thereby negotiate the relative social accountability operant for the sequences-in-progress. I then make an analytic leap by likening the demonstrable patterning and re-patterning of conversational action in (6) and (7), and (alternatively) the very nature of talk-in-interaction as a whole, to a possible notion of language play inasmuch as both are built from the principles and mechanisms underlying social convention. This amounts to theorizing the data and what they may mean from the outset, and it is done in the service of characterizing what may well be parts of the elusive notion of “play” as viewed from one possible angle. What remains is the unarguable insight that play in general and language play in particular accompany human cognitive and social development over time, and language provides both the means for playing and objects to play with. Talkin-interaction provides the social matrix in which language play with words and grammar may be collaboratively achieved by interactants. However, this study primarily sought to illustrate that (inter-)action itself can become a meta-linguistic object of language play. The data illustrate vividly that L2 learners are able to subject the actions their utterances achieve to conscious, systematic, and joint pattern manipulation when using the L2 for their social and interactional needs. If adult L2 learners can successfully engage in meta-linguistic play with language and the actions it performs by issuing turns-at-talk, the question is if such play with action on the utterance level is also manifest in human interactional development during childhood. Thus, understanding how social actors may collaboratively play with action itself as a salient social learning mechanism opens a promising avenue to the study of L2 learners’ social and interactional development over time, be it conceptualized in terms of L2 pragmatic development (Kasper and Rose 2003) or as L2 interactional development (Hall, Hellerman, and PekarekDoehler 2011). This endeavor may be gainfully coupled with studies from first language acquisition that examine how the doing of action (Sidnell and Enfield 2014) emerges in and through talk-in-interaction, and to which extent this process

68 

 Thorsten Huth

may be routinely accompanied by language play inasmuch as it is conceptualized as the conscious, systematic, and joint manipulating of social action itself.

Transcription conventions . A period indicates a fall in tone. , A comma indicates continuing intonation. ? A question mark indicates rising intonation. ?, A question mark and a comma indicate rising intonation weaker than that indicated by a question mark. : A colon indicates an extension of the sound or syllable it follows (co:lon). ::: More colons prolong the stretch (co:::lon). - A single dash indicates an abrupt ending or cutoff. mine Emphasis is indicated by underlining. CAP Capital letters are used to indicate an utterance, or part thereof, that is spoken much louder than the surrounding talk. hhh. Audible aspirations .hhh Audible inhalations. (( )) Vocalizations that are not recognizable, i.e. the transcription is not clear. >mine< Part of an utterance is delivered at a pace quicker than the surrounding talk. ( ) Items in doubt are enclosed with single parenthesis. [I used Utterances starting simultaneously are linked together with left-hand [I saw brackets. I us[ed to When overlapping utterances do not start simultaneously, left-hand [he is brackets are used to mark the point at which an ongoing utterance is joined. = Utterances are linked together with equal signs when they are latched immediately, i.e. without an interval in between them. (0.7) Intervals in the stream of talk are timed in tenths of a second and inserted within parentheses, either within an utterance or between utterances.

References Atkinson, John M. & Paul Drew. 1979. Order in court: The organisation of verbal interaction in judicial settings. London: Macmillan.



Playing with turns, playing with action? A social-interactionist perspective 

 69

Atkinson, John. M. & John Heritage (eds.). 1984. Structures of social action: Studies in conversation analysis. Cambridge, U.K.: Cambridge University Press. Austin, John. L. 1962. How to do things with words. Oxford: Oxford University Press. Barraja-Rohan, Anne-Marie & Catherine R. Pritchard. 1997. Beyond talk: A course in communication and conversation skills for intermediate adult learners of English. Melbourne: Western Melbourne Institute of TAFE. Barraja-Rohan, Anne-Marie. 2011. Using conversation analysis in the second language classroom to teach interactional competence. Language Teaching Research 15(4): 479–507. Bell, Nancy D. 2005. Exploring L2 language play as an aid to SLL: A case study of humor in NS– NNS interaction. Applied Linguistics 26(2): 192–218. Bell, Nancy D. 2011. Comparing playful and nonplayful incidental attention to form. Language Learning 62(1): 236–265. Belz, Julie A. 2002. Second language play as a representation of the multicompetent self in foreign language study. Journal of Language, Identity, & Education 1(1): 13–39. Betz, Emma M. & Thorsten Huth. 2014. Beyond grammar: Teaching interaction in the German language classroom. Die Unterrichtspraxis/Teaching German 47(2): 140–163. Bilmes, Jack. 1988. The concept of preference in conversation analysis. Language in Society 17(2): 161–181. Bilmes, Jack. 2014. Preference and the conversation analytic endeavor. Journal of Pragmatics 64: 52–71. Broner, Maggie A. & Elaine Tarone. 2001. Is it fun? Language play in a fifth-grade Spanish immersion classroom. The Modern Language Journal 85(3): 363–79. Brown, Penelope & Stephen C. Levinson. 1987. Politeness: Some universals in language usage. Cambridge, U.K.: Cambridge University Press. Bushnell, Cade. 2008. ‘Lego my keego!’: An analysis of language play in a beginning Japanese as a foreign language classroom. Applied Linguistics 30(1): 49–69. Cekaite, Asta & Karin Aronsson. 2004. Repetition and joking in children’s second language conversations: Playful recyclings in an immersion classroom. Discourse Studies 6(3): 373–392. Cekaite, Asta & Karin Aronsson. 2005. Language play, a collaborative resource in children’s L2 learning. Applied Linguistics 26(2): 169–191. Clayman, Steven E. 2013. Turn-constructional units and the transition-relevance place. In Jack Sidnell & Tanya Stivers (eds.), The handbook of conversation analysis, 150–166. Chichester, West Sussex: Wiley-Blackwell Publishing. Cook, Guy. 1997. Language play, language learning. English Language Teaching Journal 51(3): 224–231. Cook, Guy. 2000. Language play, language learning (Oxford Applied Linguistics). Oxford: Oxford University Press. Davidson, Judy. 1984. Subsequent versions of invitations, offers, requests, and proposals dealing with potential or actual rejection. In J. Maxwell Atkinson & John Heritage (eds.), Structures of social action: Studies in conversation analysis (Studies in Emotion and Social Interaction), 102–128. Cambridge: Cambridge University Press. Dervin, Fred & Anthony Liddicoat (eds.). 2013. Linguistics for intercultural education. Amsterdam: John Benjamins. Drew, Paul. 1984. Speakers’ reportings in invitation sequences. In J. Maxwell Atkinson & John Heritage (eds.), Structures of social action: Studies in conversation analysis (Studies in Emotion and Social Interaction), 129–151. Cambridge: Cambridge University Press.

70 

 Thorsten Huth

Drew, Paul. 2013. Turn design. In Jack Sidnell & Tanya Stivers (eds.), The handbook of conversation analysis, 131–149. Chichester, West Sussex: Wiley-Blackwell Publishing. Glenn, Phillip. 2003. Laughter in interaction (Studies in Interactional Sociolinguistics 18). Cambridge: Cambridge University Press. Goffman, Erving. 1981. Forms of talk. Philadelphia: University of Pennsylvania Press. Golato, Andrea. 2002. German compliment responses. Journal of Pragmatics 34(5): 547–571. Golato, Andrea. 2005. Compliments and compliment responses: Grammatical structure and sequential organization. Amsterdam: John Benjamins Publishing. Grice, Herbert P. 1975. Logic and conversation. In Peter Cole & Jerry Morgan (eds.), Syntax and semantics 3: speech acts, 41–58. New York: Academic Press. Haakana, Markku & Marja-Leena Sorjonen. 2011. Invoking another context: Playfulness in buying lottery tickets at convenience stores. Journal of Pragmatics 43: 1288–1302. Hall, Janet. C., John Hellerman & Simone Pekarek-Doehler (eds.). 2011. L2 interactional competence and development. Multilingual Matters. Hayashi, Makoto. 2013. Turn allocation and turn sharing. In Jack Sidnell & Tanya Stivers (eds.), The handbook of conversation analysis, 167–190. Chichester, West Sussex: WileyBlackwell Publishing. Heritage, John. 1984. Garfinkel and ethnomethodology. Cambridge, U.K.: Polity Press. Huth, Thorsten & Carmen Taleghani-Nikazm. 2006. How can insights from conversation analysis be directly applied to teaching L2 pragmatics? Language Teaching Research 10(1): 53–79. Huth, Thorsten. 2006. Negotiating structure and culture: L2 learners’ realization of L2 compliment–response sequences in talk–in–interaction. Journal of Pragmatics 38(12): 2025–2050. Huth, Thorsten. 2007. Pragmatics revisited: Teaching with natural language data. Die Unterrichtspraxis/Teaching German 40(1): 21–33. Huth, Thorsten. 2010. Can talk be inconsequential? Social and interactional aspects of elicited second language interaction. The Modern Language Journal 94(4): 537–553. Kasper, Gabriele. 1992. Pragmatic transfer. Second Language Research 8(3): 203–231. Kasper, Gabriele & Kenneth R. Rose. 2003. Pragmatic development in a second language. Wiley-Blackwell. Kim, Yongho & David Kellogg. 2007. Rules out of roles: Differences in play language and their developmental significance. Applied Linguistics 28(1): 25–45. Kotthoff, Helga. 1993. Disagreement and concession in disputes: On the context sensitivity of preference structures. Language in Society 22(2): 193–216. Kramsch, Claire & Patricia Sullivan. 1996. Appropriate pedagogy. English Language Teaching Journal 50(3): 199–212. Kramsch, Claire (ed.). 2002. Language acquisition and language socialization: ecological perspectives (Advances in Applied Linguistics). New York, NY: Continuum. Lantolf, James P. 1997. The function of language play in the acquisition of L2 Spanish. In Ana Terresa Perez-Leroux and William R. Glass (eds.), Contemporary perspectives on the acquisition of Spanish, vol. 2: Production, processing, and comprehension, 3–24. Sommerville, MA: Cascadilla Press. Lantolf, James P. (ed.). 2000. Sociocultural theory and second language learning (Oxford Applied Linguistics). Oxford, U.K.: Oxford University Press. Lazaraton, Anne. 1997. Preference organization in oral proficiency interviews: The case of language ability assessments. Research on Language and Social Interaction 30(1): 53–72.



Playing with turns, playing with action? A social-interactionist perspective 

 71

Lazaraton, Anne. 2014. Aaaaack! The active voice was used! Language play, technology, and repair in the Daily Kos weblog. Journal of Pragmatics 64: 102–116. Leech, Geoffrey N. 1983. Principles of pragmatics (Longman Linguistics Library 30). London: Longman. Lerner, Gene H. 1996. Finding “face” in the preference structures of talk–in–interaction. Social Psychology Quarterly 59(4): 303–321. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Levinson, Stephen C. 2013. Action formation and ascription. In Jack Sidnell & Tanya Stivers (eds.), The handbook of conversation analysis, 103–130. Chichester, West Sussex: WileyBlackwell Publishing. Liddicoat, Anthony J. & Angela Scarino. 2013. Intercultural language teaching and learning. Oxford: Wiley-Blackwell Publishing. Mey, Jacob L. 2001. Pragmatics: An introduction. London: Blackwell Publishing. Pomerantz, Anita. 1978. Compliment responses: Notes on the co-operation of multiple constraints. In Jim Schenkein (ed.), Studies in the organization of conversational interaction 79–112. New York: Academic Press. Pomerantz, Anita. 1984. Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. In J. Maxwell Atkinson & John Heritage (eds.), Structures of social action: Studies in conversation analysis (Studies in Emotion and Social Interaction), 57–101. Cambridge: Cambridge University Press. Pomerantz, Anita & John Heritage. 2013. Preference. In Jack Sidnell & Tanya Stivers (eds.), The Handbook of Conversation Analysis, 210–228. Chichester, West Sussex: Wiley-Blackwell Publishing. Pomerantz, Anne & Nancy D. Bell. 2007. Learning to play, playing to learn: FL learners as multicompetent language users. Applied Linguistics 28(4): 556–578. Rose, Kenneth R. & Gabriele Kasper (eds.). 2001. Pragmatics in language teaching (Cambridge Applied Linguistics). Cambridge, U.K.: Cambridge University Press. Sacks, Harvey, Emamuel A. Schegloff & Gail Jefferson. 1974. A simplest systematics for the organization of turn-taking for conversation. Language 50(4): 696–735. Sacks, Harvey. 1987 [1973]. On the preferences for agreement and contiguity in sequences in conversation. In Graham Button & John R. E. Lee (eds.), Talk and social organization, 54–69. Clevedon, Avon: Multilingual Matters Ltd. Schegloff, Emanuel A., Gail Jefferson & Harvey Sacks. 1977. The preference for self-correction in the organization of repair in conversation. Language 53(2): 361–382. Schegloff, Emanuel A. 1968. Sequencing in conversational openings. American Anthropologist 70(6): 1075–1095. Schegloff, Emanuel A. 1990. On the organization of sequences as a source of “coherence” in talk–in–interaction. In Bruce Dorval (ed.), Conversational organization and its development (Advances in Discourse Processes 38), 51–77. Norwood, New Jersey: Ablex Publishing. Schegloff, Emanuel A. 2007. Sequence organization in interaction: A primer in conversation analysis, vol. 1. Cambridge, UK: Cambridge University Press. Scollon, Ron & Suzanne W. Scollon. 2001. Intercultural communication, 2 edn. Malden, MA: Blackwell Publishing. Searle, John R. 1969. Speech acts: An essay in the philosophy of language. Cambridge, UK: Cambridge University Press. Sidnell, Jack & N. J. Enfield. 2014. The ontology of action, in interaction. In N. J. Enfield, Paul Kockelman & Jack Sidnell (eds.), The Cambridge handbook of linguistic anthropology, 423–445. Cambridge, UK: Cambridge University Press.

72 

 Thorsten Huth

Sidnell, Jack & Tanya Stivers (eds.). 2013. The handbook of conversation analysis. Chichester, West Sussex: Wiley-Blackwell Publishing. Taleghani-Nikazm, Carmen. 1998. Politeness in Persian interaction: The preference format of offers in Persian. Cross-roads of Language, Interaction, and Culture 1: 3–11. Taleghani-Nikazm, Carmen. 2002. A conversation analytical study of telephone conversation openings between native and nonnative speakers. Journal of Pragmatics. 34(12): 1807–1832. Taleghani-Nikazm, Carmen. 2006. Request sequences: The intersection of grammar, interaction and social context. Amsterdam: John Benjamins. Taleghani-Nikazm, Carmen & Thorsten Huth. 2010. L2 requests: Preference structure in talk– in–interaction. Multilingua 29(2): 185–202. Tannen, Deborah. 1981. New York Jewish conversational style. International Journal of the Sociology of Language 30: 133–149. Tannen, Deborah. 2013. Turn–taking and intercultural discourse and communication. In Christina Bratt Paulston, Scott F. Kiesling & Elizabeth Rangel (eds.), The handbook of intercultural discourse and communication 135–157. Chichester, West Sussex: WileyBlackwell Publishing. Tarone, Elaine. 2000. Getting serious about language play: Language play, interlanguage variation and second language acquisition. In Bonnie Swierzbin, Frank Morris, Michael E. Anderson, Carol A. Klee & Elaine Tarone (eds.), Social and cognitive factors in second language acquisition: Selected proceedings of the 1999 Second Language Research Forum, 31–54. Somerville, MA: Cascadilla Press. Waring, Hansun Zhang. 2013. Doing being playful in the second language classroom. Applied Linguistics 34(2): 191–210. Wittgenstein, Ludwig. 1973. Philosophical investigations. Wiley-Blackwell Publishing.

Tony Veale

3 The shape of tweets to come: Automating language play in social networks Abstract: Twitter has proven itself a rich and varied source of language data for linguistic analysis. For Twitter is more than a popular new platform for social interaction via language; in many ways Twitter constitutes a whole new genre of text, as users adapt to its limitations (140 character “tweets”) and its novel conventions (e.g. re-tweeting, hashtags). Language researchers can harvest Twitter data to study how users convey meaning with affect, and how they achieve stickiness and virality with the texts they compose. But Twitter presents an opportunity of another kind to the computationally-minded language researcher, a generative opportunity to study how algorithmic models might exploit linguistic hypotheses to compose novel and meaningful micro-texts of their own. This computational turn allows researchers to go beyond merely descriptive models of playful uses of language such as metaphor and irony. It allows researchers to test whether their models embody a sufficiently algorithmic understanding of a phenomenon to facilitate the construction of a fully-automated computational system, one that can generate wholly novel examples that are deemed acceptable by humans. This chapter presents and evaluates one such system, a Twitterbot named @MetaphorMagnet that generates, expresses and shares its own playful insights on Twitter.

1 Introduction A mismatch between a container and its contents can often tell us much more than the content itself, as when a person places the ashes of a deceased relative in a coffee can, or sends a brutal death threat in a Hallmark greeting card. The communicative effectiveness of mismatched containers is just one more reason to be skeptical of the Conduit metaphor (Reddy 1979) – which views linguistic constructs as containers of propositional content to be faithfully shuttled between speaker and hearer – as a realistic model of human communication. Language involves more than the faithful transmission of logical propositions between information-hungry agents, and more effective communication – of attitude, expectation and creative intent – can often be achieved by abusing our linguistic containers of meaning than by treating them with the sincerity that the Conduit metaphor assumes. Consider the case of verbal irony, in which a speaker DOI 10.1515/9781501503993-004

74 

 Tony Veale

deliberately chooses containers that are pragmatically ill-suited to the conveyance of their contents. For instance, the advertising container “If you only see one [X] this year, make it this one” assumes that [X] denotes a category of event – such as “romantic comedy” or “movie about superheroes” – with a surfeit of available members for a listener to choose from. When [X] is bound to the phrase “comedy about Anne Frank” or “musical about Nazis”, the container proves too hollow for its content, and the reader is signaled to the presence of irony. Though such a film may well be one-of-a-kind, the ill-fitting container suggests there are good reasons for this singularity that do not speak to X’s quality as an artistic event. Yet if carefully chosen, an apparently inappropriate container can communicate a great deal about a speaker’s relationship to the content conveyed within, and as much again about the speaker’s relationship to their audience. As more practical limitations are placed on the form of linguistic contain ers, the more incentive one has to exploit or abuse containers for creative ends. Consider the use of Twitter as a communicative medium: writers are limited to micro-texts of no more than 140 characters to convey both their meaning and their attitude to this meaning. So each micro-text, or tweet, becomes more than a container of propositional content: each is a brick in a larger edifice that comprises the writer’s online personae and textual aesthetic. Many Twitter users employ irony and metaphor to build this aesthetic and thus build up a loyal audience of followers for their world view. Yet Twitter challenges many of our assumptions about irony and metaphor. Such devices must be carefully modulated if an audience is to perceive a speaker’s meaning in the playful (mis)match of a linguistic container to its contents. Failure to do so can have serious repercussions when one is communicating to thousands of followers at once, with tweets that demand concision and leave little room for nuance. It is thus not unusual for even creative tweets to come packaged with an explicit tag such as #irony, #sarcasm or #metaphor. Metaphor and irony are much-analyzed phenomena in social media (Ghosh et al. 2015), but this chapter takes a generative approach, to consider the production rather than the analysis of creative linguistic phenomena in the context of a fully- autonomous computational agent – a Twitterbot – that crafts its own metaphorical and ironical tweets from its own knowledge-base of common-sense facts and beliefs. How might such a system exhibit a sense of irony that human users will find worthy of attention, and how might this system craft interesting metaphoric insights from a knowledge-base of everyday facts that are as banal as they are uncontentious? We shall explore the variety of linguistic containers at the disposal of this agent – a fully autonomous computational system on the Web named @MetaphorMagnet – to better understand how such containers can be playfully exploited to convey ironic, witty or thought-provoking views on the



The shape of tweets to come 

 75

world. With @MetaphorMagnet we aim to show that interesting messages are not crafted from interesting contents, or at least not necessarily so. Rather, effective tweets emerge from an appropriate if non-obvious combination of familiar containers with unsurprising factual fillers. In support of this view, we present an empirical analysis of the assessment of @MetaphorMagnet’s uncurated figurative outputs by human judges. Just as one can often guess the contents of a physical container by its shape, one can often guess the meaning of a linguistic container by its form. We become habituated to familiar containers, and just as we might imagine our own uses for a physical container, we often pour our own meanings into suggestive textual forms. For in language, meaning follows form, and readers will generously infer the presence of meaning in texts that are well-formed and seemingly the product of an intelligent entity, even if this entity is not intelligent and any meaning is not intentional. Remarkably, Twitter shows that we willingly extend this generosity of interpretation to the outputs of bots that we know to be unthinking users of wholly aleatoric methods. Twitterbots exploit this charity of interpretation – wherein a well-formed linguistic container is assumed to carry a well-founded meaning – by serving up linguistic forms that readers tacitly fill with their own meanings. We aim to empirically demonstrate here that readers do more than willingly suspend their disbelief, and that a well-packaged linguistic form can seduce readers into seeing what is not there: a comprehensible meaning, or at least an intent to be meaningful. We do this by evaluating two metaphor bots side-by-side: a rational, knowledge-based Twitterbot named @MetaphorMagnet vs. an aleatoric and largely knowledge-free bot named @MetaphorMinute.

2 Digital surrealists: Out of the mouths of bots Most Twitterbots are simple, rule-based systems that use stochastic tools to explore a loosely-defined space of possible textual forms, or what Oulipo (1981) calls a space of “potential literature.” Most bots are thus high-concept, low-complexity generative systems that transplant the aleatoric methods and constraints of the early surrealists, the Oulipo group and the “beat” writers – from André Breton to Raymond Queneau (1961, 1981) to William Burroughs (1963) and Brion Gysin – into the realms of digital content, social networks and online publishing. Each embodies a language game with its own generative rules, or what Breton called “la règle du jeu.” Yet Breton, Queneau, Burroughs and Gysin viewed the use of mechanical rules as merely the first of a two-stage text-creation process: in the first, random recombinant methods are used to confect candidate texts in ways that, though unguided by meaning, are also free of the baleful influence of

76 

 Tony Veale

cliché (see Ricks [1980, 1995] for a discussion of this influence); at the second stage, these candidates are carefully interpreted and filtered by a human, to select those that are novel and interesting and to reject the rest. Most bots implement the first stage but ignore the second, pushing the task of critiquing and filtering candidate texts onto the humans who read and selectively re-tweet them. Nonetheless, some bots achieve surprising linguistic results with the simplest of tools. Consider @Pentametron, a bot that generates accidental poetry by re-tweeting pairs of random tweets of ten syllables apiece (for an iambic pentameter reading) if each ends on a rhyming syllable. When the meaning of each tweet in a couplet happens to cohere with the other, as in “Pathetic people are everywhere”/“Your web-site sucks, @RyanAir”, the pairing produces an emergent meaning that is richer and more resonant than that of either tweet alone. Trend ing social events such as the Oscars or the Super Bowl are especially conducive to just this kind of synchronicity, as in this fortuitous pairing: “So far the @SuperBowl commercials blow”/“Not even gonna watch the halftime show.” If one were so minded, one could trace the lineage of @Pentametron and that of other tweetsplicing bots (such as the headline-creating @twoheadlines of Darius Kazemi) to Oulipo mechanisms such as Queneau’s (1961) flip-book for generating sonnets. In contrast, a bot named @MetaphorMinute wears its aleatoric methods on its sleeve, for its tweets – such as “a haiku is a tonsil: peachblow yet snail-paced” – are not so much random metaphors as random metaphor-shaped texts. Much like Chamberlain and Etter’s (1983) RACTER system, @MetaphorMinute employs a generative strategy that relies on word associations and randomness rather than word meanings and pragmatics. The bot instantiates a standard linguistic container for metaphors – the copula frame “X is a Y” – with random-seeming word choices, and tweets the results without any attempt at quality estimation or filtering every two minutes. Interestingly, @MetaphorMinute’s tweets are just as likely to provoke a sense of mystification and ersatz profundity as they are total incomprehension. Bots such as @Pentametron and @MetaphorMinute do not generate their texts from the semantic-level up; rather, they manipulate texts at the word-level only, and thus lack any sense of the meaning of a tweet, or any rationale as to why one tweet might be better – which is to say, more interesting, more apt or more re-tweetable – than any other. The Full-FACE poetry generator of Colton et al. (2012) also uses a templateguided version of the cut-up method to mash together semantically-coherent text fragments in a way that – much like @Pentametron – obeys certain over-arching constraints on meter and rhyme. These text fragments come from a variety of online sources, ranging from short tweets to long news articles. News stories are a rich source of readymade phrases that convey resonant images, and these can be clipped from a news text using standard NLP techniques, while tweets



The shape of tweets to come 

 77

that use affect-rich language can also be extracted automatically via standard sentiment analysis lexica and tools. Thus, a large stock of resonant similes, such as “blue as a blueberry” or “hot as a sauna” can be extracted from the Web using a search engine (Veale 2014), since the simile frame “as X as Y” is specific enough to query for, and promiscuous enough to match, a rich diversity of typical X:Y associations. These associations can then be recast in a variety of poetic forms to make their clichéd offerings seem fresh again, as in “Blueberry-blue overalls” or “sauna-hot jungle.” Indeed, the very act of juxtaposing clichés can itself be a creative act, as evidenced both by the success of the cut-up method in general and that of specific cut-ups in particular. Consider William Empson’s withering analysis of the persnickety, cliché-hating George Orwell, whom Empson called “the eagle eye with the flat feet” (quoted in Ricks [1995:356], who admires Empson’s “audacious compacting of clichés”). The Full-FACE system is just one of many computational creativity systems that use an autonomous variant of Burroughs and Gysin’s cut-up method to integrate tight constraints on form with loose constraints on meaning. Breton famously stated that “Je ne veux pas changer la règle du jeu, je veux changer de jeu.” Twitterbots do not change or transcend their own rules, but different bots do represent different language games with their own rules. So to change the game, a computational creativity developer can simply build a new bot, to exploit a different set of tropes and linguistic containers. It is rare for any one Twitterbot to incorporate a diverse set of tropes and production mechanisms; each typically follows Breton’s experimentalist approach to art in its random sampling of a specific space of possibilities. Each bot thus forms its own art installation, to showcase a single generative idea. @MetaphorMagnet, the bot at the heart of this chapter, represents a departure from this norm, insofar as it exploits a wide range of tropes and rendering strategies, it employs diverse sources of knowledge, and it applies a variety of reasoning styles to generate surprising conclusions from its stock of otherwise banal facts. But does this added sophistication – bought at the cost of increased system complexity and knowledge-engineering effort – result in tweets that are seen as more meaningful, novel, apt or retweetable by human users? It is this point that exercises us most in the coming sections.

3 Exercises in style over substance Style lends a distinctive shape and appearance to our linguistic containers. But it can do more than make one text stand out from others: it can shape the way one should feel about the contents of a linguistic container. Style can imbue a

78 

 Tony Veale

banal event with the excitement of a thriller, the immediacy of a newsflash or the comedy of a farce. It can make a crass generalization read like a dry scientific fact, or a dull fact read like a tabloid headline. As shown by Raymond Queneau’s remarkable Exercises in Style (1947/1981), style crucially shapes how readers construe the states of affairs conveyed by a text. Queneau at turns wrings pathos, humor and cultural insight from his 99 alternate stylistic renderings of the same banal tale of a scuffle on a bus. Though most Twitterbots embody just a single exercise in style, these bots collectively offer a digital realization of Queneau’s entire experimental agenda. Indeed, it is not too great a stretch to suggest that humans follow these bots for many of the same reasons they read Queneau’s Exercises or his other Oulipo work (such as Cent mille milliards de poèmes from 1961). Meanwhile, more ambitious bots are taking up the Queneau challenge to create original outputs in a wide diversity of styles that grab the eye and shape the attitude of a reader. @BestOfBotWorlds, for instance, invents tweets with faux-inspirational or religious content using forms that satirize the “style” of Jesus, Mohammad, Yoda, Donald Trump and even the Hulk. We humans obtain more mileage than we may ever care to admit from templates, constraints and other “bot”-like stylistic approaches to linguistic creativity. Consider what Matthew McGlone and Jessica Tofighbakhsh (1999) call the Keats heuristic, an insight into creative language use that owes as much to Nietzsche (“we sometimes consider an idea truer simply because it has a metrical form and presents itself with a divine skip and jump”) as to the poet John Keats (“Beauty is truth, truth beauty”). McGlone and Tofighbakhsh (2000) show that when presented with uncommon maxims or proverbs with internal rhyme (e.g. “woes unite foes”), subjects tend to view these as more insightful about the world than the equivalent paraphrases with no internal rhyme at all (e.g. “troubles unite enemies”). While the Keats heuristic is not exactly a license to pun, it is an incentive to rhyme, and to give as much weight (or more still) to superficial aspects of poetry generation as to deep semantics and pragmatics. Indeed, the heuristic is tacitly central to the operation of virtually every computational creativity approach to poetry generation (e.g. Milic 1971; Chamberlain and Etter 1983; Gervás 2000; Manurung et al. 2012; Veale 2013). If human poets ask questions first and rhyme later, computational creativity systems typically rhyme first and ask questions later, if at all. For if the human jury in the O.J. Simpson trial could be turned against bald facts with the Keatsian “If the glove don’t fit you must acquit”, readers of computergenerated poetry can be persuaded to see deliberate meaning and resonance in any output that has a “divine skip and jump.” There is something undeniably special about poetry, whether it is the gentle poetry of William Shakespeare’s “Shall I compare thee to a summer’s day” or the rough poetry of Johnnie Cochrane’s “If the glove don’t fit you must acquit”.



The shape of tweets to come 

 79

Milic (1971), an early computational creativity pioneer, argues that while poetry “is more difficult to write than prose” it offers other freedoms to writers due to the willingness of readers to “interpret a poem, no matter how obscure, until he has achieved a satisfactory understanding.” What then of the enigmatic tweets of bots like @MetaphorMinute, whose obscurity is a function of random word choice and whose surface forms are not designed to make any sense at all? Milic argues that computer poetry serves a useful role other than its obviously generative one, by alerting us to “the curious behavior of familiar words in unfamiliar combinations.” Behavior that makes perfect sense when dealing with the writings of a gifted human poet, such as our tendency to “interpret an utterance by making what concessions are necessary on the assumption that a writer has something in mind of which the utterance is the sign”, is, argues Milic, “inappropriate when the speaker is a computer.” Yet Twitterbots benefit from such concessions and assumptions whether or not followers know them to be bots. This Eliza effect (see Weizenbaum 1966; Hofstadter 1995) is especially pronounced in the coining of would-be metaphors, leading Milic to note “how readily we accept metaphor as an alternative to calling a sentence nonsensical.” @MetaphorMinute and other aleatoric bots wring maximal value from this insight by devising texts that they themselves cannot distinguish from nonsense. This begs an important question: are the meanings imposed on a random text by a creative human of comparable value to those conveyed by a Twitterbot with its own model of the world and its own insights to tweet?

4 Filling linguistic containers with metaphorical meanings What might it mean for a bot to have “something in mind of which [its] utterance is the sign”? When it comes to metaphor generation, we might expect that our bot would generate its figurative tweets from a conceptual model of the world as it sees it, in a way that accords with a sound theory of how and why humans actually use metaphor. For the latter, the field of Artificial Intelligence offers us a range of models to choose from. Computational approaches to metaphor divide into four broad classes: the categorial, the corrective. the analogical and the schematic. Categorial approaches view metaphor as a means to re-conceptualize one idea by placing it into a taxonomic category strongly associated with another (see Hutton 1982; Way 1991; Glucksberg 1998). Corrective approaches view metaphor as an inherently anomalous deviation from literal language, and strive to recover the correspond ing literal meaning of any figurative statement that violates its lexico-semantic norms (see Wilks 1978; Fass 1991). The analogical approaches aim to capture the relational parallels that allow our representation of an idea in one domain, the

80 

 Tony Veale

source, to be systematically projected onto our mental representation of an idea in another, the target (see Gentner et al. 1989; Veale and Keane 1997). Finally, schematic approaches aim to explain how related linguistic metaphors arise as surface manifestations of deep seated cognitive structures called Conceptual Metaphors (Lakoff and Johnson 1980; Carbonell 1981; Martin 1990; Veale and Keane 1992). Each approach has its own merits, but none offers a complete computational solution. Bots that aim for a general competence in metaphor must thus implement a selective hybrid of multiple approaches. Yet each approach also requires its own source of knowledge. Categorial approaches require a comprehensive taxonomy of flexible categories that can embrace atypical members on demand. Corrective approaches are built on a substrate of literal case-frames onto which deviant usages can be correctively projected. Analogical approaches assume an inventory of graph-theoretic representations of concepts, from which a structure-mapping engine can eke out its sub-graph isomorphisms. Schematic approaches rely on a stock of Conceptual Metaphors (CMs) – such as Life is a Journey or Theories are Buildings – to unearth the deep structures beneath the surface of diverse linguistic forms. Though hybrid approaches demand multiple sources of knowledge, there exist public Web services that integrate this knowledge with the appropriate means of using it for metaphor. The Thesaurus Rex Web service of Veale and Li (2013) provides a highly divergent system of fine-grained categorizations that allows a 3rd-party client system to e.g. determine that War and Divorce have each been viewed as kinds of destructive thing, traumatic event and severe conflict in the texts of the Web. The Metaphor Eyes Web service of Veale and Li (2011) is a rich source of relational norms – also harvested at scale from Web texts – such as that businesses earn profits and pay taxes, or that religions ban alcohol and believe in reincarnation. The Metaphor Magnet service of Veale (2014) offers a rich source of the stereotypical properties and behaviors of familiar ideas, and provides a means to retrieve salient CMs from the Google n-grams (Brants and Franz 2006) which can then be elaborated to create novel linguistic metaphors. @MetaphorMagnet relies on each of these public Web services to generate the conceptual conceits that underpin its figurative tweets. For instance, it uses Thesaurus Rex to provide the categorization insights that it then packages as oddone-out lists or as faux-dictionary definitions. It uses the Metaphor Eyes service to provide the relational structures it needs to perform structure mapping and thus concoct original analogies and dis-analogies. And it uses the Metaphor Magnet service to access the stereotypical properties and behaviors of ideas, and to juxtapose these properties via resonant contrasts and norm contraventions. Once the conceptual chassis of a metaphor is constructed in this way, it is then packaged in an apt linguistic form.



The shape of tweets to come 

 81

5 Shaping a tweet: Automated exercises in Twitter style CMs such as Life Is A Journey and Politics Is A Game are more than productive deep-structures for the generation of whole families of linguistic metaphors; they also provide the conceptual mappings that shape our habitual thinking about such familiar ideas as Life, Love, Politics and War. Politicians and philosophers exploit conceptual metaphors to frame an issue and shape our expectations; when a CM fails to match our own experience, we reject it and switch to a more apt metaphor. So a metaphor-generating bot can thus create a thought-provoking opposition by pitting one CM against another that advocates a conflicting view of the world. The following tweet from @MetaphorMagnet uses this approach to contrast two views on #Democracy: To some voters, democracy is an important cornerstone. To others, it is a worthless failure. #Democracy= #Comerstone #Democracy= #Failure The CM Democracy Is A Cornerstone (of society) is often used to frame political discussions, and can be seen as an specialization of the CM Society Is A Building, itself an elaboration of the CM Organization Is Physical Structure (see Grady 1997). Yet the importance of cornerstones to the buildings they anchor finds a sharp contrast in the assertion that Democracy Is A Failure. Each of these affective claims is so commonly asserted that they can be found in the Google n-grams, a large database of short fragments of frequent Web texts. The 4-gram “democracy is a cornerstone” has a frequency of 91 in the Google n-grams, while the 4-gram “democracy is a failure” has a frequency of 165. These n-grams, which suggest potential CMs for @MetaphorMagnet, are elaborated with added detail via the Metaphor Magnet Web service, which tells the bot that the stereotypical cornerstone is important and the stereotypical failure is worthless. The following tweet makes similar use of CMs found in the Google n-grams, but renders the conflict in a different linguistic container: Remember when tolerance was promoted by crusading liberals? Now, tolerance is violence that only fearful appeasers can avoid. The bot is guided here by the suggestive Google 3-gram “Tolerance for Violence” (frequency=1353), but it does not directly contrast the ideas #Tolerance and #Violence. Instead, it finds a potential analogy in this juxtaposition, between the promoters of #Tolerance (which it renders as crusading liberals) and the opponents of #Violence (which it renders as fearful appeasers). The choice of stereotypical properties (crusading and fearful) is driven by the bot’s need to create a resonant semantic opposition. The bot omits the hashtags #Tolerance=#Violence from this

82 

 Tony Veale

tweet due to the confines of Twitter’s 140-character limit. But it can also choose to render a complex conceit across two successive tweets, as in the following: Remember when research was conducted by prestigious philosophers? #Research=#Fruit #Philosopher=#Insect Now, research is a fruit eaten only by lowly insects. #Research=#Fruit #Philosopher=#Insect @MetaphorMagnet uses a number of packaging strategies to turn a figurative comparison into an ironic observation, ranging from the use of an explicit #Irony hashtag (which is commonplace on Twitter) to the use of “scare” quotes to focus on the part of a tweet most deserving of disbelief. The following tweet showcases both of these strategies: #Irony: When some chefs prepare “fresh” salads the way apothecaries prepare noxious poisons. #Chef=#Apothecary #Salad=#Poison‬ Irony offers a concise means of contrasting two points of view: that which is expected and the disappointing reality. By comparing the preparation of salads – the “healthy” option on most menus – to the preparation of poisons, this analogy undermines the expectation of healthfulness and suggests that some salads are noxious and chemical-filled. The real world is filled with situations in which naturally antagonistic properties are found in surprising proximity. These situations, if expressed in the right linguistic form, can be elevated to the level of situational irony. Consider, for instance, the following @MetaphorMagnet tweet: #Irony: When the timers that are found in enjoyable games activate gruesome bombs. #Enjoyable=#Gruesome‬ It is important to stress that @MetaphorMagnet does not simply fill linguistic templates with related words. Rather, the above tweet is constructed at the knowledge-level, by a bot that intentionally seeks out stereotypical norms that are related (e.g. by a pivotal idea timer) yet which can be placed into antagonistic juxtapositions around this pivot. In effect, the goal of the linguistic rendering is to package a knowledge-level conceit – typically a conflict of ideas and properties – in a tweet-sized narrative. For example, the following tweet is stylistically rendered as a narrative of change: To join and travel in a pack: This can turn pretty girls into ugly coyotes. #Girl=#Coyote Twitter offers unique social affordances that allow a bot to elevate almost any contrast of ideas into a dramatic narrative. Rather than talk of generic liberals



The shape of tweets to come 

 83

or appeasers, a bot can give these straw men real names, or at least invent fake names that look like the real thing and which, as Twitter handles, seem wittily apropos to the views that are espoused. In this way, by imagining its central conceit as a topic of a vigorous debate by real people, a bot can turn an abstract metaphor into a concrete situation with its own colorful participants. Consider the social debate that is made personal in this tweet from @MetaphorMagnet: .@war_poet says history is a straight line .@war_prisoner says it is a coiled chain #History=#Line #History=#Chain The handles @war_poet and @war_prisoner are invented by @MetaphorMagnet to suit, and amplify, the figurative views that they are advanced in the tweet, by using a mix of relational knowledge (from the Metaphor Eyes service) and language data (via the Google n-grams). Since poets write poems about the wars that punctuate history, and poems contain lines, the 2-gram “war poet” is recognized as an apt handle for an imaginary Twitter user who might advance a view of history as a line. In this case the handle @war_poet really does name a real Twitter user, but this only adds to the sense that Twitterbot confections are a new kind of interactive theatre and performance art (see Dewey 2014). The most profound aspects of this contrast are not appreciated by @MetaphorMagnet itself, or at least not yet. For example, the bot does not yet appreciate what it means for history to be a straight line, and while it knows enough to invent the intriguing handle @war_prisoner, neither does it appreciate what it might mean to be a prisoner of history, enslaved in a repeating cycle of war. Our bots will always evoke in we humans more than they themselves can ever appreciate, yet this may itself be a key part of computational creativity’s allure.

6 Content versus container: Evaluating metaphor generation @MetaphorMagnet differs from @MetaphorMinute in a number of key ways. For one, its mechanics are informed by Lakoff and Johnson’s Conceptual Metaphor Theory and a range of computational approaches. For another, it draws on considerable semantic and linguistic resources, from a large knowledge-base of conceptual relations and stereotypical beliefs to the linguistic diversity of the Google n-grams. Note that all of @MetaphorMagnet’s tweets – all its hits and all its misses – are open to public scrutiny on Twitter. But to empirically evaluate the success of the bot as a knowledge-based, theory-driven producer of novel, meaningful and retweet-worthy metaphors, we turn to the crowdsourcing platform CrowdFlower, where we conduct a comparative evaluation of

84 

 Tony Veale

@MetaphorMagnet and its closest knowledge-free counterpart, @MetaphorMinute. The latter, designed by noted bot-maker Darius Kazemi, uses a wholly aleatoric approach to metaphor generation yet has over 500 followers on Twitter that do not mind its one-every-two-minutes scattergun approach to generation. @ MetaphorMinute crafts metaphors by filling a template with nouns and adjectives that are chosen more-or-less at random, to produce inscrutable tweets such as “a cubit is a headboard: stational yet tongue-obsessed.” We chose 60 tweets at random from the past outputs of each Twitterbot. CrowdFlower annotators, who were each paid a small sum per judgment, were not informed of the origin of any tweet, but simply told that each was selected from Twitter because of its metaphorical content. We did not want annotators to actively suspend their disbelief by knowingly dealing with bot outputs. Annotators were paid to rate the content of each tweet along three dimensions, Comprehensibility, Novelty and likely Retweetability, and to rate all three dimensions on the same scale: Very Low to Medium Low to Medium High to Very High. Ten annotations were solicited for each dimension of each tweet, though the responses of likely scammers (non-engaged annotators) were later removed from the dataset. Tabs. 1 through 3 present the distributions of mean ratings per tweet, for each dimension and each Twitterbot. Tab. 1: Comprehensibility of @MetaphorMagnet and @MetaphorMinute Comprehensibility Very Low Med. Low Med High Very High

Metaphor Magnet

Metaphor Minute

11.6% 13.2% 23.7% 51.5%

23.9% 22.2% 22.4% 31.6%

More than half of @MetaphorMagnet’s tweets were ranked as having very high comprehensibility, while less than one third of @MetaphorMinute’s tweets are so ranked. More surprising, perhaps. is the result that annotators found more than half of @MetaphorMinute’s wholly random metaphors to have mediumhigh to very-high comprehensibility. This Twitterbot’s use of abstruse terminology, such as stational and peachblow, may be a factor here, as might the bot’s use of the familiar copula container X is Y for its metaphors, which may well seduce annotators into believing that an apparent metaphor really does have a comprehensible meaning, if only one were to expend enough mental energy to actually discern it.

The shape of tweets to come 



 85

Tab. 2: Novelty ratings of @MetaphorMagnet and @MetaphorMinute Novelty Very Low Med. Low Med High Very High

Metaphor Magnet

Metaphor Minute

11.9% 17.3% 21% 49.8%

9.5% 12.4% 14.9% 63.2%

The dimension Novelty yields results that are equally surprising. While half of @MetaphorMagnet’s metaphors are rated as having very-high novelty in Tab. 2, almost two-thirds of @MetaphorMinute’s tweets are just as highly rated. However, we should not be overly surprised that @MetaphorMinute’s bizarre juxtapositions of rare or unusual words, as yielded by its unconstrained use of aleatoric techniques, are seen as more unusual than those word juxtapositions arising from @MetaphorMagnet’s controlled use of attested Web n-grams and stereotypical knowledge. As shown by Giora et al. (2004), novelty is neither a source of pleasure in itself nor is it a reliable benchmark of creativity. Rather, pleasurability derives from the recognition of useful novelty, that is, novelty that can be understood and appreciated relative to the familiar. On Twitter, useful exploitation is frequently a matter of social reach. A tweet is novel and useful to the extent that it attracts the attention of Twitter users and is deemed worthy of re-tweeting to others in one’s social circle. Our third dimension, Re-Tweetability, reflects the likelihood that an annotator would ever consider re-tweeting a given metaphorical tweet to others. Though we ask annotators to speculate here – neither bot has enough followers to perform a robust statistical analysis of actual retweet rates – the results largely conform to our expectations. The results presented in Tab. 3 show retweetability to be a matter of novelty and comprehensibility together, and not just a matter of novelty alone. Though annotators are not generous with their Very-High ratings for either bot, @MetaphorMagnet’s tweets are judged to be significantly more re-tweetable than the largely random offerings of @MetaphorMinute. Tab. 3: Retweetability of @MetaphorMagnet and @MetaphorMinute Retweetability Very Low Med. Low Med High Very High

Metaphor Magnet

Metaphor Minute

15.5% 41.9% 27.4% 15.3%

41% 34.1% 15% 9.9%

86 

 Tony Veale

This is just as well, given the considerable gap in complexity and sophistication that exists between the two bots. But this is an encouraging result not just for theory-informed Twitterbots like @MetaphorMagnet and their creators, but for Twitter itself. Twitter offers a compelling platform for research into interactive play through language, not least because its human users appreciate these phenomena when they see them. But as pointed out in Reddy (1979), the conduit metaphor of language is an imperfect one, and the linguistic containers that shuttle back and forth between speakers may convey much more or much less than they appear to convey. The appearance of comprehensibility may not always result in actual comprehension, and so, while a Computational Creativity system may cleverly use packaging and style to foster a belief that a given tweet has a coherent meaning, it cannot insert this meaning into the head of a reader. Meaning is the product of interpretation, and interpretation is often hard. Milic (1971) notes that in a context that licenses a poetic interpretation, such as one in which a reader is told that a particular text is a metaphor, readers are more likely to accept that the text – as inscrutable as it may be – has a metaphorical meaning rather than dismiss it as nonsense. Recall that over 75% of @MetaphorMagnet’s tweets and over 50% of @MetaphorMinute’s tweets are judged as having medium-high to very-high comprehensibility. We thus need to look deeper, to determine whether raters can actually back up these judgments with actual meanings. So in a second CrowdFlower experiment, we make raters work harder, to reconstruct a partial tweet by adding the missing information that will make it whole and apt again. That is, we employ a cloze test format for this experiment, by removing from each tweet the pair of key qualities that anchor the tweet and make its comparison of ideas seem meaningful and apt. For @MetaphorMagnet, for example, we remove the properties detailed and vague in this tweet: To some freedom fighters, freedom is a detailed recipe. To others, it is a vague dream. #Freedom=#Recipe #Freedom=#Dream For @MetaphorMinute, we blank out hippy and revisional in the following: a flatfoot is a houseboat: hippy and revisional For each tweet from each bot, we blank out a pair of original qualities as above; this pairing is the answer that is sought from human judges. We also choose 4 distractor pairs for each original pair, by selecting pairs from other tweets from the same bot. As in our first experiment, we chose 60 tweets at random from the past outputs of each bot, and 10 ratings were solicited for each. Annotators were presented with a tweet in which the key properties were blanked out, and given

The shape of tweets to come 



 87

five randomly ordered pairs of possible fillers (the original pair and four distractors from other tweets) to choose from. Tab. 4: Relative Aptness of @MetaphorMagnet and @MetaphorMinute Aptness Very Low Med. Low Med High Very High

Metaphor Magnet

Metaphor Minute

0% 22% 58% 20%

84% 16% 0% 0%

To make the results of the experiment comparable to those of the 1st experiment (Tabs. 1, 2, 3), we obtain the mean aptness of each tweet, so that e.g. if 7 out of 10 raters correctly choose the original pairing, then that tweet is deemed to have an aptness of 0.7. We then place these aptness scores into bands, where the Very Low band = 0 to 0.25, Medium Low = 0.26 to 0.5, Medium High = 0.51 to 0.75, and Very High = 0.76 to 1. By calculating the distribution of tweets to each band, we can determine e.g. the percentage of tweets from each bot that are put into the Very High band. Our hypothesis is rather straightforward: if tweets are linguistic containers that are carefully crafted to convey a particular meaning, then it should be easier to select the missing pair of qualities that make this meaning whole again; if, on the other hand, the tweet is all there is, and its content is chosen mostly at random, then raters will choose the right pairing with no more success than random selection. The results reported in Tab. 4 bear out this hypothesis. The Eliza effect (Hofstadter 1995) can lead us to appreciate a bot’s tweets as meaningful but it cannot tell us what this meaning should be. Though the results above may seem a foregone conclusion, as @MetaphorMagnet’s tweets are designed to communicate a fully recoverable meaning while those of @MetaphorMinute are not, this is surely what it means to engage in real communication: to design an utterance so that an intended meaning is re-created, in whole or in part, in the mind of an intelligent, receptive audience.

7 Concluding remarks: Believing is seeing Whenever a machine uses style and packaging to convey a sense of understanding and profundity with otherwise shallow linguistic forms, as in Weizenbaum’s (1965) infamous ELIZA system (which fooled some users into believing it was a

88 

 Tony Veale

fully-functional and caring psychotherapist), the label “ELIZA Effect” proves to be an apt one (Hofstadter 1995). However, we humans are also subject to an ELIZA effect of our own, insofar as we often do others the courtesy of assuming their utterances to be freighted with real meaning and creative intent, and will often work hard to uncover that meaning for them. At one time or another, we have all relied on catch-phrases, clichés, slogans, idioms, canned jokes and other half-empty linguistic containers to suggest to others that we have deeper meanings in mind, or have something more profound to offer, than we actually do. In a famous polemical essay from 1946, George Orwell excoriates speakers of English for their reliance on jargon, foreign words and empty phraseology as a substitute for thoughts of real substance, while Geoff Pullum (2003) upbraids modern speakers for a grating over-reliance on “multiuse, customizable, instantly recognizable, time-worn, quoted or misquoted phrases or sentences that can be used in an entirely open array of different jokey variants by lazy journalists and writers.” These “phrases for lazy writers in kit form” are not that different from the template-based language games played by superficial Twitterbots, and though we humans fill our templates – such as “X is the new black”, “In X no one can hear you scream” or “if the Eskimos have N words for snow then Xs surely have as many for Y” – with lexical fillers that are contextually apt, we employ our templates to be just as provocative, and to imply or to suggest more than we actually mean. By fitting our meanings to familiar structures, we give our readers the cues they need to see the meanings we want them to see. As the cliché goes, seeing is believing, yet when metaphor and irony are involved, the converse is also true, for these turn believing into a kind of seeing. Ultimately, metaphor and irony are cognitive devices for generating and conveying a specific viewpoint. So at the level of ideas, every conceptual metaphor offers a means of viewing one idea though the lens of another, while at the level of words every linguistic metaphor offers a means of concisely conveying this viewpoint to another, and of helping others to view the world from the same vantage point. If the situation with irony is that much more complicated, it is because irony offers a stereoscopic viewpoint that conveys and contrasts two conflicting perspectives at once to highlight the disparity between the world as it appears and the world as it should be, or at least as it was promised to be. Yet whether one is thinking and communicating via metaphor or via irony, construction proceeds in much the same direction, from perception to conceptualization to expression. The hashtag #irony is more often used on Twitter as an observation than as a warning, and corresponds more to the conversational gambit “Isn’t it ironic …” than to an insurance policy against potential misunderstanding. Yet a creative writer cannot make a tweet ironic by adding the hashtag #irony any more than one can make it a metaphor by adding #metaphor or turn it into a



The shape of tweets to come 

 89

witticism by adding #funny. Though first-generation bots such as @MetaphorMinute achieve the appearance of metaphor, by generating random metaphorshaped tweets that merely use the copula container for its X is Y metaphors, there are no containers that can imbue random juxtapositions with ironic or figurative meanings, and no hashtag that can magically substitute for a lack of original insight. To generate a creative figurative statement that really is designed to be playfully figurative, such as a witticism that is both ironic and metaphorical, an automated system must proceed in the same way as a human: from insight to conceptualization to careful packaging in an appropriate linguistic container. In the chapter we have shown how an automated system that is both theory-guided and knowledge-driven – a Twitterbot called @MetaphorMagnet – navigates this course for itself. Modular theoretical frameworks, such as the General Theory of Verbal Humor (GTVH) used here, have their sectional boundaries severely tested by the task of squeezing provocative figurative forms into the textual confines of a tweet. When packing so much into so little, every single aspect of language production – from scenario construction to schema selection to framing strategy to word choice – must work so closely with every other that no single concern can ever be truly autonomous or “modular.” Twitter is not just the best of bot worlds then, but an ideal environment in which to study the cognitive linguistics of playful language.

References Brants, Thorsten & Alex Franz. 2006. Web 1T 5-gram database, Version 1. Linguistic Data Consortium. Burroughs, William S. 1963. The cut-up method. In LeRoi Jones (ed.), The moderns: An anthology of new writing in America. New York: Corinth. Carbonell, Jaime G. 1981. Metaphor: An inescapable phenomenon in natural language comprehension. Report 2404. Pittsburgh: Carnegie Mellon Computer Science Dept. Chamberlain, William & Thomas Etter. 1983. The police-man’s beard is half-constructed: Computer prose and poetry. New York: Warner Books. Colton, Simon, Jacob Goodwin & Tony Veale. 2012. Full-FACE Poetry Generation. Paper presented at the 3rd International Conference on Computational Creativity, University College Dublin, May 30-June 1. Dewey, Caitlin. 2014. What happens when @everyword ends? Intersect, Washington Post, May 23. http://www.washingtonpost.com/news/the-intersect/wp/2014/05/23/whathappens-when-everyword-ends/ (accessed Nov 27, 2015). Fass, Dan. 1991. Met*: a method for discriminating metonymy and metaphor by computer. Computational Linguistics 17(1): 49–90. Gentner, Dedre, Brian Falkenhainer & Janice Skorstad. 1989. Metaphor: The good, the bad and the ugly. In Yorick Wilks (ed.), Theoretical issues in NLP, Hillsdale, New Jersey: Lawrence Erlbaum Associates.

90 

 Tony Veale

Gervás, Pablo. 2000. Wasp: Evaluation of different strategies for automatic generation of Spanish verse. Paper presented at the AISB-2000 Symposium on Creative & Cultural Aspects of Artificial Intelligence. University of Birmingham, April 17–18. Ghosh, Aniruddha, Guofu Li, Tony Veale, Paolo Rosso, Ekaterina Shutova, John Barnden & Antonio Reyes. 2015. SemEval-2015 Task 11: Sentiment Analysis of Figurative Language in Twitter.  Paper presented at SemEval-2015, the 9th International Workshop on Semantic Evaluations, Denver, Colorado, June 4–5. Giora, Rachel, Ofer Fein, Jonathan Ganzi, Natalie A. Levi & Hadas Sabah. 2004. Weapons of mass distraction: Optimal innovation and pleasure ratings. Metaphor and Symbol 19(2): 115–141. Glucksberg, Sam. 1998. Understanding metaphors. Current Directions in Psychological Science 7: 39–43. Grady, Joseph. 1997. Foundations of meaning: Primary metaphors and primary scenes. Berkeley, University of California PhD dissertation. Hofstadter, Douglas. 1995. The ineradicable Eliza effect and its dangers. In Douglas Hofstadter, Fluid concepts and creative analogies: Computer models of the fundamental mechanisms of thought (Preface 4), New York: Basic Books. Hutton, James (translator). 1982. Aristotle’s poetics. New York: Norton. Lakoff, George & Mark Johnson. 1980. Metaphors we live by. Chicago, Illinois: Chicago University Press. Martin, James H. 1990. A computational model of metaphor interpretation. New York: Academic Press. Manurung, Ruli, Graeme Ritchie & Henry Thompson. 2012. Using genetic algorithms to create meaningful poetic text. JETAI 24(1): 43–64. McGlone, Matthew S & Jessica Tofighbakhsh. 1999. The Keats heuristic: Rhyme as reason in aphorism interpretation. Poetics 26(4): 235–44. McGlone, Matthew S & Jessica Tofighbakhsh. 2000. Birds of a feather flock conjointly (?): rhyme as reason in aphorisms. Psychological Science 11(5): 424–428. Milic, Louis T. 1971. The possible usefulness of computer poetry. In Roy A. Wisbey (ed.), The computer in literary and linguistic research. Cambridge, UK: Publications of the Literary and Linguistic Computing Centre of the University of Cambridge. Oulipo. 1981. Atlas de littérature potentielle. Paris, France: Gallimard. Orwell, George. 1946. Politics and the English language. Horizon 13(76), April. Pullum, Geoffrey. 2003. Phrases for lazy writers in kit form. Language Log. http://itre.cis. upenn.edu/myl/languagelog/archives/000061.html (accessed November 29, 2015). Queneau, Raymond. 1961. Cent mille milliards de poèmes. Paris, France: Gallimard. Queneau, Raymond. 1981. Exercises in style. 2nd Edition (Translated from the French by Barbara Wright). New York: New Directions Books. Reddy, Michael J. 1979. The conduit metaphor: A case of frame conflict in our language about language. In Andrew Ortony (ed.), Metaphor and thought, 284–310. Cambridge, UK: Cambridge University Press. Ricks, Christopher B. 1980. Clichés. In Leonard Michaels & Christopher Ricks (eds), The state of the language. Berkeley, California: University of California Press. Ricks, Christopher B. 1995. The force of poetry. Oxford University Press. Veale, Tony & Mark T. Keane. 1992. Conceptual Scaffolding: A spatially founded meaning representation for metaphor comprehension. Computational Intelligence 8(3): 494–519. Veale, Tony & Mark T. Keane. 1997. The competence of sub-optimal structure mapping on ‘hard’ analogies. Paper presented at IJCAI’97, the 15th International Joint Conference on Artificial Intelligence, Nagoya, Japan, August 22–29.



The shape of tweets to come 

 91

Veale, Tony & Guofu Li. 2011. Creative introspection and knowledge acquisition. In Proc. of AAAI-2011, Paper presented at the 25th Conference of the Association for the Advancement of Artificial Intelligence, San Francisco, California, August 7–11. Veale, Tony & Guofu Li. 2013. Creating similarity: Lateral thinking for vertical similarity judgments. Paper presented at ACL 2013, the 51st Annual Meeting of the Association for Computational Linguistics, Sofia, Bulgaria, August 4–9. Veale, Tony. 2013. Less rhyme, more reason: Knowledge-based poetry generation with feeling, insight and wit. Paper presented at ICCC 2013, the 4th International Conference on Computational Creativity. University of Sydney, Australia, June 12–14. Veale, Tony. 2014. Running with scissors: Cut-ups, boundary friction and creative reuse. Paper presented at the 22nd International Conference on Case-Based Reasoning, Cork, Ireland, September 29 – October 1. Way, Eileen Cornell. 1991. Knowledge representation and metaphor: Studies in cognitive systems. Dordrecht, The Netherlands: Kluwer Academic Press. Weizenbaum, Joseph. 1966. ELIZA – a computer program for the study of natural language communication between man and machine. Communications of the ACM 9(1): 36–45. Wilks, Yorick. 1978. Making preferences more active. Artificial Intelligence 11(3): 197–223.

Elizabeth Holt

4 “This system’s so slow”: Negotiating sequences of laughter and laughables in call-center interaction Abstract: This chapter examines sequences of interaction in calls to a gas-supply company. The sequences follow long silences as employees wait for the computer to access the customers’ account details. Employers then comment on the slowness of the system and are responded to with laughter. In the majority of instances there are further references to the system that constitute a move away from the serious business of the call and are accompanied by laughter. Thus, there is a move from explicitly institutional talk to less formal interaction. In this delicate environment – i.e. where the slowness of the system leads to a hiatus in dealing with the customers’ requests – they evince moments of affiliation and carefully negotiated transitions away from enacting institutional roles. Thus, they demonstrate that what might loosely be called ‘non-serious’ interaction is not necessarily a complete departure from serious interaction: serious actions are achieved. Further, it is possible to see how they are the product of subtle negotiation over turns rather than sudden and absolute transitions.

1 Introduction Recurrently in interaction sequences of talk occur that may be loosely termed ‘playful’, ‘humorous’ or ‘non-serious’. These are, in the first place, lay categories. They are vague and overlapping. What we might term ‘playful’ contributions to talk clearly covers a broad spectrum of utterances. At its most clear-cut end occur instances whereby speakers collaborate to produce sequences of talk that are not ‘real’ (e.g. they enact characters [Holt 1999] or produce actions that are obviously not to be taken at face value [Huth this volume]) (see also Bateson 1972). Humorous interaction is often deemed to be characterized by certain kinds of incongruity (see for example Mulkay 1988, Attardo 2004). The term ‘non-serious’ has also been used to refer to related types of contributions to interaction (Holt 2013). Our task as analysts is to learn more about the phenomena they describe, and to use that knowledge to reflect on the application and implications of these terms. Ultimately it might be possible to develop technical understandings of some of them (Holt 2016). The sequences on which this chapter focuses fall in the gray area between clearly non-serious contributions and serious ones. As such they represent the m ­ ajority of DOI 10.1515/9781501503993-005

94 

 Elizabeth Holt

instances that we might intuitively recognize as relevant to an understanding of ‘playful’, ‘humorous’ or ‘non-serious’ interaction: clear-cut cases of play or humor are less prevalent than these ambivalent contributions. Their analysis demonstrates a number of factors relevant to the study of these phenomena in interaction. First, they show that many instances are not clearly non-serious or serious, but combine elements of both. Second, they show that attempts to identify or define relevant contributions on the basis of a single attribute, such as incongruity, are in danger of missing much of their complexity. Third, that a sequential approach is necessary to a full understanding of these contributions: the sequences analyzed here demonstrate that shifting from serious talk to less serious interaction is the product of careful negotiation and collaboration over several turns, and that more overtly non-serious contributions occur following more ambivalent ones that pave the way. Fourth, laughter is a crucial part of these sequences, its analysis demonstrates its role in the negotiation and creation of a shifts towards non-seriousness. The sequences in question are drawn from a corpus of calls to a gas-supply company. The turns under analysis follow long silences as employees wait for the computer to access the customers’ account details. Employers’ comments about the slowness of the system are responded to with laughter and, in the majority of instances, further references to the system that are less serious and are accompanied by laughter. Here is an example. (1) [Ramsay:A:47:2:15/3/96] (E is the company employee, C is the caller. C has received a very high bill) 49 C: is (.) °sorry° (.) two (.) two (.) two (.) 50 nine (.) two two (.).hh treble oh one 51 (.) 52 E: °right (let’s see)° 53 (2.0) 54 E: just waiting fer this (.)particular 55 system to get on the mo:ve 56 ((7.0 sound of E tapping keyboard)) 57 E: this is one of the slo:west systems that 58 we’ve got 59 C: ↑Uhhhhhuh n(h)o pro(h)blem(h).hhhhhh (.) 60 61 C: it’s just that I nearly had a heart attack 62 when I got this (one) °huh huh huh° .hh[hh 63 E: [I can 64 quite understand why:. 65 (.) 66 E: now (.) I think wha- (.) what the details…



“This system’s so slow” 

 95

At the start of the extract the company employee takes the customer’s reference number in order to call up their details on her screen. At line 53 there is a two second silence, then the employee offers an account. At line 56 there is a seven second silence. The employee then offers another account, assessing the system as “one of the slowest systems that we’ve got”. The customer laughs and then says “no problem”, interspersed with beats of laughter. He then goes on to describe his reaction to receiving the large bill the company sent, followed by more laughter. His assessment of his reaction to receiving the high bill is characterized by overstatement which contributes to its non-seriousness (Ford and Fox 2010; Holt 2011, 2013). The laughter in this (and the other instances in the collection) is crucial in its construction as ‘non-serious’. Laughter is often the clearest indication (for participants and analysts) of playfulness or non-seriousness. Thus, research on laughter is relevant to our endeavor. There is a long tradition of such research in conversation analysis. However, it demonstrates the need for caution in assuming too close a relationship. CA has shown that laughter is recurrently associated with actions that are not humorous, playful or non-serious (such as troubles-tellings [Jefferson 1984] and complaints [Holt 2012]). Thus, researchers remain agnostic about the nature of the target of the laughter, using the term ‘laughable’ to refer to it. Laughter is seen as an important contribution to interaction in its own right rather than as an indication of the presence of humor or play. But this is not to say that analyzing the laughter is not crucial to an understanding of sequences of playful interaction. In such a conception we move away from the simple assumption that laughter follows humor, to a mutual constitution model that suggests that the occurrence of laughter marks its referent (usually retrospectively) as laughable – and, potentially, as humorous. Funniness becomes understood not as an inherent property of a message, or the internal state of a social being, but rather as a jointly negotiated communicative accomplishment. (Glenn 2003: 33)

The collection of extracts on which this chapter is based underlines the fact that laughter is not merely or even necessarily a reaction to humor, but it does subtle work in constituting turns as potentially non-serious. The sequences are characterized by laughter and laughables. But they are not purely non-serious: serious work gets done here. Further, there is no sudden transition into non-seriousness. Rather, there is a process of subtle negotiation. The sequences begin with a turn which has no laughter as part of its construction. It is multifaceted; performing a number of serious actions. But, in each instance, this initial turn is responded to with laughter. Also, in most instances, turns more explicitly marked as laughables and further laughter continue these sequences. Thus, these extracts provide insight into the subtle ways in which sequences involving laughter and

96 

 Elizabeth Holt

less serious turns are managed. They demonstrate that no one feature of the talk (e.g. incongruity) can be used on its own to explain the presence of the laughter. Rather, turns are multifaceted and ambiguous as to their actions, and the laughter cements potential moves in prior turns away from more serious talk. But the transition away from seriousness is not complete: in these sequences serious actions are performed that contribute to the contiguity of the interaction and the relationship between the speakers.

2 The data This chapter is based on investigation of a corpus of calls to a gas supply company. The company deals mainly with business accounts, although some homeowners with larger consumptions also come under its jurisdiction. The calls are initiated by the customers. The majority of calls are about issues to do with the customer’s account, such as disputing a bill, following up a previous enquiry, chasing missing bills or invoices, wanting to pay off an outstanding amount before the end of the financial year, notifying the company of a move in premises, and requesting information about gas consumption. In most cases callers refer to complainable matters such as missing letters, failure of the company to respond to enquiries, problems in contacting the appropriate employee or discrepant information. Most calls initially follow a similar pattern with customers explaining their reason for calling and employees eliciting information (account numbers or addresses) to enable them to access details of the customers’ account on the computer at the beginning of the conversation. Calls in the corpus are taken by three employees (two men and one woman). In the nine extracts that constitute the current collection, company employees/ call-takers refer to the slow running of the system. This is followed by laughter from the customer/call-maker. A broad characterizations of the sequence is as follows: A long silence Employee: Reference(s) to the slowness of the system, built without laughter. Caller: A laugh response, often accompanied by verbal affiliation Employee/Caller: In most instances, at least one further laughable with laughter, followed by affiliation or further laughter. Employee: Termination of the sequence with a return to business oriented talk I use conversation analysis to investigate these sequences. I explore the design of the turns and their contribution to the ongoing talk. I begin with consideration of the first few turns – where call-takers assess the system and callers laugh in response. I consider what may occasion the laughter: in other words, I examine the



“This system’s so slow” 

 97

action and design of the turns and why laughter may be an appropriate response. In analyzing the responses I focus on the action of the laughter and its relationship to the prior turn. Finally, I consider further contributions to the sequences and how actions of the first turns are continued and extended in subsequent turns.

3 The sequence The sequence on which this chapter focuses occur in a delicate environment: waiting for the computer to access the relevant information leads to a hiatus in the proceedings and a long silence in the interaction. This is illustrated by the following three extracts. (2) [Ramsay:A:20:14:26/3/96] 1 E: let me just check for you 2 → (15.0) 3 E: I do apologise for having to wait so long 4 (.) 5 E: the computer system’s going very 6 slow[today 7 C: [Oh its alright hhuhhuh (we 8 hav[e the same problem) 9 E: [it- £well it normally does to be (h) 10 hon(h)es[t£ 11 C: [hh-hehehe.hh 12 (9.0) 13 E: er Barland (Paint Company) (3) [Ramsay:A:127:45:14/3/96] 1 E: okay just have a look (.) °oh four one three° 2 → (3.5) 3 E: this system is so: slo:w today 4 C: hhuh huh huh .hhhhhh 5 E: right okay er which amount… (4) [Ramsay:A:78:46:14/3/96] (At the start C is offering further information from the bill to try to help locate the account) 1 C: if [you want that] 2 E: [wha– what ]is that 3 C: it’s one five nine two one six

98 

 Elizabeth Holt

4 → (3.0) 5 E: ( )it’s okay it’s just a matter of getting 6 the system to actually mo:ve som[etimes 7 C: [hhehh oh right 8 (hh)I know the feelin’ .hhehh 9 (7.0) 10 E: ( )the tenth of August… In these excerpts, and in extract (1) above, long silences occur as a result of the employee waiting for the appropriate information to appear. Thus, there is a hiatus in the proceedings and the wait is a potential source of frustration and awkwardness for the participants. These are followed by the employee’s assessment of the slow running of the system in which the employee orients to this potential annoyance by accounting for the wait and in (2) explicitly apologizing. In each instance there is then an affiliative response by the caller including or comprising of laughter. In some instances the environment in which these sequences occur is doubly delicate because callers have rung to make a complaint which they do at the start of the call1. The conversation from which Extract (1) is taken is a case in point. (1) [Ramsay:A:47:2:15/3/96] 1 E: good morning ((company name)) ↑how may I help. 2 C: oh good morning er: .hh I was speaking from 3 ((place name)). hh er regarding (.) 4 er an invoice I've received this morning 5 E: Oh kay 6 C: er (2.0) I- I- I- I just don't understand it at all it's an invoice for four hun- four thousand 7 8 two hundred and four pounds seventy 9 E: ri::↑ght. okay .hh if you can (have a note 10 of) your name and the reference number I’ll 11 have a look at it for you 12 C: yeh .hh erm can I just explain to you first of all I rece:ived a gas bill 13 14 ( ) gas bill .hh [for= 15 E:     [right 16 C: =six hundred and two pounds seventy five. (.) 17 er we were away on holiday when this arrived 1 In some the reason for the call has not become apparent at this stage since callers begin by giving their referencing number before stating why they are calling.



“This system’s so slow” 

18 .hh dated the fifth of March .h[h er 19 E: [°righ-° 20 date of reading fourth of March it says but it 21 was an estimated .hhh and then on the seventh 22 of March I received an amended account ·hh 23 which says this bill replaces previous 24 account .hh and there was a credit for 25 nine hundred and twenty one pounds ninety 26 eight .hh leaving me twelve pence credit. 27 .hh and it said cheque to follow (.) .hh on this. (.) and then this morning ·hh I 28 29 received (.) from your phone number .hhh erm 30 the er bill for four thousand two hundred and 31 four pounds seventy. (.) which I you know I 32 just don’t understand because we pay 33 qua↑rterly 34 E: right yes it’s its looks like there’s 35 something obviously going on on there 36 C: [·hhhh 37 E: [bu- as I say I mean we don’t (with no) 38 reference number I I can’t have 39 a look [at anyth[ing to see what’s going= 40 C: [.hh [er41 E: =on [so if you’d like to give me those details 42 C: [ri43 yeah now from your copy you mean 44 E: yea[h 45 C: [yea- ·hh right the (.) er:::m invoice 46 reference number 47 E: right. (2.0) 48 49 C: is (.) let’s (see) (.) two (.) two (.) two (.) 50 nine (.) two two (.) .hh treble oh one 51 (.) 52 E: °ri:ght let’s see° 53 (2.0) 54 E: just waiting fer this (.)particular 55 system to get on the mo:ve 56 ((7.0 sound of tapping keyboard)) 57 E: this is one of the slo:west systems that

 99

100 

 Elizabeth Holt

58 59 C: 60 61 C: 62 63 E: 64 65 66 E:

we’ve got ↑Uhhhhhuh n(h)o pro(h)blem(h).hhhhhh (.) it’s just that I nearly had a heart attack when I got this (one) °huh huh huh° .hh[hh [I can quite understand why:. (.) now (.) I think wha- (.) what the details…

Beginning at line 3 the caller starts to explain the nature of the problem he has rung to report; that he has received an inaccurate and extremely high gas bill. The employee does not directly address the problem straight away but asks for his name and reference number (lines 9 to 11). The caller, however, continues to detail his complaint, and the employee asks again for the reference number at lines 37 to 41. The caller gives his number at lines 49 to 50 followed by a two second pause, the employee’s first reference to the slow running of the system, a seven second pause and a further reference to the speed of the system at lines 57 and 58, before the customer’s laughter and affiliation at line 59. Thus, these latter turns occur in an environment of some delicacy: the caller has rung with a complaint about an erroneous bill, which he has detailed at some length to the employee despite her request, early in the call, for his name and reference number as a prerequisite to investigating the situation. Then waiting for the computer to access the information results in several seconds of silence which has the potential to be awkward for both participants and frustrating for the customer. Thus, it is interesting to note that in these inauspicious circumstances the customer produces laughter as part of an affiliative turn at line 59. The occurrence of laughter in a delicate environment is not, however, unprecedented. In fact, conversation analytic research on laughter has demonstrated a recurrent relationship between them (see for example Alasuutari 2009; Fatigante and Orletti 2013; Haakana 2001; Osvaldsson 2004; Wilkinson 2007). Laughter within a turn can modify or have some impact on the talk that it accompanies. For example, Jefferson (1984) showed how laughter in troubles-tellings can do troubles-resistance. Laughter can also occur as a response. In complaints, minimal, often equivocal laughs are somewhat affiliative: they fail to explicitly contribute to the complaint, being disengaged and termination-relevant, without explicitly disagreeing with the assessment they follow. In the instances in this collection, however, the laughter is more affiliative. Further, it is regularly followed by contributions that are more explicitly affiliative and extend the sequence whereby participants retreat from business-oriented interaction to talk that is less formal and non-serious.



“This system’s so slow” 

 101

The following extract shows that following an extended silence, an apology and an assessment of the system the participants collaborate to produce an extended sequence involving affiliation and laughter. (5) [Ramsay: 7.5.96 B JO 2.34] 1 (8.0) 2 E : okay so th↑at’s all been changed.= 3 C: =°right° 4 E: sorry I have to go through certain procedures and 5 it’s a bit long winded 6 C: ↑ heh heh 7 E: >hehe< that’s the problem though with 8 computers [°isn’t it.° 9 C: [yes I kno:w he he 10 E: not designed with computer and the users in mind. 11 C: I know [honestly I g- I get so: cross be[cause 12 E: [he he   [he he he 13 C: they’re so slow at times [aren’t they 14 E: [well I’ve got customer’s 15 on the other (hhh)e(hhh)n[d 16 C [y(hhh)e he he [he 17 E [>hehe< 18 E but I can’t get cross whereas ££the customer’s 19 get cross with me££ 20 C oh well I get very cros[s I have (other) t]hings 21 E [ >hehe< he ] 22 C I am sure it’s beyo- it’s beyond me so he he 23 (he)= 24 E =okay so I’m just going to go in and change the 25 er (1.0) er name (can do) >can I have the new 26 business name< Prior to this extract the caller has asked the employee to change her billing address. There is then a long silence as he does this on the system (line 1). At line 2 he reports that the address has been changed. He then apologizes for the delay and negatively assesses the system. The customer laughs and the employee also laughs then adds a further comment about computers in general. The caller affiliates and laughs again (line 9). In line 10 the employee continues to complain about computers. The caller adds to this complaint by assessing her own experience and negatively assessing computers at lines 11 and 13, while the employee

102 

 Elizabeth Holt

produces several beats of overlapping laughter. This affiliative sequence is extended over several more turns as the participants continue to refer to their experiences of problems caused by computers. So, in this extract a delicate environment – a long silence followed by an apology from the employee - leads to an extended sequence of affiliation, laughter and contributions that move away from directly dealing with the business at hand to sharing experiences. Analysis of the turns in this and the other extracts in the collection throws light on how the participants manage these transitions from orienting to the business at hand in potentially delicate environments to collaboratively creating sequences of less serious talk and laughter. Thus, in the following sections I track through the sequence exploring the contribution of the turns and showing how it can result in a departure from clearly institutional talk towards sequences of affiliation and less formal interaction.

4 Assessments of the system Following the silences the sequences begin with turns referring to the slowness of the system which are responded to with laughter. Yet these first turns do not have laughter as part of their construction. Thus, the laugh-response is not orienting to inviting laughter (Jefferson 1979) in the prior turn. It is pertinent, then, to consider what makes laughter an appropriate response2 and in this way begin the transition away from business-oriented talk to less serious turns. Though these turns do not have laughter as part of their construction, it is possible that elements of their design contribute towards making laughter an appropriate response (along with other appropriate responses) (Holt 2011). In terms of identifying recurrent features of both their design and action, we can begin by noting that they consist of either implicit or explicit assessments of the speed of the computer system. In the first three extracts the assessment of the system is explicit. (1) 57 E:→ this is one of the slo:west systems that 58 we’ve got

2 Of course laughter can be inappropriate, but in these instances the laughter is treated as appropriate. In fact, as shown, participants capitalize on the laughter by adding further laughter and laughables.



“This system’s so slow” 

 103

(2) 5 E:→ the computer system’s going very 6 slow[ today (3) 3 E:→ this system is so: slo:w today In these extracts the call takers assess the speed of the system in a negative and upgraded manner: in each case the system is not just described as ‘slow’ but as “very slow”, “so: slo:w” and “one of the slo:west systems”. In their analysis of a collection of informal face-to-face interactions involving reciprocal laughter, Ford and Fox (2010) identified exaggeration, assessments and “extreme description” (p.362) as recurrent features of turns treated as laughables. Similarly Drew (1987), in his analysis of teasing in interaction identified overstatement as a recurrent aspect of their design (see also Holt, 2011). Thus, upgraded turns are sometimes associated with turns treated as laughables. In extracts (1) to (3) the upgraded assessments of the system are explicit. In other extracts the assessments are more implicit. (4) [Detail] 5 B: ( )it’s okay it’s just a matter of getting 6 the system to actually mo:ve som[etimes (6) [Ramsay:B:00:67:18/3/96] 1 E: …let’s go and have a look at the (bill that I’m 1 looking at) it takes a little bit longer to get 2 into the bills ha- we’ve (.) got a screen 3 that tells us all the meter readings. 5 C: ri:ght 6 E: erm I was looking on there: (0.5) erm let me go 7 and have a look in the bill[: 8 C: [okay 9 (0.2) 10 E: ble:ssed (system) 11 C: huh huh huh [.hhh 12 E: [takes a bit 13 longer [to get into the bills] In (4) in the way this turn is formulated – “actually mo:ve sometimes” – there is an implication that the system is slow. Similarly, in (6) implicit in E’s mild profanity – “ble:ssed (system)” – is a negative assessment of the computer system,

104 

 Elizabeth Holt

which is made more explicit following C’s laughter with “takes a bit longer to get into the bills”. These explicit or implicit negative assessments of the system contribute to the potential action of the employees’ turns in a number of ways. First, by assessing the system in this way they act as potential complaints. Extract (3) is a succinct instance with the assessment in line 3 constituting the entirety of the employee’s complaint regarding the workings of the system. (3) [Detail] 2 (3.5) 3 E: this system is so: slo:w today 4 C: hhuh huh huh .hhhhhh 5 E: right okay er which amount… In other extracts, however, the assessment contributes towards a longer sequence with further complaint-relevant units regarding the workings of the system. Extract (1) is a case in point. (1) [Detail] 53 (2.0) 54 E: just waiting fer this (.)particular 55 system to get on the mo:ve 56 ((7.0 sound of E tapping keyboard)) 57 E: this is one of the slo:west systems that 58 we’ve got 59 C: ↑Uhhhhhuh n(h)o pro(h)blem(h).hhhhhh Following the two second pause at line 53 E gives an account which is also an implicit negative assessment of the system – “waiting fer this (,) particular system to get on the mo:ve”. Then, after another long pause, she adds a more explicit negative assessment, thus making a more overt complaint. This extract also clearly illustrates another action of the assessment components and the sequences they contribute to: by explaining the cause of the delays, they also act as accounts. Thus, lines 54 and 55 of extract (1) explicitly account for the prior delay, and lines 57 and 58 add to this by assessing the speed of the system. In extract (2) the assessment unit again acts as an account, this time, contributing to a sequence in which the employee first apologizes for the delay. (2) [Ramsay:A:20:14:26/3/96] 2 (15.0) 3 E: I do apologise for having to wait so long



“This system’s so slow” 

4 5 E: 6 7 C: 8

 105

(.) the computer system’s going very slow[ today [Oh its alright hhuhhuh (we hav[e the same problem)

Following the 15 second delay at line 2 E apologizes and describes the wait as “so long”. He then adds another unit in which he further accounts for the delay and negatively assesses the computer system. Thus, the turns consist of upgraded negative assessments regarding the workings of the company’s computer system. These negative assessments are complaint-relevant in that they refer to an aspect of the employee’s working environment which is impacting their capacity to expedite the caller’s request. At the same time, they also act as accounts for the often long silences that precede these turns. In so doing these assessments of the system and the sequences to which they contribute constitute a departure away form more explicit orientation to the task at hand. Prior to the silences and talk about the system, the participants engage in answering the caller’s request: the talk focuses on the business at hand. However, turns involving assessments of the system constitute a step away from direct orientation to the caller’s request. To illustrate this, consider again the talk that precedes extract (1). This shows how the interaction leading up to the silences and the assessment of the system is oriented to the business at hand: the caller states the nature of his enquiry and the employee ascertains his reference number in order to solve the problem regarding the discrepant information he has received about his account. Thus, talk up to this point is institutional interactional in nature. Drew and Heritage (1992) outline a number of elements that characterize institutional talk. The first of these is as follows: Institutional interaction involves orientation by at least one of the participants to some core goal, task or identity (or set of them) conventionally associated with the institution in questions. In short, institutional talk is normally informed by goal orientations of a relatively restricted conventional form. (Drew and Heritage 1992: 22, original emphasis.)

Talk in the call leading up to the silence at line 53 is oriented to the goal of first establishing and then beginning to try to solve the caller’s problem involving the conflicting and inaccurate information he has received. However, talk following the pause is less explicitly oriented to this reason for call. It is not a total departure: the call-taker’s assessment of the system at lines 57 and 58 accounts for the wait and her inability to continue working to solve his problem, and in this respect it does orient to the business at hand. However, by assessing the system it is a departure from explicit orientation to the reason for call: it refers to her

106 

 Elizabeth Holt

working environment. Further, the departure becomes more overt as the sequence progresses. At line 54 the employee begins with an account for the pause -”just waiting fer this (.) particular system to get on the mo:ve”. This is slightly more oriented to the business at hand, as she explicitly accounts for delay in dealing with the current request. The assessment at line 57 and 58, following the longer pause, creates a further step away in that she assesses the system without explicit orientation to the current problem. As well as contributing to a step away from explicit orientation to the business at hand, the turn at lines 57 and 58 contrasts with talk leading up to the sequence in a number of other ways. First, it focuses on circumstances concerning the employee’s experience. Goffman (1959) describes people’s behavior in settings as falling into two broad types: backstage and onstage behavior. When onstage performers present an idealized version of the role, engaging in what might be seen as ‘professional’ behavior. In contrast, backstage a performer may have more license to behave in ways that could be deemed unacceptable front stage, such as complaining about customers (in roles such as waiter/waitress for example), and joking about onstage events. In applying this distinction to the gas supply call center, the backstage would equate to conversations between colleagues, for example (Holt, 1999). Conversations between employees and customers are ‘onstage’. However, distinctions between these two domains are not neatly bounded. Thus, the metaphor is useful in seeing talk in these turns as moving from less clearly onstage behavior to slightly more backstage performance, whereby the employee provides some insight into the ‘behind-the-scenes’ workings of the company. Similarly Goffman’s (1961) metaphor ‘role distance’ might also provide some insight. The call-taker takes a step away from simply enacting the role of gas supply company employee by implicitly complaining about the company’s system, and thereby subtly and slightly distancing themselves from the role. In conversation analysis identity or role is recognized as salient to understanding participation in interaction. These roles or identities are seen as occasioned moment-by-moment and oriented to in the talk. Interactionally relevant identities can range from ones linked to the action and participation of the interactant, such as speaker, storyteller, troubles-tellers, to gender and place identities. Laughter and laughables have been found to be central to the construction of these identities (see Clift 2013; Jefferson 1984, 2004; Liebscher and Dailey-O’Cain 2013). In the extracts here employees orient to their institutional identities as they attempt to solve the callers’ requests. But, in the turns involving negative assessments of the system, they begin to take a subtle step away by evoking other identities such as complainant. Thus, by assessing the system these turns give an insight into the employee’s experience and the workings of the institution. Whilst previous talk has related



“This system’s so slow” 

 107

to the customer’s request, this constitutes a shift in focus towards the employee’s concerns. It creates a transition away from a strictly institutional role towards something more informal. Further, in sharing potentially complaint-relevant information about the workings of the company with the caller it distances the call-taker from the organization. This transition to less explicitly institutional talk is contributed to by a shift to less formal language: thus, the design of the turns as well as the actions they perform coalesce to manage this shift. In (1) “get on the mo:ve” is less formal than much of the preceding talk. Extract (5) exemplifies this transition even more clearly. (6) [Detail] 6 E: erm I was looking on there: (0.5) erm let me go 7 and have a look in the bill[: 8 C: [okay 9 (0.2) 10 E: ble:ssed (system) 11 C: huh huh huh [.hhh The mild profanity at line 10 constitutes a transition to less formal talk. Intonation can also help to distinguish these sequences. In (3) there is a noticeable change of tone by the employee at line 3. (3) [Detail] 3 E:→ this system is so: slo:w today In contrast to E’s preceding talk, the assessment is delivered with more emphasis and at a slower tempo. The emphasis on “slo:w” contributes to the upgraded nature of the assessment, and the entire turn is more animated than previous talk. Thus, both the nature of the talk and its delivery contribute to a move away from more explicitly goal-driven institutional interaction to talk that is a little less formal and more conversational. So, the turns referring to the slow running of the system are complex and multifaceted. They perform a series of actions: they negatively assess the system, and, in so doing act as potential complaints and accounts for the pauses. At the same time they initiate or contribute to a step away from explicitly institutional oriented talk towards talk that is more informal and involves a shift away from orientation to institutional role on the part of the employee. So we cannot explain the laughresponses in terms of a simple relationship between humor in the first turn and a laugh-response. Certain elements of these turns may coalesce to make the laughter appropriate. They consist of upgraded assessments which are recurrent features of

108 

 Elizabeth Holt

turns treated as laughables. Also, it has become clear that these turns are a departure from the more explicitly institutional goal-driven talk that precedes them, and in this respect they are somewhat incongruous. Ford and Fox (2010) also identified incongruity as a recurrent feature of interactional turns treated as laughables. But they are not explicitly ‘funny’ and they contribute to the talk in serious and potentially delicate ways. Thus, to investigate this further I now turn to the responses in order to explore both what the laughter orients to in the prior turn and its action in terms of the sequence to which it contributes.

5 The laugh-response In each of the instances in the collection, callers respond to the assessments of the system with laughter. The laughter can be unaccompanied, as in (3) and (5), or, more frequently, accompanied by lexical elements. (1) [Detail] 54 E:   just waiting fer this (.)particular 55   system to get on the mo:ve 56 ((7.0 sound of E tapping keyboard)) 57 E:   this is one of the slo:west systems that 58    we’ve got 59 C:→  ↑Uhhhhhuh n(h)o pro(h)blem(h).hhhhhh (2) [Detail] 3 E:   I do apologise for having to wait so long 4  (.) 5 E:   the computer system’s going very 6   slow[today [Oh its alright hhuhhuh (we 7 C:→ 8    hav[e the same problem) (3) [Detail] 3 E:   this system is so: slo:w today 4 C:→    hhuh huh huh .hhhhhh (4) [Detail] 5 B:  ( )it’s okay it’s just a matter of getting 6   the system to actually mo:ve som[etimes 7 C:→   [hhehh oh right 8   (hh) I know the feelin’ .hhehh



“This system’s so slow” 

 109

(5) [Ramsay: 7.5.96 B JO 2.34] 1 E:    okay so th↑at’s all been changed.= 2 C:    =°right° 3 E:    sorry I have to go through certain procedures and 4    it’s a bit long winded 5 C:→   ↑heh heh 6 E:    >hehe< that’s the problem though with 7    computers [°isn’t it.° In (1) the caller laughs and then says “n(h)o pro(h)blem(h)” with laugh particles interspersing these words. In (2) the caller responds with “Oh it’s alright” followed by laughter. In (3) three beats of laughter and a loud out-breath by the caller follow the assessment of the system. In (4) an extended breathy laugh precedes “oh right” at line 7. In (5) two beats of laughter follow the employee’s comment about the “long winded” system at lines 4 and 5. To understand turns at talk it is always necessary to consider their relationship to the ongoing sequence. However, this is particularly important in regards to laughter. Laughter is not part of the lexical code, and thus, its ‘meaning’ or, rather the nature of its contribution, is especially difficult to pin down. But, what laughter is doing becomes easier to understand when we consider its relationship to the turn it targets. As Schenkein (1972) points out, we make sense of heheh in the same/similar way as other utterables –“by reference to the activities which they are recognizably used to accomplish” (p. 345). Crucial to this process is the relationship between the laughter and the sequence in which it is situated, especially in terms of the immediately preceding or accompanying talk. Laughter is indexical; it is “heard to be tied in a most powerful way to the immediately prior utterance” (Schenkein 1972, p.365). The placement of laughter is crucial in determining its referent because it recurrently occurs after, or concurrent with, the turn that is treated as a laughable (Glenn 2003: 48–9). Thus, in order to understand laughter in talk, it is necessary to look both at the target of the laughter – to consider how it may occasion the laughter- and the laughter (and any accompanying talk) to explore the nature of its contribution to the ongoing sequence. In these instances an important contribution of the laughter is affiliation. The association between laughter and affiliation has been noted by many authors (including Clarke and Wilkinson 2009; Kangasharju and Nikko 2009; Walker 2013). According to Schenkein (1972: 371) laughs: are one of the ways persons can go about proffering or displaying affiliations with one another in the course of some conversation-in-progress. That some second-speaker hehe can be heard to support some intendedly nonserious first- speaker’s utterance reveals on its occurrence a coincidence of thought, attitude, sense of humor, and the like.

110 

 Elizabeth Holt

However, this is not to say that laughter is always affiliative. It can also be disaffiliative (see for example Romaniuk, 2013; and Ticca, 2013) or somewhere between these two extremes (see, for example, Jefferson, Sacks and Schegloff 1987; Greatbatch and Clark 2003; Holt 2012). The fact that it is non-verbal means that it can be ambiguous as to its affiliative status. In an analysis of minimal laughter following a complaint about a third party, Holt (2012) found that it was used to resist fully aligning or affiliating with the complaint while still maintaining some social cohesion. In these instances the laughter affiliates with the employee’s comment on the system. As has become clear, the assessments of the system occur in delicate environments: delays prevent the employees from continuing to expedite the callers’ requests. Further, in most of the calls (as in [1] above) callers report complaint-relevant matters concerning their interaction with the company in the lead up to the hiatuses caused by the workings of the system. Following the negative assessments, callers can respond in several ways: they can remain silent; treat the prior turn as news; or as troubles-implicative in terms of smooth expedition of the request. By laughing the caller orients to the slow running as non-serious in terms of its troubles-implicative nature. Further, the caller aligns with the transition away from explicit orientation to the business at hand (the reason for call). The laughter also affiliates with the employee’s slight shift in participation status away from company representative to a less fully institutional persona. A more detailed consideration of extract (1) illustrates these actions. (1) [Detail] 57 E:    this is one of the slo:west systems that 58    we’ve got 59 C:→    ↑Uhhhhhuh n(h)o pro(h)blem(h).hhhhhh The high pitched, single extended beat of laughter at the start of C’s turn aligns with the previous turn as a telling, affiliating with it as newsworthy though not necessarily a serious problem. The laughter continues throughout the next component and ends in breathiness at the end of C’s turn. Thus, while producing “no problem”, which orients to the previous as an account and possibly an implicit apology, the laughter particles continue to treat it as less serious than might have been the case had it been uttered without the laughter. By responding in this way C also takes a step away from explicitly orienting to the business at hand (he does not, for example, take this opportunity to return to describing the nature of his problem). By laughing, rather than orienting to the delay as a problem, he adopts a stance that is socially cohesive, potentially creating rapport with the call-taker rather than a more confrontational approach.



“This system’s so slow” 

 111

Here, and in other extracts, the affiliative nature of the laughter is both evinced and contributed to by the fact that it is accompanied by other, less ambiguously affiliative elements. Thus in (1) the laughter is followed by “n(h)o pro(h) blem(h)”, in (2) laughter follows “Oh its alright”, and in (4) the caller laughs and then says “oh right”. These units add to the ambiguous affiliation of the laughter by orienting to the prior turn as news – in (2) and (4) the ‘oh’s act as change-ofstate particles (Heritage 1984); they orient and align with the prior as an informing or account, and as part of an apology in (1), or as an implicit apology in (2). In (2) and (4) the callers proceed (in the same turn) to produce elements which explicitly affiliate with the employee by stating that they have the same experience –“(we hav[e the same problem)” and “I know the feelin’”. Turns in which the callers empathize with the employee are recurrent in the corpus and generally follow the caller’s initial response to the employee’s assessment. Thus, they are the subject of the next section as we track through these sequences, exploring recurrent elements. They also provide further evidence of the affiliative nature of the laughter, in that they show that it often occurs in environments characterized by more explicit affiliation by the same participant following the laughter. In sum, in this section it has become apparent that laugh responses affiliate with the multifaceted turns that precede them. The laughter, rather than straightforwardly responding to the presence of some element in the preceding turn, contributes to the actions embedded within prior turns. In so doing the laughs constitute these turns as laughables and align with a move away from explicitly goal-oriented institutional interaction to something less serious. In most of the extracts, this movement is made more explicit in subsequent turns involving laughables and further laughter.

6 Further laughables Following the laughter and (sometimes) verbal affiliations by the customer, in most of the extracts one or other of the participants contributes a further laughable. These are more explicitly marked as laughables through the presence of laughter in the same turn. In (1) C assesses his response to receiving the high bill. (1) [Detail] 54 E: just waiting fer this (.)particular 55 system to get on the mo:ve 56 ((7.0 sound of E tapping keyboard)) 57 E: this is one of the slo:west systems that

112 

 Elizabeth Holt

58    we’ve got 59 C:    ↑Uhhhhhuh n(h)o pro(h)blem(h).hhhhhh 60    (.) 61 C:→    it’s just that I nearly had a heart attack 62 →    when I got this (one) °huh huh huh° .hh[hh 63 E:    [I can 64    quite understand why:. 65    (.) 66 E:    now (.) I think wha- (.) what the details… Following C’s laugh response and affiliation at line 59, the caller then adds to the sequence by describing his reaction on receiving the high bill. His turn mirrors the call-takers turn at lines 57 and 58 in several respects. It is an upgraded assessment: “nearly had a heart attack” is an exaggerated assessment of the impact of the bill. Like the call-taker’s assessment of the system it continues the transition away from explicitly task related talk towards more “backstage” information. By portraying his reaction to the bill it creates a transition away from a straightforwardly institutional role towards a more complex and nuanced one. Like E’s turn about the system, it is a potential complaint, since it assesses the impact in a highly negative way. It may also implicitly contribute towards accounting for his call and his earlier complaint relevant turns regarding the company’s communication with him. Also, like the call-taker’s assessment, it constitutes a transition to more informal talk usually associated with ordinary interaction. Figurative expressions such as this are much more common in ordinary conversation than in institutional interaction (Holt 1991). At the end of this turn, C laughs. In explicitly treating this turn as a laughable C may modify a potentially problematic action (Shaw, Hepburn, Potter 2013). His formulation of his reaction to receiving the high bill is potentially delicate in that it is an implicit complaint of his treatment by the company E represents. Thus the laughter explicitly constitutes the turn as non-serious rather than a serious complaint. Furthermore it is also a mild troubles-telling. According to Jefferson (1984) troubles tellers laugh to show they are coping. By describing his reaction to the bill in this way, as well as adding moderating laughter, C produces a turn that is not an explicit complaint aimed at E (as a representative of the company) but allows for affiliation by E over the way he has been treated. In response, the employee does not laugh but instead contributes a strong affiliation with “I can quite understand why:”. Reciprocating the laughter may have been somewhat disaffiliative in that it may not have oriented sufficiently to the complaint and troubles-telling aspects of the caller’s turn. In troubles-tellings,



“This system’s so slow” 

 113

recipients of troubles-tellers’ laughter do not laugh, thus showing affiliation with the teller (Jefferson 1984). Following E’s affiliative turn, she then initiates a transition back to the business at hand at line 18. In (1) it is the caller that extends the sequence by contributing a further laughable in the form of a turn followed by laughter. In (2) it is the call-taker that extends the sequence. (2) [Detail] 3 E:    I do apologise for having to wait so long 4   (.) 5 E:    the computer system’s going very 6    slow[today 7 C:    [Oh its alright hhuhhuh (we 8    hav[e the same problem) 9 E:→    [it- £well it normally does to be (h) 10 →    hon(h)es[t£ 11 C:    [hh-hehehe.hh 12  (9.0) 13 E:    er Barland (Paint Company) In lines 7 and 8 it seems3 that C adds a further unit about computer systems which affiliates with E by portraying himself or his company as having the same problem. Thus, he explicitly affiliates with the call-taker. His affiliation is overlapped by E’s turn at line 9 and 10 in which he adds to his negative assessment of the system. He upgrades the complaint by assessing its slow running as not just occurring “today”, but “normally”. The turn is delivered with smile voice (see Ford and Fox 2010) and there is a little interpolated laughter towards the end of the turn. Thus, this assessment is more explicitly constituted as a laughable. And, orienting to it in this way, C responds with several beats of laughter (line 11). C’s laughter brings the sequence to a close. Following it there is a nine second pause before E continues with orientation to the business at hand. Thus, in these two extracts further turns are contributed to the sequence that are explicitly marked as laughables by turn-completion laughter. Participants negotiate a more explicit transition into less serious talk, capitalizing on the first highly ambivalent turn (the assessment of the system) followed by affiliative laughter. 3 The brackets around this component indicate that it is the transcriber’s best guess at what is said here. The next turn by the employee begins near the start of this hard-to-hear component.

114 

 Elizabeth Holt

In the next extract this subtle negotiation into less serious talk is extended over several turns to create a more dramatic and extended departure from explicitly business oriented interaction. (6) [Detail] 10 E: ble:ssed (system) 11 C: huh huh huh [.hhh 12 E: [takes a bit 13 longer [to get into the bills] 14 C: [yeah I know its ]like us its not15 wul the system’s not ri:ght is it °hhe[hh .hh 16 E: [no: 17 E: no i- erm I was saying to somebody this morning 18 I’m sure it’s cos it’s got the hump (with th-) 19 with us cos we’ve got (.) two billing systems 20 in ther[e now 21 C:      [really heh h[eh heh 22 E: [yeh n’ I say- I say it’s 23 got the hump 24 C: huh huh hu[h [.hhh 25 E: [hhu[hh 26 (0.3) 27 E: yeh look’n at it it’s three seven oh five… After E’s implicit negative assessment of the system and C’s laughter, E adds an account which contributes towards explaining her mild profanity at line 10. In overlap C affiliates by claiming the same experience and assessing the system. This is followed by laughter at the end of her turn. Thus, she affiliates with E by contributing a laughable which is explicitly marked as such with turn final laughter. E continues by telling a story about a conversation she had earlier in the day where she referred to the problematic system. This includes the colloquial expression “it’s got the hump”4 (line 18). Thus, at this point she has shifted footing into talk that is more conversational and is not goal-driven institutional talk. At line 21 C produces a newsmark and three beats of laughter. E overlaps the laughter with a repeat of the figurative expression. This is followed by shared laughter. After a brief pause, E initiates a transition back to business (line 27). Thus, in this extract an initial implicit complaint about the system leads to a sequence involving further laughables and laughter. The shared laughter at lines 4 The expression means that the computer is annoyed or is sulking.



“This system’s so slow” 

 115

24 and 25 is highly affiliative and consolidates the transition towards talk that is more informal and conversational. Extracts where these sequences are extended show, then, that potential actions within the first pair (involving, for example a transition to less institutional and less serious talk) are compounded in later turns with contributions that are more explicitly laughable, less serious and move further away from goaldriven institutional talk.

7 Conclusion These extracts evince a subtle transition from explicit attention to the reason for call towards talk characterized by laughter and laughables. Subtle negotiation is evident in the fact that the early turns (in particular) are highly ambivalent in terms of their status as potential laughables. They have no laughter as part of their construction, and they perform (or potentially perform) a number of actions, including negatively assessing the system, accounting for the delay and taking a step away from explicitly institutional talk. Thus, they are multifaceted and perform serious tasks. However, they are responded to with laughter. The laughter orients to and ratifies this move away from explicitly institutional talk towards something more personal, less formal, and less serious. It affiliates with the call-taker in a tricky environment (i.e. not being able to expedite the caller’s request due lack of access to the appropriate computer screen) and in a transition away from explicitly goal-driven institutional talk. Sequential analysis of these sequences has demonstrated that laughter is central in negotiating a move from serious, goal-driven interaction towards less serious, more affiliative talk. Further, the contribution of the laughter is complex and often hard to pin down. It is not possible to identify a single feature of the prior turns that leads to the laughter. Thus, for example, there is not a simple relationship between incongruity and laugh responses. In some ways the turns accompanied or responded to with laughter are incongruous with other turns in the call. For instance, we have seen that assessments of the system contrast with turns more focused on the business at hand. They are also incongruous in that the call-taker takes a step away from their institutional persona to negatively assess their working environment. Other turns in these sequences are also incongruous, in for example, personifying the system and describing it using the colloquial phrase “got the hump”. But these incongruities do not, on their own, explain the laughter. Other aspects of turns are also relevant, and it has been productive to examine the action of both the turns responded to with laughter and the laughter itself to understand its occurrence. Further, turns responded to or accompanied

116 

 Elizabeth Holt

by laughter are not clearly humorous; rather, it is the presence of the laughter than helps constitute these turns as potentially non-serious. Authors have pointed to a relationship between laughter and the creation of intimacy or rapport (Lavin and Maynard 2001; Jefferson, Sacks and Schegloff 1987). In these extracts, potential laughables in these delicate moments create the opportunity for callers to affiliate with the employees. This is particularly useful in this environment. Not only is the immediate environment tricky in that the calltaker is temporarily prevented from continuing to solve the caller’s request, but callers generally report complaint-relevant matters in the run up to these sequences. Producing a highly ambivalent turn that makes affiliation through laughter an appropriate response can contribute towards the creation of rapport if the caller affiliates in this way. But it also provides for other responses, including less affiliative ones. It may, therefore, enable call-takers the chance “test the water”: to see whether callers are likely to adopt or continue with a less affiliative and thus what may be seen as a less ‘friendly’ stance, or to move towards a more affiliative or ‘friendly’ one. In this respect, although these sequences constitute a departure from dealing with the business at hand, they may be very productive in terms of the success of the call, both from an institutional and a social point of view.

References Alasuutari, Maarit. 2009. What is so funny about children? Laughter in parent-practitioner interaction. International Journal of Early Years Education, 17(2): 105–118. Attardo, Salvatore. 2004. Linguistic theories of humor. Berlin: Mouton de Gruyter. Bateson, Gregory. 1972. A theory of play and fantasy. In Bateson, Gregory, Steps to an Ecology of Mind, 177–193. New York, Ballantine. Clarke, Michael & Ray Wilkinson. 2009. The collaborative construction of non-serious episodes of interaction by non-speaking children with cerebral palsy and their peers. Clinical Linguistics & Phonetics, 23: 583–597. Clift, Rebecca. 2013. No laughing matter: Laughter and resistance in the construction of identity. In Phillip Glenn & Elizabeth Holt (eds.), Studies of Laughter in Interaction, 223–236. London: Bloomsbury. Drew, Paul. 1987. Po-faced receipts of teases. Linguistics, 25: 219–253. Drew, Paul. & John Heritage. 1992. Analyzing talk at work: an introduction. In Paul Drew & John Heritage (eds.), Talk at work, 3–65. Cambridge: Cambridge University Press. Fatigante, Marilena. & Franca Orletti. 2013. Laughter and smiling in a three-party medical encounter: Negotiating participants’ alignment in delicate moments. In Phillip Glenn & Elizabeth Holt (eds.), Studies of laughter in interaction, 161–184. London: Bloomsbury. Ford, Celia. E., & Barbara A. Fox. 2010. Multiple practices for constructing laughables. In Dagmar Barth-Weingarten, Elisabeth Reber & Margert Selting (eds.), Prosody in interaction,339–368. Amsterdam: John Benjamins.



“This system’s so slow” 

 117

Glenn, Phillip. J. 2003. Laughter in interaction. Cambridge: Cambridge University Press. Goffman, Erving. 1959. The presentation of self in everyday life. New York: Doubleday Anchor. Goffman, Erving. 1961. Encounters: Two studies of self in everyday life. New York: Doubleday Anchor. Greatbatch, David & Timothy Clark. 2003. Displaying group cohesiveness: Humour and laughter in the public lectures of management gurus. Human Relations, 56(12): 1515–1544. Haakana, Markku. 2001. Laughter as a patient’s resource: Dealing with delicate aspects of medical interaction. Text, 21(1/2): 187–219. Heritage, John. 1984. A change-of-state token and aspects of its sequential placement. In J. M. Atkinson & John Heritage (eds.), Structures of social action: Studies in conversation analysis, 299–345. Cambridge: Cambridge University Press. Holt, Elizabeth J. 1991. Figures of speech: An exploration of the use of idiomatic phrases in conversation. Unpublished thesis: University of York. Holt, Elizabeth. 1999. Just gassing: An analysis of direct reported speech in a conversation between employees of a gas supply company. Text, 19: 505–538. Holt, Elizabeth. 2011. On the nature of ‘laughables’: Laughter as a response to overdone figurative phrases. Pragmatics, 21(3): 393–410. Holt, Elizabeth. 2012. Using laugh responses to defuse complaints. Research on Language and Social Interaction, 45(4): 430–448. Holt, Elizabeth. 2013. ‘There’s many a true word said in jest’: Seriousness and non-seriousness in interaction. In Phillip Glenn & Elizabeth Holt (eds.), Studies of laughter in interaction, 69–90. London: Bloomsbury. Holt, Elizabeth. 2016. Laughter at last: playfulness and laughter in interaction. Journal of Pragmatics, 100: 89–102. Jefferson, Gail. 1979. A technique for inviting laughter and its subsequent acceptance/ declination. In George Psathas (ed.), Everyday language: Studies in ethnomethodology, 70–96. New York: Irvington. Jefferson, Gail. 1984. On the organization of laughter in talk about troubles. In J. Maxwell Atkinson & John Heritage (eds.), Structures of social action; Studies in conversation analysis, 346–369. Cambridge: Cambridge University Press. Jefferson, Gail. 2004. A note on laughter in “male-female” interaction. Discourse Studies, 6: 117–133. Jefferson, Gail, Harvey Sacks, & Emanuel Schegloff. 1987. Notes on laughter in pursuit of intimacy. In Graham Button & John R. E. Lee (eds.), Talk and social organization, 152–205. Clevedon: Multilingual Matters. Kangasharju, Helena & Tuija Nikko. 2009. Emotions in organizations: Joint laughter in workplace meetings. Journal of Business Communication, 46(1): 100–119. Lavin, Danielle & Douglas W. Maynard. 2001. Standardization vs. rapport: Respondent laughter and interviewer reaction during telephone surveys. American Sociological Review, 66: 453–479. Liebscher, Grit & Jennifer O’Cain. 2013. Constructing identities through laughter. In Phillip Glenn & Elizabeth Holt (eds.), Studies of laughter in interaction, 237–254. London: Bloomsbury. Mulkay, Michael. 1988. On humour: Its nature and its place in modern society. Cambridge: Polity Press. Osvaldsson, Karin. 2004. On laughter and disagreement in multiparty assessment talk. Text, 24(4): 517–545.

118 

 Elizabeth Holt

Romaniuk, Tanya. 2013. Interviewee laughter and disaffiliation in broadcast news interviews. In Phillip Glenn & Elizabeth Holt (eds.), Studies of laughter in interaction, 201–220. London: Bloomsbury. Schenkein, James N. 1972. Towards an analysis of natural conversation and the sense of heheh. Semiotica, 6: 344–377. Shaw, Chloe, Alexa Hepburn, & Jonathan Potter. 2013. Having the last laugh: On post-completion laughter particles. In Phillip Glenn & Elizabeth Holt (eds.), Studies of laughter in interaction, 91–106. London: Bloomsbury. Ticca, Anna C. 2013. Laughter in bilingual medical interactions: Displaying resistance to doctor’s talk in a Mexican village. In Phillip Glenn & Elizabeth Holt (eds.), Studies of laughter in interaction, 107–130. London: Bloomsbury. Walker, Gareth. 2013. Young children’s use of laughter after transgressions. Research on Language and Social Interaction, 46(4): 363–382. Wilkinson, Ray. 2007. Managing linguistic incompetence as a delicate issue in aphasic talk-ininteraction: On the use of laughter in prolonged repair sequences. Journal of Pragmatics, 39(3): 542–569.

Olga Zayts and Stephanie Schnurr

5 Laughter as a “serious business”: Clients’ laughter in prenatal screening for Down’s syndrome Abstract: This chapter examines the use of laughter in the context of prenatal screening (PS) for Down’s syndrome in Hong Kong. Scholars interested in humor typically approach laughter as a phenomenon that accompanies funny, amusing, and humorous situations. In our previous work on nurses’ laughter in PS (Zayts and Schnurr, 2011) we have shown that laughter may also be used to perform ‘serious business’, for example, it may be employed by nurses to help them facilitate clients’ decision-making process regarding testing for Down’s syndrome. This chapter focuses on the second part of the laughter sequences, in particular it examines what is interactionally achieved through the reciprocation of the nurses’ laughter by their clients. Drawing on 34 video-recorded consultations between nurses and pregnant women, and using conversation analysis, we show that the reciprocated laughter in these sequences can be affiliative and serve to establish rapport between the participants. It can also be disaffilitative, particularly in interactional contexts when participants engage in negotiating their epistemic and deontic statuses and authority. The negotiation of epistemic statuses is observed in consultations with more experienced and knowledgeable clients who use laughter to terminate the topic of ‘inquiring about their knowledge’. The negotiation of deontic authority happens in decision-making phases of these consultations and through laughter the clients affirm their right and sufficient knowledge to make a decision.

1 Introduction In this chapter we focus on the use of laughter in interactionally challenging situations, in particular we examine how laughter initiated by one interlocutor is responded to by another participant of an interaction. Scholars interested in humor and other creative or playful uses of language often approach laughter as an interactional phenomenon expressing or accompanying funny, amusing or humorous situations. Research in conversation analysis in the last 40 years, however, has demonstrated that laughter is a complex interactional phenomenon that performs a wide range of functions, from a simple response to humor to doing more “serious business”, particularly in contexts that are not DOI 10.1515/9781501503993-006

120 

 Olga Zayts and Stephanie Schnurr

i­ mmediately thought of as “funny” (Glenn 2003; Holt 2013). Moreover, laughter is used in highly orderly ways, and it may accompany a wide range of social activities. This chapter demonstrates the importance of examining laughter in various social activities so as not to conflate it with and limit to ‘having fun’ and ‘amusement’. The specific context that we focus on here is prenatal screening (hereafter, PS) for Down’s syndrome. Previous research in medical contexts has shown that laughter is often used to “ease” embarrassing, sensitive or painful aspects that are typical of medical encounters, and to create an alignment between participants (e.g. Haakana 1999, 2001, 2002). In our previous work on laughter in PS we have shown that laughter may be employed by nurses to help them achieve the main institutional goal of these interactions, namely facilitating autonomous choice by their clients, i.e. pregnant women, regarding whether they want to take prenatal screening tests or not (Zayts and Schnurr 2011). We have examined two specific contexts in which the nurses’ laughter regularly occurs: when the women refuse outright to undergo prenatal screening (which goes against the nurses’ beliefs that it is in the women’s best interests to take tests and to confirm whether the child they carry has Down’s syndrome or not); and when the women address direct questions to the nurses regarding what tests they should take (which interferes with the professional principle of nondirective [or non-intrusive] way of conducting PS consultations). In the first case the nurse’s laughter performs the “serious business” of “laughing off” and overcoming the women’s resistance to taking the tests; and in the second case it allows the nurses to avoid the precarious situation of having to give advice to the women when institutionally they are not supposed to do so. In this chapter we focus on the second part of these laughter sequences, namely the women’s responses to the nurses’ laughter. More specifically, we examine the interactional contexts in which the nurses’ laughter is reciprocated by the women and contrast some of those cases with cases where it is not. Our particular focus is on what is interactionally achieved through the reciprocation of nurses’ laughter by their clients in these encounters.

2 Laughing together The phenomenon of speakers “laughing together” has received a lot of attention in the conversation analytical literature. In her early work Jefferson (1979) discusses how laughter that occurs at the end of a speaker’s turn, or laughter particles within a speaker’s turn, invite laughter by another speaker. The first speaker’s laughter thus serves as a “laugh invitation” to the second speaker.



Laughter as a “serious business” 

 121

Such a laugh invitation may lead to its acceptance by the second speaker, but this is not always the case (Haakana 2002). A laugh invitation may also be declined, or it may be responded to by silence by the second speaker (in which case the first speaker may extend a laughter sequence) (Jefferson 1979). To avoid the extension of a laughter sequence, the second speaker may also respond to a laugh invitation with talk (Jefferson 1979). There are certain environments that have been widely cited in the conversation analytical literature where laugh invitations are more likely to be declined. Jefferson (1984), for example, analyzes the cases of trouble-tellings when speakers laugh to show their resistance to the trouble. The acceptance of such laugh invitations by the second speaker may be seen as not taking the trouble seriously. Another context where reciprocation of laughter may be seen as inappropriate is laughter that accompanies a speaker’s self-deprecating utterance. Reciprocating laughter in that case may be seen as an agreement with the assessment (Glenn 1991/ 1992). Some studies have also noted a more nuanced nature of responses to laughter that may include smiling or a smiling voice, and the differences in actions that these other responses perform. For example, smiling may be seen as a “milder” way of expressing speakers’ affiliation with the previous utterance when that utterance is constructed as delicate and troublesome (Haakana 1999, 2010; see also Glenn and Holt 2013). Glenn and Holt (2013) note that the analysis of whether the first laugh constitutes a laugh invitation or not requires a close analysis of the “laughable”1 (which they define as an action that makes laughter relevant or a turn that is followed by laughter) and the sequential environment of where laughter occurs. Holt (2011: 408 as cited in Glenn and Hall 2013) suggests that “[t]urns characterized by several elements recurrently associated with laughable may be more likely to be seen as inviting laughter”. By a similar token, she suggests that a speaker’s turns that are concerned with “serious business” and that contain laughter are less likely to be responded to by laughter. Extending this idea to the context of PS consultations, the cases where nurses’ laughter is reciprocated by the clients (despite these consultations being concerned with the “serious activities” of managing information- and advice- giving and decision-making about testing) become ­particularly interesting for close scrutiny. The existing research has also shown a close link between laughter and the social context of an interaction. In relation to institutional contexts, laughter appears to depend on participants’ institutional roles and responsibilities

1 The term “laughable” does not presuppose that the action or the turn that the laughable refers to is funny or amusing.

122 

 Olga Zayts and Stephanie Schnurr

(e.g. see contributions in Glenn and Holt 2013). Haakana (1999: 132) maintains that “[l]ay persons and professionals use laughter in different ways” that are related to their institutional identities. Glenn (2010: 1487) echoes this idea that “in institutional interactions the distribution and the sequential organization of laughter can reflect and constitute asymmetries of respective roles and tasks”. He maintains that these asymmetries are manifest in the distribution of laughter and its responses; and also what activities laughter accompanies. Laughing together can be seen as a strategy of building up rapport and establishing alignment, particularly in those encounters that lead to successful outcomes (see, for example, Adelswärd’s [1989] study of Swedish job interviews where more laughter was observed in those interviews that were successful). Laughter can also be a powerful resource of reinforcing the asymmetries of the participants’ roles and responsibilities. For example, in the analysis of employment interviews Glenn (2010) shows that the interviewers are typically the ones who initiate laughter, and while it may be reciprocated by the interviewees, the interviewers are the ones who decide whether to extend laughter or to bring the interaction “back to business”. Jacknick (2013) in the analysis of ESL classroom interactions shows that laughter may be used as an interactional resource to negotiate participant’s epistemic authority. In particular, the author discusses how students employ laughter to question the teacher’s answers thus challenging her authority. Through laughing (or not laughing) together, participants display their understanding of and negotiate their roles and responsibilities in an interaction. And while there are general patterns that can be observed in specific contexts, participants may adapt them to local “contingencies and activities” (Glenn 2010: 1496) of an on-going interaction. In the analysis presented below we examine what roles and responsibilities the participants negotiate and what “contingencies and activities” they orient to when they engage in shared laughter in PS consultations.

3 Data and method The data for this chapter comes from a large interactional project on prenatal screening for Down’s syndrome in Hong Kong that was conducted in 2006–2012. In that project we collaborated with a team of medical professionals from a Prenatal Diagnostics and Counselling Department in one public hospital. We focus here on 34 consultations that we also analyzed in our previous study (Zayts and Schnurr 2011) – albeit with a focus on the nurses’ rather than the patients’ ­laughter. These consultations vary in length from 9 to 22 minutes, and the overall recording time is around 7 hours. All consultations were video-recorded and



Laughter as a “serious business” 

 123

t­ranscribed using transcription conventions traditionally used in conversation analytic research (Jefferson, as published in Atkinson and Heritage 1984; ten Have 2007). The occurrences of laughter were transcribed on a particle-by-particle basis (see Jefferson 1985 for the advantages of this method in analyzing laughter). In the transcripts we also noted the sequential positioning of laughter, laughter initiation and completion points and non-verbal actions of the participants relevant to the analysis of laughter (such as smiles) (Hepburn and Varney 2013). The participants of these recorded interactions include Hong Kong Chinese nurses (abbreviated as N in the transcripts) and clients: pregnant women (abbreviated as W in the transcripts) who attended the consultation on their own or were accompanied by other family members (typically their husbands). The women in the selected consultations come from the Philippines, so in the consultations English was used as a lingua franca, or common language of the participants. We did not purposefully select the consultations with participants from the Philippines. They comprise the largest group in our corpus because of the sociodemographics of clients in the hospital where the data were collected: the hospital is public and the services are provided at a minimal charge, therefore, the hospital is popular with less affluent groups of the population. The women were 35 years or older and were referred for prenatal screening services due to their advanced maternal age which is associated with higher risks of having a child with Down’s syndrome. The screening services in Hong Kong are a routine part of prenatal care and women are offered a choice of: (a) indirect or non-invasive tests (a blood test or a sonographic measurement of a fetus’s neck); (b) direct or invasive tests (amniocentesis or CVS); and (c) a no-test option. The first option carries no risk to the mother or the fetus, but offers a lower detection rate and a higher false positive rate than the direct tests that carry a slight risk (1 percent) of miscarriage. As we have discussed in our previous work (e.g. Zayts and Schnurr 2011; Zayts and Pilnick 2014), among the Filipino clients it is not unusual to opt out of testing altogether or to agree to a non-invasive and safe testing option due to their Roman Catholic background and the religiously motivated opposition towards termination of pregnancy (which is the only “medical” intervention available to these women if an abnormality is identified). The widely accepted professional principle among medical professionals involved in the provision of prenatal screening services in Hong Kong and other countries is that the final choice of what testing (if any) to pursue lies with prospective parents. This principle confines the role of medical professionals to providing good quality information about available testing and facilitating clients’ decision-making. At the hospital prior to seeing a medical professional (typically a nurse; sometimes a doctor) women watch a video produced by the Hong Kong Department of Health that covers the nature of the syndrome, available tests, and

124 

 Olga Zayts and Stephanie Schnurr

support provided to families with Down’s syndrome children. This is followed by a face-to-face consultation. The consultations, although different depending on the family and medical history of clients, typically go through the following stages: (a) consultation opening; (b) history taking; (c) educational stage in which the nurses provide information about Down’s syndrome, available testing options, and their pros and cons; (d) decision-making stage in which clients make a decision about what testing option (if any) to pursue; and (e) consultation closing. These stages may be dispersed throughout a consultation (e.g. when participants “revisit” a decision they have made at different stages). In our previous study of nurses’ laughter (Zayts and Schnurr 2011) we have observed that most laughter sequences occurred in the education and decision-making stages, and we have attributed this to the fact that this is where most sensitive and potentially ­face-threatening talk occurs and where the clients’ choice regarding testing is negotiated. In our analysis we employ conversation analysis to illustrate the patterns in which nurses’ laughter is responded to by their clients. Conversation analysis is concerned with the “interactional accomplishment of particular social activities” (Drew and Heritage 1992: 17). These activities are comprised of sequentially ­organized social actions, which are described in relation to their context, social organization and any alternative means by which these actions (and activities) can be realized. We have applied conversation analysis to analyzing PS encounters in our previous work (e.g. Zayts and Schnurr 2011; Zayts and Pilnick 2014; Pilnick and Zayts 2014) and have noted that the advantage of this methodology is on the close attention to the processes involved in PS, for example, examining not only whether autonomous choice by the clients has been achieved but also how it has been achieved. Our analysis of the interactional data is also ethnographically informed. In particular, in our analysis we selectively draw on the information about participants’ backgrounds that was obtained in the interviews with the participants prior to and after the consultations. Such an approach of combining the analysis of interactional data with ethnographic observations yields additional insights into the analyzed interactions (for the advantages of this method see Maynard 2003; Silverman 1999; Pomerantz 2005; Waring 2012). The approach also proves particularly useful in analyzing such complex phenomenon as laughter in all its instantiations so as to shed light on its various conversational functions, including both ‘playful’ and ‘serious’ ones. In our previous work on nurse’s laughter (Zayts and Schnurr 2011) we ­followed Haakana (2002) in identifying laughter sequences for analysis. In particular, we examined every occurrence of nurses’ laughter in the data that was not preceded by laughter in the previous turn regardless of the position of ­laughter within the turn (i.e. regardless of whether it occurred in the turn-final



Laughter as a “serious business” 

 125

position, within the turn, or as a responsive laugh). In examining the patterns of the r­ eciprocation of nurses’ laughter by their clients in this paper, we revisit these data, albeit with a focus on the response that the nurses’ laughter receives. As responses we consider laughter as well as smiling/smiling voice as the phenomena closely linked to laughter (which can be considered as “milder” response to laughter). Table 1 below represents the distribution of laughter and its ­responses in our data corpus. Tab. 1: Distribution of laughter and its responses in PS encounters (Zayts and Schnurr 2011) Occurrences of laughter and its responses Women Laughter in initiated turns Smiling/smiling voice in ­reciprocated turns No reciprocation (i.e. no laughter/smiling) Responses

54 19 16 N = 89

61% 21% 18% 100%

Nurses 72 13 27 N = 112

64% 12% 24% 100%

As Tab. 1 demonstrates there are more cases of nurses’ (volunteered) laughter (72 cases) in the data corpus than women’s laughter (54 cases). These numbers point to an interesting trend in the analyzed data where the nurses volunteer laughter more often than the women. There are, however, fewer occurrences of the reciprocation of laughter by the nurses (13 cases) than by the women (19  cases); and the total number of the occurrences of the laughter is higher in the case of the women (82% as compared to 76% for the nurses). Previous studies of laughter in medical contexts (e.g. Haakana 2001) have shown that typically it is patients who laugh more and in most cases their laughter is not reciprocated by doctors. As Haakana notes (2001: 196), particularly when dealing with delicate issues, “by not laughing the doctors seem to be doing the right thing”. In the PS context, the higher occurrence of the laughter initiated by nurses in comparison to the ­client-initiated laughter may be explained by the very different nature of these encounters in comparison to other medical contexts. As the main goal of these PS consultations is to provide information to clients about available testing and facilitating clients’ decision-making, naturally the nurses get to do more talking in these consultations than the clients, and they also initiate laughter more frequently. In contrast, in primary care contexts where other studies of laughter were conducted (e.g. West 1984; Haakana 2002) patients present their acute medical problems to doctors and therefore talk and laugh more. In terms of the total number of the occurrences of laughter,

126 

 Olga Zayts and Stephanie Schnurr

our study supports the idea that in general clients laugh more than the medical providers. In what follows we will discuss what functions these instances of reciprocated laughter may perform. The analysis presented in the next sections is structured along three main functions of reciprocated laughter in the analyzed corpus: managing risk talk (4.1) and negotiating participants’ epistemic (4.2) and deontic (4.3) status and stance. Except for Example 1 that comes from the history-taking stage, all other examples come from the educational and the decision-making stages where, as we have previously observed (Zayts and Schnurr 2011), most of these laughter sequences occur.

4 Analysis 4.1 Shared laughter as a means of managing risk talk In PS consultations most talk relates to various risks. For example, in the history-taking stage participants discuss a woman’s advanced maternal age and the associated higher risks of Down’s syndrome, as well family history and lifestyle practices that may potentially increase a woman’s risk of having a child with abnormalities. The educational stage, on the other hand, involves the explanation of several types of risks: the “risk of occurrence” (the probability of abnormalities happening in a fetus; and the risks associated with testing [e.g. miscarriage]), the “risk of knowing” (psychosocial and interpersonal implications that finding out about abnormalities via testing may have on a client and the family) (Sarangi et al. 2003), and the “risk of not knowing” (psychosocial implications of not pursuing any testing [e.g. increased anxiety]) (see also Pilnick and Zayts 2014; Yau and Zayts 2014). Naturally, this relatively high amount of risk talk may easily lead to a heightened anxiety among the clients that may be both explicitly lexicalized by them (e.g. clients admitting that they are worried) and conveyed through various non-verbal means (e.g. crying). To address their clients’ anxiety the nurses routinely employ laughter along with other verbal and non-verbal strategies to reassure their clients and to put them at ease. In Example 1 the woman’s anxiety is related to taking a drug in the first trimester of her pregnancy to induce menstruation. As it turns out, the drug may lead to abnormalities of the heart and the limbs in a fetus. Together with other complicating factors in the woman’s family and personal health history, this puts her at a higher risk of having a child



Laughter as a “serious business” 

 127

with abnormalities. In our analysis below we look at what is achieved in the interaction through the nurse’s laughter and particularly its reciprocation by the woman. Example 1 The woman will be 42 years old at her expected time of delivery. She has a history of terminating a pregnancy in the past and a family history of mental retardation. 1 N: How many tablets did you take? 2 W: (° °) I think it’s four ((looks worried)) 3 N: Four tablets. (1.9) ((writing)) >All together,< (.) at one time, or:= 4 W: =No. 5 (1.5) 6 N: °Three weeks. Fourteen weeks.° ((writing)) (2.0) Ok. Nah, um, (2.1), 7 according to the literature review, that, (.) because we come across this drug (.) quite often, ok? Um, we know that this drug may-sometimes, may 8 9 cause um, abnormality with the heart, (.) and the limbs, alright? So, we can 10 have ultrasound to look for any-any abnormality with the heart and the limbs 11 W: On the baby? 12 N: Yes. ((nods)) 13 (4.2) 14 N: → Seems that you are very worried? ((N smiles and pats W on the arm)) 15 hh huh [huh huh ] 16 W: → [Hh hmm] Y(h)es:: ((smiles)) 17 N: Uh, (1.5) we have to check ultrasound first. ((N pats W on the shoulder 18 and keeps her hand there)) Alright? [And then-] 19 W: [Now?] 20 N: Ehm, maybe next week. Now we have the blood test. 21 W: Mm hmm. 22 N: Uh ((N removes her hand)) it’s quite common. °Ok?° It’s, we come oftenwe often come → across this drug. °Ok?° You don’t-you don’t have to 23 be too worried. [We w(h)ill → ch(hh)eck ultra(h)sound first] 24 25 W: → [((smiles)) ]. Ok. ((N makes arrangements for the ultrasound examination)) This extract starts with the nurse clarifying how many pills the woman has taken and informing her about the serious side effects of the drug as reported in the specialist literature. This is followed by a statement that the hospital can perform an ultrasound to check for abnormalities. The mother’s non-verbal behavior (a worried look) and a very quiet (inaudible to the researchers) response

128 

 Olga Zayts and Stephanie Schnurr

to the nurse convey her anxiety. From line 14 onwards the nurse begins to actively reassure the mother by employing a number of verbal and nonverbal strategies. She acknowledges the mother’s state (seems that you are very worried, line 14) and explicitly reassures her (you don’t- you don’t have to be too worried, line 23–24). She even pats the mother on the arm and the shoulder (and keeps her hand on the shoulder while she talks). While touching may largely be considered inappropriate outside the examination stages of medical encounters, we have observed in our data that the nurses often pat and touch the clients in an attempt to reassure them. There are a number of studies that have indicated that it is typically easier for the nursing staff to establish rapport and a more trusting relationship with their clients (e.g. Horrocks, Anderson, and Salisbury 2002). More close contact with the clients than expected may also be explained by the less asymmetrical relationship between the nurses and the women in the analyzed consultations. The nurse also talks about the drug as being quite common and something that the medical professionals often come across (in an attempt to reassure the mother that she knows what she talks about and how to deal with it). Of particular interest to us are two occurrences of laughter and their reciprocation. The first occurrence of the nurse’s laughter follows her statement that the mother seems to be worried. It is an extended laughter (that starts with audible aspiration and extends to three full beats of laughter). The second instance of the nurse’s laughter occurs immediately after her explicit reassurance that the mother does not need to worry (lines 23–24). This sequential location of the nurse’s laughter signals that laughter is indeed used to mitigate the risk talk and to reassure the woman. In both cases the woman’s responses are more “reserved”. In the first instance, the woman produces an audible aspiration followed by a confirmation of the nurse’s observation that she is worried which incorporates a laughter particle. The woman also smiles (line 16). In the second instance the woman just smiles (line 25). Previous studies have shown that more “reserved” responses to laughter, such as smiling, may be due to a delicate nature of talk (Jefferson 1985; Haakana 2010), as is indeed the case in this example. This “reserved” response may also signal the mother’s orientation to the institutional asymmetry between the nurse’s and her own role. The mother’s response, nevertheless, signals her affiliation with the nurse. The laughter and its reciprocation thus serve to mitigate the risk talk in this exchange and establish alignment between the participants with regards to what can be done about managing the woman’s risks. Since ensuring the client has enough information to make an informed choice is the main agenda of PS consultations, in addition to managing the risk talk and reassuring the women, another important aspect that the nurses’ job



Laughter as a “serious business” 

 129

involves is ensuring that the women fully understand the information about Down’s syndrome and available testing options. As the nurses are dealing with a very diverse client population in terms of their birth history (first-time and experienced mothers), sociocultural backgrounds (speakers with higher or lower language proficiency; native and non-native speakers of English and Chinese) and socioeconomic statuses (high income, college educated and above, and low income, low education level), the information delivery inevitably involves negotiation of the participants’ existing and new knowledge. It also involves the negotiation of what knowledge participants need to have in order to make an informed decision. Information delivery thus becomes a jointly accomplished activity where the nurses, on the one hand, “tailor” the information to each individual woman depending on her previous knowledge (Zayts and Kang 2010) and the women, on the other hand, co-construct with the nurses what they know (and need to know) in order to make an informed decision. Since ‘not knowing’ (or “non-understanding”) statuses may be potentially face-threatening, participants routinely employ laughter as a strategy of negotiating their knowledge.

4.2 Shared laughter as a means of negotiating participants’ epistemic statuses and stances In the analysis that follows we draw on some recent conversation analytical work on epistemics that discusses how participants negotiate knowledge in an interaction (Heritage 2012). This work makes a distinction between epistemic status (one’s knowledgeability about a certain domain of knowledge) and epistemic stance (speakers’ position towards each other’s epistemic statuses in an interaction), and epistemic authority (whether participants have [primary] rights to certain knowledge). Heritage (2012: 1) observes that “epistemic statuses of the speaker and hearer are a fundamental and unavoidable element in the construction of social action”. In relation to the information delivery, the epistemic statuses of the speaker and hearer are crucial as they determine what information is delivered to a client and in what ways this is done. Example 2 At the time of delivery the woman will be 40 years old. She has previous experience of prenatal screening as she had 4 pregnancies and delivered 2 children in the past. 1 N: → First of all, do you still remember what is Down’s syndrome? ((smiles)) 2 W: → Hehe. 3 N: Because you were seen here two years ago?

130 

 Olga Zayts and Stephanie Schnurr

4 W: Ye::s. We just watched the video. 5 N: → Yeah heh 6 W: → Hehe[he] 7 N: [No]w. This is a condition which is related to the chromosomal abnormality, ok? The kids having Down’s syndrome have mental retardation. Mm. We know 8 9 that there is no relationship with family history. Even if there is no family 10 history of Down’s syndrome, as your age is a little bit high, then your risk of 11 having a Down’s syndrome baby is higher as well, OK? Like you, at delivery you will be forty years, and the risk, you know? [You know what the risk is?] 12 13 W: [One in four- (.) eight hundred]. One in eight hundred. 14 N: One in one hundred. ((N continues with the information delivery)). In our corpus to tap into their clients’ knowledge or to establish their epistemic statuses, the nurses often employ “initial inquiries” (Zayts and Kang 2010), a type of question-response sequence that precedes the information delivery (e.g. Do you know/remember…?). In Example 2 the woman is an experienced mother who has undergone prenatal screening in her previous pregnancies, so a certain level of knowledge about Down’s syndrome can be safely presumed. The nurse, however, in line with her institutionally assigned role as information provider, needs to deliver that information to the mother. While in the interactions with less experienced women (first-time mothers) there is a divide between the epistemic statuses of the speaker (the nurse, a knowledgeable medical professional) and the hearer (a woman; non-knowledgeable lay person) that warrants information delivery, with more experienced women the speakers and hearers’ epistemic statuses converge to an extent, and the delivery of information that is already known becomes a repetitive and somewhat redundant activity. When the nurses still carry on with the information-giving, they typically mitigate this activity. In this example this mitigation is done nonverbally through smiling (line 1) as well as verbally through the account that the nurse gives for her action (because you were seen here two years ago [therefore, you could have forgotten this information], line  3). The nurse’s smile is reciprocated by the woman with two full beats of laughter (line 2). In line 4 the woman confirms that she remembers what Down’s syndrome is and notes that she also has just watched the video (together with other women seen in the hospital for prenatal screening). This latter acknowledgement makes the nurse’s information delivery even less relevant. It, therefore, leads to another short instance of laughter from the nurse (one laughter beat, line 5) that is again reciprocated by the woman (an extended laughter sequence comprising of three



Laughter as a “serious business” 

 131

laughter beats, line 6). By reciprocating the nurse’s laughter, the woman, on the one hand, signals her understanding of the peculiar situation that the nurse finds herself in: questioning the woman’s memory and still having to provide the information that the woman appears to remember well. On the other hand, the reciprocation of laughter does something else: it terminates the topic of “inquiring about the woman’s epistemic status” and allows the participants to move on to the next topic (Holt 2010). Interesting about this example is that the woman’s laughter on both occasions is more “pronounced” than the nurse’s laughter (the nurse’s smile is responded to with two full beats of laughter [lines 1–2]; and one beat of laughter is responded to with three full beats of laughter [lines 5–6]). As we can see in this interaction, lay participants may also take on a knowledgeable epistemic stance, thus questioning the epistemic authority of professionals. In this example, while the topic is terminated by the woman through laughter, the nurse is the one who terminates the woman’s laughter and initiates a new topic (that is marked by the discourse marker now, line 7) thereby regaining her institutional (and interactional) authority. This is typical of our data corpus: while both nurses and women engage in laughter, it is usually the nurses who bring laughter to an end and resume the “business at hand”. Heritage (2012) observes that participants’ epistemic statuses are not static: they may change as an interaction progresses. In our corpus, laughter is also employed by the participants during the actual information delivery to adapt to their changing epistemic statuses from ‘not knowing’ to ‘knowing’ after they receive the information about PS from the nurses. Example 3 comes from the end of the information delivery stage. The nurse has just explained in detail all testing options to the woman. At this point in the interaction it may be assumed that the woman understands what her choices are and is able to make a decision. Example 3 The woman in this consultation is 42 years old. This is her first pregnancy. Prior to this extract, the nurse has explained the woman’s increased risk of having a child with Down’s syndrome due to her age. The nurse has also introduced CVS and amniocentesis as possible testing options. 1 N: So, so far are these choices clear? ((smiles)) ((points to the information 2 sheet in front of her)) 3 W: Yes. ((smiles)) 4 N: So (.) uh (.) it’s up to you hh whether you- which test you want (2.0) 5 6 N: .h or do you need to discuss with your husband?

132 

 Olga Zayts and Stephanie Schnurr

At the beginning of this extract the nurse inquires whether the choices that she has presented are clear to the woman (line 1). Since admitting to non-understanding of something that has just been explained is potentially a face-threatening issue, the nurse mitigates her question with a smile. The woman confirms that the choices are clear to her, thereby also establishing and confirming her epistemic status of someone who knows (line 3). The confirmation is accompanied by the reciprocation of the nurse’s smile (line 3) that brings this exchange to an end, and the nurse moves on to the next topic of making a choice. The new topic is marked by the discourse marker so (line 4). Thus in this example, similarly to Example 2, the shared smiling among the participants displays their orientation towards bringing one topic to an end due to the now ‘knowing status’ of the woman and shifting to another topic. Recurrently, we see a different pattern of laughter in our data when there is a clear “epistemic imbalance” between the participants as Example 4 ­demonstrates: Example 4 The woman in this interaction is 38 years old. This is her second pregnancy but she has no previous experience of Down’s syndrome screening (her first child was born in the Philippines). This example comes after the nurse has taken the woman’s medical and family history. 1 N: Ehm: have you watched video (.) today? 2 W: Mm hmm. 3 N: Ah:: (1.5) ((takes notes)) Do you have any idea of Down Syndrome? 4 (1.3) 5 W: Mm: °no.° 6 N: → N(hh)o:. Have you seen children like this before? 7 P: Mm hmm. In Example 4 the nurse confirms with the woman that she has watched the video on Down’s syndrome (line 1) and employs an initial inquiry (Do you have any idea of Down Syndrome?, line 3) to tap into how much the woman understands about Down’s syndrome. While the woman did watch the video, she acknowledges, after a pause and a slight hesitation that she does not know about Down’s syndrome. This leads to the nurse’s reserved laughter: she reiterates the woman’s response that incorporates a laughter particle (line 6). The nurse’s laughter here signals the dispreferred nature of the woman’s response. In other words, there is an expectation that after watching the video, women have some idea about Down’s syndrome that will facilitate the subsequent information delivery. The nurse then attempts another initial inquiry (Have you seen children like this before?, line 6) to elicit whether the woman indeed has no previous knowledge or



Laughter as a “serious business” 

 133

experience of seeing children with Down’s syndrome. What is interesting about this example is that the nurse’s laughter is not reciprocated by the woman. This signals that the woman does not treat the nurse’s laughter as an invitation to laugh. The lack of reciprocation of laughter suggests that the participants are orienting to two different issues in this interaction. While the nurse’s laughter is oriented towards the fact that the woman who has just watched a video about Down’s syndrome says that she does not know what it is, the woman’s laughter signals her orientation to the topic of Down’s syndrome as a serious matter. Sequentially, the absence of the reciprocation means that the topic of “how much the woman knows about Down’s syndrome” is not terminated at that point and the nurse continues to inquire about the woman’s state of knowledge. Examples 2 to 4 demonstrate how through laughter and its reciprocation (or non-reciprocation) the participants negotiate their epistemic statuses and stances prior to and during the information delivery to facilitate this process. Alongside the epistemic dimension of communication, participants also ­negotiate their deontic authority that relates to the “right to determine others’ future actions” (Stevanovic and Peräkylä 2012: 297). As Stevanovic and Peräkylä (2012: 298) note, epistemic and deontic authority are interrelated concepts: “­epistemic authority is about knowing how the world ‘is’; deontic authority is about determining how the world ‘ought to be’”. In our previous work (Pilnick and Zayts 2016) we have discussed that in PS consultations negotiation of deontic authority relates to who gets to make a decision about what tests (if any) a pregnant woman should take. We have shown that on some occasions medical professionals exercise their deontic authority overtly and tell the women what test they should take. More typically though, decisions about testing involve a careful negotiation of the participants’ deontic rights. Even when the professionals choose to direct their clients in the decision-making towards the options that they (the professionals) favor, this is done in more covert ways by, for example, giving recommendations, listing (preferred) testing options, or foregrounding specific psychosocial concerns (e.g. these women’s worries) as the basis for the decision to undertake a specific test (see also Pilnick and Zayts 2014). In our last two examples we examine how participants employ laughter as another “covert” strategy of negotiating their deontic authority. 4.3 Clients’ laughter as a means of negotiating deontic authority Decision-making in PS consultations involves decisions about what information a woman needs to know before making an informed decision about testing, and the actual decision-making about testing. In Zayts and Schnurr (2011) we have

134 

 Olga Zayts and Stephanie Schnurr

observed that nurses routinely employ laughter in responding to the women’s dispreferred activities as a means of resisting those activities. As Glenn (2003: 141) notes, “structurally resistance means acting to discontinue the activity proposed or in progress”. In Example 5 the nurse laughs in response to the woman’s refusal to receive more information about one of the tests, amniocentesis, as she has decided not to pursue invasive testing. The nurse, however, is required to give full information about all available tests as part of her job (and as a necessary step in helping the woman make an informed decision). We focus on what is achieved through the reciprocation of the nurse’s laughter by the woman. Example 5 The woman is 38 years old and this is her first pregnancy. Prior to the interaction included in the extract the nurse explained to the woman what Down’s syndrome is, what the woman’s risk is of bearing a child with the condition, and the pros and cons of different testing options. Following the information delivery, the woman has made the decision not to pursue any testing and to continue the pregnancy. 1 N: Would you like to have more information on amniocentesis? 2 W: ((shakes her head throughout the utterance)) 3 No (it’s not-it’s not ne[cessary)] 4 N: → [You d(h)on’t] wa(h)nt [hah >hah hah hah< hah] 5 W: →    [heh heh huh huh huh] 6  → I’m s(h)o(h)7 N: → Yo(h)u don’t w(h)ant. 8 W: → I don’t wa(h)nt t(h)o. 9 N: Ok, now. Uh, in fact, this screening test can- uh 10 estimate your risk again. If it is high:, then, we will 11 advise you to go for amniocentesis. 12 W: Mm hmm. 13 N: If the risk is low? Then we will not arrange further testing. Detection rate is eighty-six percent. So you 14 would prefer not to know of any amniocentesis= 15 16 W: =No. By announcing her decision to the nurse (not included in the extract) the woman has established her deontic authority. The nurse then inquires whether the woman would still like to receive information about amniocentesis (that she has not yet delivered) (line 1). On the surface, the nurse’s question does not appear to threaten the woman’s deontic authority (that is, the right to make a decision about the continuation of pregnancy), the implication of the nurse’s question, however, goes beyond a simple inquiry about providing more information. As



Laughter as a “serious business” 

 135

Stevanovic and Peräkylä (2012: 309) note, providing more information may be used as an indirect strategy “to resist the unfavorable deontic implications of the first speaker’s utterance”. A woman’s overt refusal to receive the information (that is strengthened by shaking her head as she speaks) (lines 2–3) is interrupted by the nurse’s confirmation of the decision that contains laughter particles (You d(h)on’t] wa(h)nt) and five full beats of laughter (line 4). The nurse’s laughter signals her disaffiliation with the woman’s action of refusing information. In our previous work we discussed that in such cases the nurses use laughter to “laugh off” their clients’ refusals (Zayts and Schnurr 2011). On the one hand, the nurse’s laughter signals her resistance; on the other hand, it also mitigates this process of negotiating participants’ deontic rights, making it less confrontational. The woman reciprocates the nurse’s laughter (five beats of laughter, line 5) and then attempts to provide what looks like an account for her decision (I’m s(h)o(h)-, line 6). As Lyman and Scott observed (1967: 46), speakers typically use accounts to justify or excuse their “untoward” actions and to “bridge the gap between actions and expectations” (Lyman and Scott 1968: 46). The reciprocation of laughter and the presence of the incomplete account in the woman’s response signal the woman’s awareness of the dispreferred nature of her refusal and serve to mitigate it. This is followed by another repetition of the woman’s decision not to receive the information about amniocentesis by the nurse (line 7) who inserts laughter particles into her speech. The nurse’s laughter is again reciprocated by the woman who confirms her decision (line 8). This example thus presents an extended laughter sequence in which the nurse’s laughter is repeatedly reciprocated by the woman. The nurse’s extension of laughter can be seen as a continued attempt to express her resistance towards the woman’s decision and perhaps an invitation to reconsider it. The reciprocation of the nurse’s laughter by the woman, on the one hand, mitigates the dispreferred action of refusing the information. On the other hand, it may also be seen as an attempt to bring this topic to an end, thereby signaling that this is her final decision. While it is the woman who terminates the laughter sequence, it is the nurse who brings the consultation “back to business”: she resumes the talk about different tests, bringing up amniocentesis yet one more time as the procedure that the woman will be advised to consider if her screening test results are positive (i.e. the woman’s risk of having a baby with Down’s syndrome is identified as high). The nurse’s last attempt to explain the procedure (lines 14–15) is again refused by the woman (line 16).

136 

 Olga Zayts and Stephanie Schnurr

In Stevanovic and Peräkylä’s (2012) terms, it can be described as an example of “deontic incongruence” where the nurse’s and the woman’s agendas are ­misaligned and each other’s actions are treated as dispreferred. Through laughter and its reciprocation the participants engage in a complex process of negotiating their deontic rights. The laughter in this example allows both participants to make claims to their deontic rights, and to minimize the effect of their dispreferred actions. In the last example we examine how the participants employ laughter to negotiate their deontic authority in the actual decision-making process. Example 6 (This example is also analyzed in Zayts and Schnurr 2011: 12) The woman in this encounter is 38 years old and this is her fourth pregnancy. Prior to this extract, the woman has expressed her firm decision not to undergo any testing due to religious reasons. Nonetheless, she has accepted the nurse’s invitation to provide information on available tests. The extract comes at the end of the information delivery when the woman announces her final decision. 1 N: → Or:, you can adopt the third choice, that is no test ((smiles through the 2 utterance)). 3 W: Yes. 4 N: → Yes. That’s what you want? ((smiles)) 5 (0.6) 6 W: That’s what we want. 7 N: → O(h)k heh heh ((14 lines are omitted in which the nurse makes arrangements with the woman to have an ultrasound at eighteen weeks of gestation)) 22 N: So you will not change your mind? Ah huh huh [huh huh] 23 W: [°N(h)o:°] 24 N: Anyway, you will continue with the pregnancy? 25 W: Yes. 26 N: Yeah, °I see.° So, you will not em:: ultrasound only. 27 W: Ultrasound please. The example starts with the nurse telling the woman about one possible choice – not pursuing any testing (lines 1–2). The nurse smiles as she talks about this option. As we have already noted, in the interviews the nurses commented on their strong preference of testing to no testing; therefore in this example, smiling can be considered a “milder version” of expressing that this is the nurse’s dispreferred



Laughter as a “serious business” 

 137

option. The nurse also smiles when she confirms the woman’s brief response of not wanting to pursue testing (line 4); and produces an extended laughter (line 7) to the woman’s stronger assertion about her decision (line 6). Example 6 is thus another example of “deontic incongruence” where the nurse resists the woman’s deontic right to make the decision of not pursuing any testing. The absence of the reciprocation of laughter by the woman in the first part of this exchange signals that the woman treats this exchange as “non-laughable” or serious. The nurse’s question in line 22 (So you will not change your mind?) is in a declarative form which is typically used to confirm a response (in this case, the woman’s decision) rather than to question it. Nevertheless, the very fact that the question is being asked, raises the issue of the woman’s deontic right. The nurse’s extended laughter (5 full beats) that follows the question mitigates the face-threatening nature of her action. The woman’s brief (and quiet) response that confirms her decision incorporates a laughter particle (°N(h)o:°, line 23). The low volume and the laughter mitigate the refusal. Laughter in the woman’s response also serves as an attempt to terminate the nurse’s repeated actions of questioning her right to make the decision, and the discourse marker “anyway” in the nurse’s next turn signals the end of this topic and the move to the next one (Lenk 1998).

5 Discussion and conclusion In this chapter we have examined the second pair part of the laughter sequences negotiated by nurses in PS consultations. Building on existing research of shared laughter, particularly in institutional contexts, we have shown that there are certain regularities in how the pregnant women in our data reciprocate (or do not reciprocate) the nurses’ laughter. These r­ egularities appear to be closely related to the specific activities of the PS consultations that the participants engage in, such as communicating risk information and making a decision about potentially pursuing testing for Down’s syndrome. We have shown that shared laughter may be affiliative and serve to establish rapport between the participants. This happens in those cases when the nurses employ laughter to mitigate the risk talk that is prevalent in the analyzed interactions, and to reassure the women. In such cases, the women laugh with the nurses (Glenn 2003) to express their affiliation with the nurses’ actions. However, as we have shown, shared laughter may also express participants’ disaffiliation. In the analyzed interactions, these are those cases where participants engage in

138 

 Olga Zayts and Stephanie Schnurr

negotiating their epistemic and deontic statuses. As Jacknick in her analysis of ESL classroom interactions notes: Despite expectations based on institutional role, epistemic authority is not owned by one party or another; rather, it is claimed by participants with different roles [...] in different ways, and laughter is instrumental in displaying (dis)alignment with claims (2013: 199).

Applying this idea to the context of PS, there are certain roles that are institutionally assigned to the participants, such as information- provider and information- recipient (for a detailed discussed see Zayts and Schnurr 2014). Based on those roles, one can expect a certain “epistemic divide” between the nurses and the women. However, as we have shown in our analysis, in consultations with more “experienced” expectant mothers (for example, those who know about Down’s syndrome and screening and testing options from their experience in previous pregnancies), participants’ epistemic statuses may converge to an extent. In these cases, the activity of information delivery that the nurses are institutionally required to perform then becomes somewhat redundant. In such cases, shared laughter is used by the participants strategically to negotiate their epistemic statuses. The nurses laugh to mitigate the potential face-threat of the inquiries about the women’s epistemic status; whereas the women laugh to terminate the topic of “inquiring about their knowledge” and to establish their epistemic authority. Examples like these reflect the complexity of the process of negotiation of participants’ epistemic statuses in institutional interactions: while professionals have an a priori “privileged” epistemic status due to their professional knowledge and expertise, one cannot automatically assume that there is necessarily an “epistemic imbalance” (Heritage 2012: 32) between professional and lay participants. As a contrast, we have shown that when there is an epistemic imbalance between the participants, the women do not reciprocate the nurse’s laughter, and therefore, the activity of inquiring about their epistemic status may be continued by the nurses. Finally, shared laughter may also be employed by the participants to negotiate their deontic authority. The issue of deontic authority is particularly sensitive in the context of PS consultations where clients (women and their partners or significant others) are expected to make an autonomous informed choice of what testing (if any) to pursue. As we have shown elsewhere (e.g. Pilnick and Zayts fc), autonomous informed choice is difficult to attain interactionally: even when the medical professionals are not advising their clients directly what to do, they could still express their preferences in more indirect ways. In Example 6, for instance, the nurses’ resistance to the woman’s decision not to pursue any testing is expressed through repeated questioning of the woman’s choice, as well as non-verbal resources such as smiling and laughter. As we have shown in our



Laughter as a “serious business” 

 139

analysis of Example 5, the negotiation of deontic authority may involve negotiating the actual right to make the decision (Example 6), as well as the epistemic grounds (i.e. the knowledge) that would be sufficient to make an informed decision (Example 5). By focusing on the client’s responses to the nurses’ laughter this chapter has shed further light onto the interactional complexities of laughter in the medical context where, as we have shown, laughter performs a range of important functions which assist interlocutors in negotiating and ultimately achieving their interactional goals. In these contexts, laughter is not only used to mitigate and facilitate risk talk, as might have been expected, but, perhaps more interestingly, it is also frequently employed as a means to negotiate interlocutors’ epistemic statuses and stances, as well as their deontic authority. And we believe that it is in particular these latter functions relating to epistemic and deontic stances that provide promising avenues for future research in the endeavor to better understand the interactional complexities of laughter, and thereby contributing more broadly to our understanding of ‘language play’.

References Adelswärd, Viveka. 1989. Laughter and dialogue: The social significance of laughter in institutional discourse. Nordic Journal of Linguistics 12(2): 107–136. Atkinson, John Maxwell & John Heritage (eds.). 1984. Structures of social actions: Studies in conversation analysis. Cambridge: Cambridge University Press. Drew, Paul & John Heritage. 1992. Analyzing talk at work: An introduction. In Paul Drew & John Heritage (eds.), Talk at work: Interaction in institutional settings, 3–65. Cambridge: Cambridge University Press. Glenn, Philipp. 1991. Current speaker invitation of two-party shared laughter. Research on Language and Social Interaction 25(1–4): 139–162. Glenn, Philipp. 2003. Laughter in interaction. Cambridge, England: Cambridge University Press. Glenn, Philipp. 2010. Interviewer laughs: Shared laughter and asymmetries in employment interviews. Journal of Pragmatics 42(6): 1485–1498. Glenn, Philipp & Elizabeth Holt. 2013. On laughing; Studies of laughter in interaction. London: Bloomsbury Academic. Haakana, Markku. 1999. Laughing matters: A conversation analytical study of laughter in doctor-patient interactions. Helsinki: Department of Finnish Languages, University of Helsinki PhD dissertation. Haakana, Markku. 2001. Laughter as a patient’s resource: Dealing with delicate aspects of medical interaction. Text Interdisciplinary journal for the study of discourse 21(1–2): 187–219.

140 

 Olga Zayts and Stephanie Schnurr

Haakana, Markku. 2002. Laughter in medical interaction: From quantification to analysis, and back. Journal of Sociolinguistics 6(2): 207–35. Haakana, Markku. 2010. Laughter and smiling: Notes on co-occurrences. Journal of Pragmatics 42(6): 1499–1512. Have, Paul ten. 2007. Doing conversation analysis: A practical guide, 2nd edn. London: SAGE Publications. Hepburn, Alexa & Scott Varney. 2013. Beyond ((laughter)): Some notes on transcription. In Philipp Glenn & Elizabeth Holt (eds.), Studies of laughter in interaction 25–38. London: Bloomsbury. Heritage, John. 2012. Epistemics in action: Action formation and territories of knowledge. Research on Language and Social Interaction 45(1): 1–29. Holt, Elizabeth. 2013. Conversation analysis and laughter. In Carolle A. Chapelle (ed.), The encyclopedia of applied linguistics 1033–1038. Oxford, UK: Wiley-Blackwell. Horrocks, Sue, Elizabeth Anderson & Chris Salisbury. 2002. Systematic review of whether nurse practitioners working in primary care can provide equivalent care to doctors. British Medical Journal 324: 819–823. Jacknick, Christine. 2013. “Cause the textbook says …”: Laughter and student challenges in the ESL classroom. In Philipp Glenn & Elizabeth Holt (eds.), On laughing: Studies of laughter in interaction 185–200. London: Bloomsbury Academic. Jefferson, Gail. 1979. A technique for inviting laughter and its subsequent acceptance or declination. In George Psathas (ed.), Everyday language: Studies in ethnomethodology 79–96. New York: Irvington. Jefferson, Gail. 1984. On the organization of laughter in talk about troubles. In John Maxwell Atkinson & John Heritage (eds.), Structures of social actions: Studies in conversation analysis 346–369. Cambridge: Cambridge University Press. Jefferson, Gail. 1985. An exercise in the transcription and analysis of laughter. In Teo A. van Dijk (ed.), Handbook of discourse analysis (Discourse and Dialogue 3) 25–34. London: Academic Press. Lenk, Uta. 1998. Marking discourse coherence: Functions of discourse markers in spoken English. Tübingen: Gunter Narr Verlag. Maynard, Douglas W. 2003. Bad news, good news: Conversational order in everyday talk and clinical settings. Chicago, IL: University of Chicago Press. Pilnick, Alison & Olga Zayts. 2014. ‘It’s just a likelihood’: Uncertainty as topic and resource in conveying ‘positive’ results in an antenatal screening clinic. Symbolic interaction 37(2): 187–208. Pilnick, Alison & Olga Zayts. 2016. Advice, authority and autonomy in shared decision making in antenatal screening: The importance of context. Social Science and Medicine 38 (3): 343–359. Pomerantz, Anita. 2005. Using participants’ video stimulated comments to complement analyses of interactional practices. In Hedwig te Molder & Jonathan Potter (eds.), Talk and cognition: Discourse, mind and social interaction 93–113. Cambridge: Cambridge University Press. Sarangi, Srikant, Bennert Kristina, Howell Lucy & Angus Clarke. 2003. ‘Relatively speaking’: Relativisation of genetic risk in counselling for predictive testing. Health, risk and society 5(2): 155–170. Scott, Marvin B. & Stanford M. Lyman. 1968. Accounts. American Sociological Review 33(1): 46–62.



Laughter as a “serious business” 

 141

Silverman, David. 1999. Warriors or collaborators: Reworking methodological controversies in the study of institutional interaction. In Srikant Sarangi & Celia Roberts (eds.), Talk, work and institutional order: Discourse in medical, mediation and management settings 401–425. Berlin: Mouton de Gruyter. Stevanovic, Melisa & Anssi Peräkylä. 2012. Deontic authority in interaction: The right to announce, propose and decide. Research on Language and Social Interaction 45(3): 297–321. Waring, Hansun Zhang, Sarah Creider, Tara Tarpey & Rebecca Black. 2012. A search for specificity in understanding CA and context. Discourse Studies 14(4): 477–492. West, Candace. 1984. Routine complications: Trouble with talk between doctors and patients. Bloomington: Indiana University Press. Yau, Alice & Olga Zayts. 2014. “I don’t want to see my children suffering after birth”: The ‘risk of knowing’ talk and decision-making in prenatal screening for Down syndrome in Hong Kong. Health, Risk and Society 16(3): 259–276. Zayts, Olga & M. Agnes Kang. 2010. Information delivery in prenatal genetic counseling: On the role of initial Inquiries. Journal of Asian Pacific Communication 20(2): 243–259. Zayts, Olga & Stephanie Schnurr. 2011. Laughter as a medical provider’s resource: Negotiating informed choice in prenatal genetic counseling. Research on Language and Social Interaction 44(1): 1–20. Zayts, Olga, V. Y. Wake & Stephanie Schnurr. 2012. Chinese prenatal genetic counseling discourse in Hong Kong: Health care providers’ (non)directive stance, or who is making the decision. In Yuling Pan & Dániel Z. Kádár (eds.), Chinese discourse and interaction: Theory and practice 228–247. London: Equinox. Zayts, Olga & Alison Pilnick. 2014. Genetic counseling in multilingual and multicultural contexts. In Heidi Hamilton & Wen-Ying Sylvia Chou (eds.), Handbook on language and health communication 557–572. London and New York: Routledge. Zayts, Olga & Stephanie Schnurr. 2014. More than ‘information provider’ and ‘counselor’: Constructing and negotiating roles and identities of nurses in genetic counseling sessions. Journal of Sociolinguistics 18(3): 345–369.

Michael Haugh

6 Jocular language play, social action and (dis)affiliation in conversational interaction Abstract: Language play is often understood as talk that is aimed at mutual entertainment and enjoyment for participants, rather than the accomplishment of particular interactional business or forming and maintaining interpersonal relationships. While language play was initially studied with respect to creative use of linguistic forms (play with language), it was subsequently extended to also include both semantic and pragmatic forms of language play (play in language). In this paper, episodes of jocular language play are examined in interactions amongst Australian and American speakers of English. Jocular language play involves episodes of playful talk which are initiated when a participant somehow diminishes something of relevance to self, other, or a non-co-present third party within a playful frame, either through some kind of play with language or through some kind of made-up or exaggerated scenario or proposal. It is proposed that in addition to the mutual entertainment or enjoyment for participants that language play affords, it can also allow participants to negotiate or modulate stances that are implemented through sensitive or delicate social actions. It is argued that through jocular play, participants may affiliate, or mask disaffiliating, with the affective or moral stances of other participants in ways that are sensitive to the emerging trajectory of the social action in question.

1 Introduction It has long been recognized that the use of language is not restricted to the accomplishment of particular transactional or relational outcomes. In some instances language is used in more creative ways for the pleasure of those users. This form of language use comes under the broad umbrella of language play, namely, instances where “people manipulate the forms and functions of language as a source of fun for themselves and/or for the people they are with” (Crystal 1996: 328). Cook (1997) explains that language “‘play very often has something to do with enjoyment and relaxation; that its meanings and relationships are different from those in the society around it; and that the consequences of what happens in the play world are not directly relevant to the real world” (p. 227). This conceptualization of language play reflects the way in which play and playful are understood in ordinary vernacular speech (in English at least). According to the Oxford DOI 10.1515/9781501503993-007

144 

 Michael Haugh

English Dictionary Online (2015), for instance, playful, which is derived from Old English plega, can be understood as to be “light-hearted; fond of fun, games, and amusement; giving or expressing pleasure or amusement; frolicsome, sportive”, while according to the Merriam-Webster Dictionary Online (2015), it also encompasses, in addition, talk or conduct that is “humorous, jocular”. Language play from an English speaker’s perspective thus generally combines elements of fun and enjoyment with pretence.1 One of the key distinctions made with respect to language play is that between play with and play in language (Cook 1997, 2000). The former refers to “cases in which the language itself is manipulated for play” (Bell 2012: 191). The latter refers to “instances in which the language is used to engage in play” (Bell 2012: 191). Cook (1997, 2000) takes a fairly broad view of what counts as play in language, to the extent that it is sometimes difficult to distinguish between language play and conversational humor more generally (Coates 2007; Dynel 2011). In this chapter, however, language play is understood as overlapping with conversational humor only to the extent that it involves either some kind of wordplay/ play on language, or some kind of made-up or exaggerated scenario or situation, which is construed by participants as non-serious. The focus will be specifically on instances of jocular language play amongst American and Australian speakers of English, that is, instances of play with language or play in language which are construed as amusing by those participants. In the following excerpt from a conversation between two postgraduate students at a university in Sydney, for instance, a colorful comment by one of the participants on the appearance of the coffee they have just bought occasions an episode of playful mocking of that coffee. (1) ICE-AUS: S1A-020: 0:00 1 J: this coffee looks like ↑up chuck. 2 L: it does= 3 J: =I- i- it’s just (.) it’s see:thing. 4 L: he he he he 5 J: huh l- >look at it< it’s just (.) ugh. 6 (0.8) 7 L: it’s- it’s soymilk isn’t it. 8 J: ↓my soymilk coffee’s got the runs.

← ←

1 Language play can thus be understood as falling under the broader umbrella of “non-­ seriousness”, which encompasses a wide range of evaluative frames that are related to but distinct from playful and jocular, including laughable, cheeky, flippant, facetious, absurd, irreverent, mock, sportive and so on.



Jocular language play 

9 L: HA HA HA [ha ha ha 10 J: [he he he 11 J: he it’s got those litt(hh)le fle(hh)cks of 12 tomato it’s le(hh)fto(h)ver 13 L: ca(hh)rrot. 14 J: HHH ↑HA HA HA 15 (0.2) 16 L: he he he 17 J: hhh. (0.2) o:h. (giardia) of the coffee. 18 some (gia(hh)rdia) with you coffee. hhh. 19 L: taste’s al↑right though. 20 J: o:h ye:ah. I °s’pose°. mhm 21 L: °he he° (0.2) °he he°= 22 J: =it’s gro:ss 23 (0.2) 24 L: £just don’t loo(hh)k at it.£ 25 J: he he ↑he. 26 (0.2) 27 L: mmm tastes alright [though.] 28 J: [just ] (.) lay back and 29 think of 30 (0.4) 31 L: England (0.4) as I al(hh)ways [do. 32 J: [(l- hhh 33 L: he he (Haugh 2014: 92–93)

 145

← ←

The episode of jocular language play here is occasioned by an instance of objectdirected jocular mockery (Haugh 2010, 2014) in line 1 by Jason, namely, the exaggerated proposal that the coffee looks like “upchuck”. Following agreement with this stance by Lisa, the descriptions of the coffee become increasingly more exaggerated and absurd in depicting the coffee as “seething” (line 3), as having “the runs” (line 8), followed by a co-constructed description of it that likens it to the contents of vomit (lines 11–13), and the absurd claim that the coffee has “giardia” (lines 17–18). The episode is concluded through another co-constructed formulaic utterance (lines 28–31), where Lisa affiliates with Jason’s stance that drinking the coffee is something that just has to be endured, as well as indicating the closing of the sequence in question (Drew and Holt 1988). This particular sequence thus has many of the features of language play noted by Coates (2007: 38), including the co-construction of utterances (e.g. lines 11–13,

146 

 Michael Haugh

28–31) and metaphor (e.g. line 8), to which exaggerated and absurd descriptions (e.g. lines 1, 3, 11–13, 17–18) might be added (Norrick 2004; Ritchie 2011; Ritchie and Schell 2009). The episode of language play is also interactionally achieved as jocular through various phonetic features of the proposed descriptions of the coffee that mark them as laughables (Jefferson 2004b; Holt 2010; Straehle 1993), as well as the laughing responses and shared laughter on the part of both participants (e.g. lines 4, 9–10, 14–15, 21, 25, 33) (Coates 2007; Glenn 2003; Jefferson, Sacks and Schegloff 1987). Yet while the instances of object-directed mockery that collectively constitute this episode of language play are evidently designed and responded to as jocular, Lisa nevertheless takes a somewhat different stance to Jason on the coffee itself, as Haugh (2014: 93–94) notes. Specifically, while Jason takes the stance that the coffee is “just ugh” (line 5) and “gross” (line 22), Lisa takes that the stance that although the coffee looks unappealing, it nevertheless “tastes alright” (lines 19, 27). Jason responds somewhat equivocally to Lisa’s stance in hedging his agreement with it (line 20), and subsequently disattending it (lines 28–29). The way in which the two participants interactionally achieve diverging stances on the coffee, constitutes evidence that “Lisa may be treating the negative assessments of the coffee as implying a negative assessment of her choice of coffee, and thus potentially a negative assessment of her” (Haugh 2014: 94). In other words, implied criticism of Lisa’s choice of coffee by Jason, on the one hand, and disaffiliation with this criticism by Lisa, on the other hand, is accomplished here within the context of a playful or non-serious frame. It is apparent, then, that ostensibly non-serious instances of jocular language play may nevertheless allow participants to accomplish serious interactional business, in this case, the two participants taking somewhat different affective stances about coffee that is made with soymilk. In this chapter, the ways in which participants may accomplish various kinds of sensitive social actions, such as accusations, criticisms, complaints and disagreements, through jocular language play are examined. The focus, in particular, is on the ways in which participants may affiliate or disaffiliate with the stances or action preference implemented by those social actions (Stivers, Mondada and Steensig 2011). The analysis thus builds on earlier work attesting to the claim that language play is not simply a matter of “fun” (Crystal 1996: 328), “relaxation and enjoyment” (Cook 1997: 227) or “the creation of solidarity” (Coates 2007: 32), but rather may also involve “constructing and reaffirming (group and individual) identities, negotiating shifting alignments, mitigating face-threats, and expressing changes in stance and footing” (Bell 2012: 193). The chapter begins, in the following section, by briefly reviewing prior work on language play and non-seriousness to lay the groundwork for an analysis of



Jocular language play 

 147

how delicate or sensitive social actions can be accomplished through jocular language play. The dataset and analytical approach utilized in this paper are then outlined in section 3, before going on to describe how participants may (dis)affiliate through two related forms of jocular language play, namely, jocular play with language (section 4) and jocular play in language (section 5). The implications of this analysis for studies of language play more broadly are then briefly considered in the conclusion.

2 Language play and non-seriousness One of the first academic studies of playful interaction was undertaken by Bateson (1955) who noted that actions can be accompanied by different metamessages, including claims to playfulness. In general, play involves some form of pretence where the “actions, in which we now engage, do not denote what would be denoted by those actions which these actions denote” (Bateson 1955: 41). This was based on the observation that while animals may engage in playful nips, such nips “do not denote what would be denoted by the bite for which it stands”, and so constitutes a “fictional bite” (Bateson 1955: 43). This was then figuratively extended to human verbal interaction. It follows, then, that language play can work at two levels in that not only is the speaker not necessarily committed to the truth of what is said, but what is said itself may be a fabrication. This playful meta-message is generally referred to in subsequent work as either the “key” of an interaction (Hymes 1972) or an interpretive “frame”.2 This initial work by Bateson on play has inspired a wide range of researchers across different disciplines. One aspect of verbal play that has received considerable attention is the way in which participants come to understand the talk in question is being couched as non-serious. While Goffman (1974) initially argued that play is keyed by the speaker at the beginning of an episode, subsequent work has indicated that play is interactive in nature. Indeed, the consensus to date appears to be that language play is something that is collaboratively accomplished by participants (Bell 2012; Coates 2007; Glenn and Knapp 1987; Gordon 2008; Hopper 1995; Ritchie and Schell 2009). Kochman (1983), for instance, argues that recipients are just as important as speakers for maintaining a play frame, while Hopper (1995) shows that playfulness is very often keyed throughout episodes

2 The “key” of action refers to the “tone, manner or spirit in which an act is done” (Hymes 1972: 62). The “frame” of an action refers to the “definitions of a situation” in which it occurs (Goffman 1974: 11).

148 

 Michael Haugh

of language play, and that instances where these cues are found primarily at the very start of such episodes are relatively rare. It is for this reason that establishing a play frame is likened to a collaborative musical activity such as playing jazz in an ensemble (Coates 2007). A related analytical theme is that this keying or framing of an action as serious or non-serious is important for participants in interpreting what sort of next response it makes contingently relevant (Sacks 1972). For example, the expected response to an invitation that is delivered as “joking” rather than “serious” is not the acceptance or declination of the said invitation, but rather laughter or some other way of orienting to the invitation as non-serious (Holt 2013). Given what is regarded as play is invariably understood with respect to what is regarded as serious, or non-playful talk-in-interaction, then, it follows that any social action cast as non-serious or playful can be responded to seriously (Gordon 2008; Holt 2013). This means that for any participant who is initiating jocular language play the “challenge is to achieve playfulness while casting actions that can bear serious interpretation – but not too serious, not too close to the bone” (Hopper 1995: 66).3 In other cases, however, seriousness and nonseriousness may be so intertwined as to be inseparable, and so nonseriousness offers a means by which participants can negotiate social action (Holt 2013). Another related analytical theme with regard to the analysis of language play is thus that playful actions may disguise non-playful or serious actions. In other words, participants may engage in more complex forms of play underlying which is not the assumption “this is play”, but rather the question “is this play?” (Bateson 1955). Bateson (1972) suggested that it is in this sense that play can be regarded as “labile” (p. 182), while Goffman (1981) proposed that playful and non-playful frames may be “laminated” in various ways. Gordon (2008) argues that this lamination of play frames may involve either “blending” or “sequential transformation” (p. 320). The former refers to instances where what is meant by the speaker has more than one possible valence (e.g. simultaneously serious and playful), while the latter refers to cases where participants retrospectively alter prior understandings through subsequent signals that they were being playful. Such findings echo work on conversational humor more broadly (Attardo 1994; Martin 2007), where it is suggested that serious and humorous frames may be laminated in ways that allow “speakers to convey serious meanings, while appearing to be ‘only joking’” (Dynel 2011: 227). In particular, non-serious talk

3 To be evaluated as “too close to the bone” refers to something that might be regarded as true and thus hurtful. Given what is regarded as “too close” is inherently subjective it can only be analyzed with respect to recipient response(s) (Hopper 1995).



Jocular language play 

 149

can be used to accomplish serious interactional work in ways that allow the speaker to avoid being held fully accountable (Emerson 1969; Norrick 1993), or to mitigate potentially sensitive acts such as criticisms, reprimands or directives (Dynel 2011) In the remainder of this chapter, building on this previous work, it is argued that to construe actions as (ostensibly) playful or non-serious can be drawn upon as resource by participants in negotiating interpersonally sensitive or delicate social actions. This is illustrated by analyzing the ways in which participants not only perform and respond to particular social actions, such as accusations, complaints, criticisms, disagreements and teases, through jocular language play, but may also (dis)affiliate with the stances or action preference implemented by those social actions.

3 Data and method 3.1 Dataset A collection of 119 sequences containing more than 150 instances of jocular mockery was first assembled. These candidate instances of jocular mockery were sourced from a variety of different corpora of interactions amongst Australian and American speakers of English, including the Australian component of the International Corpus of English (ICE-AUS), the Griffith Corpus of Spoken Australian English (GCSAusE), the Australian Radio Talkback corpus (ART), all of which are now freely available through the Australian National Corpus (www.ausnc.org.au), along with interactions sourced from the Australians Getting Acquainted Corpus (AGAC), and the Corpus of Australians and Americans Talking (CAAT).4 All candidate examples of jocular mockery were transcribed using standard conversation analysis (CA) conventions (Jefferson 2004a). A sub-collection of 20 episodes of jocular language play, which were occasioned by instances of jocular mockery, was then assembled from this broader collection. These episodes of jocular language play were operationally defined as sequences in interaction that were occasioned when a participant diminished something of relevance to self, other, or a non-co-present third party within a playful frame through either (1) manipulation of, or allusion to, language form (i.e. jocular play with language), or (2) co-creation of some kind of made-up or exaggerated scenario or proposal (i.e. jocular play in language).

4 See Haugh (2014: 82; 2016) and Haugh and Carbaugh (2015) for further details about these respective corpora.

150 

 Michael Haugh

3.2 Analytical framework The candidate collection of episodes of jocular language play was then analyzed within an interactional pragmatics framework. Interactional pragmatics is an approach to the analysis of pragmatic phenomena that is informed by research and methods in ethnomethodological conversation analysis (Arundale 2010; Haugh 2012, 2015). However, the focus of interactional pragmatics is not only on the analysis of interactional meanings, but also the analysis of reflexive awareness on the part of participants about those meanings (i.e. metapragmatic awareness), as well as variation in their situated instantiation. Interactional meanings themselves encompass pragmatic meanings (i.e. what participants are taken to be referring to, presuming, saying, implying, inferring and so on), pragmatic acts (i.e. what kinds of socially recognizable acts and activities those participants are taken to be doing), and the interpersonal relationships, identities, evaluations and stances they are taken to be instantiating (Culpeper and Haugh 2014: 267). The current analysis focused on the actual recordings of the interactions concerned, along with their associated transcripts. In the analysis that follows, the notion of (dis)affiliation was also drawn upon in the course of analyzing the design and response features of episodes of jocular language play. Affiliation and disaffiliation concern the affective level of cooperation (Stivers, Mondada and Steensig 2011: 20), building upon and extending Heritage’s (1984) initial characterization of affiliation as a stance that is “supportive of social solidarity” (Heritage 1984: 269). Affiliative responses are, more specifically, those “actions with which a recipient displays that s/he supports the affective stance expressed by the speaker” (Lindström and Sorjonen 2013: 351), displays empathy (Stivers 2008), as well as cooperates with the action preference of the sequentially prior action (Stivers, Mondada and Steensig 2011: 21; Steensig 2013: 944). Disaffilative responses, in contrast, are actions where a recipient does not display support or empathy for the prior speaker’s affective stance, or does not cooperate with the action preference of the prior utterance (Lindström and Sorjonen 2013; Stivers 2008; Stivers, Mondada and Steensig 2011; Steensig 2013), and so may be evaluated as “destructive” of social solidarity (Heritage 1984).

4 (Dis)affiliating through jocular play with language Jocular language play can encompass instances of play with language, or what is sometimes known as wordplay. While this includes instances where there is a play on the semantic level of language use in the form of puns and so on, it may



Jocular language play 

 151

also include playful or creative uses of formulaic language (Bell 2012: 189–190; Carter 2004). Such instances of wordplay have often been analyzed as being in the service of enjoyment or relaxation amongst adults (e.g. Cook 1997), as a way of reinforcing group boundaries and social relationships (Ritchie and Schell 2009), or as a means of developing proficiency in a second language (e.g. Bell 2005; Pomerantz and Bell 2011). However, in this section, we examine cases where participants appear to be engaging in wordplay not only for the sake of their own amusement, but also in order to negotiate, disguise or modulate sensitive social actions where the participants are taking some kind of affective or moral stance. The following excerpt is taken from an initial interaction between Natalie, a postgraduate female in her mid-twenties, and Irene, a primary school teacher in her early thirties. Irene has been talking about her mother who is doing some postgraduate study about women writers in Germany. Given Irene’s description of what her mother has been studying arises subsequent to Natalie talking about what she has been researching as part of her doctoral thesis, they appear to be working up a common topic in the course of getting acquainted, namely, “obscure” topics that are deserving of further research. (2) AGA: NJIK: 1:50 57 I: but ye:ah? I’d never heard of it. and it was just 58 something really obsc- obscure, and it’s amazing 59 when people are doing (0.9) different, well hers 60 is a thesis and [so ] 61 N: [mm.] 62 (0.8) 63 I: people pick up these things and [re]discover all= 64 N: [mm] 65 I: =these really interesting things that like, we wouldn’t 66 [know generally] 67 N: [that we don’t] tend to- to [find] out about normally. 68 I: [mm. ] 69 N: par↑ticularly uh where (.) where women are concerned I 70 think you [do find] 71 I: [↑ye:s ] 72 (0.4) 73 N: all these really? interesting ↑people and they’ve 74 never? had much (.) much (.) much focus? or much 75 attention? and 76 I: mm.

152 

 Michael Haugh

77 N: >whereas< (.) there’re all sorts of obscure men who 78 get books written 79 (0.4) 80 I: ] 81 N: [>(just) about themsolutely< (.) dea:d (.) [£whi:te ↑men£ (.) that’s]= 86 I: [ha ha ha ha ha ha ha ha ] 87 N: =what people say.

←

At the point the excerpt begins Irene is expressing a positive affective stance towards her mother’s choice of topic as “amazing” (line 58) and “really interesting” (line 65), on the grounds that her mother is rediscovering things that are not generally well known (lines 63–66). In overlapping talk, Natalie picks up on this stance in going on to formulate a subsequent generalized complaint (Drew 1998; Heinemann and Traverso 2009), namely, an implicit expression of discontent that interesting women are not treated as worthy of attention (lines 69–75), yet “obscure men” get whole books written about them (lines 77–81). In that way, she construes Irene’s prior description of the research her mother is doing as touching off this complaint. Irene initially strongly affiliates with the negative affective stance implemented by Natalie through the complaint (line 80). However, Irene subsequently engages in wordplay where “history” is articulated as “his story” (line 83). This is evidently designed as jocular through the “smile voice” that marks it as a laughable (Jefferson 2004b; Holt 2010). The episode of jocular language play is further progressed by Natalie through the formulaic expression, “dead white men” (line 85), which occasions laughter from Irene, as well as indicating a potential close of the complaint sequence itself (Drew and Holt 1988). The laughter from Irene expresses appreciation of this allusion to a broader complaint discourse by Natalie (Glenn 2003; Jefferson, Sacks and Schegloff 1987). Notably, then, in the course of this sequence, a complaint about the disproportionate amount of attention paid to “obscure men” at the expense of “interesting women” by scholars and writers is jointly accomplished. However, in affiliating with this complaint, Irene effects a subtle shift in footing from a serious to a non-serious stance through jocular wordplay. In this way, she modulates the degree of seriousness of the complaint. That is, while she treats it is as something about which an expression of moral discontent is indeed legitimate (Drew 1998), it is also something about which they can share in laughter about, and so is perhaps not something to be taken too seriously (Holt 2013). In this

Jocular language play 



 153

way, Irene transforms the affective stance at play here from one that is po-faced and serious, at least as formulated by Natalie, to one that treats the complainable as something that can be laughed at, a stance with which Natalie herself subsequently affiliates. In next example, which also comes from an initial interaction where the participants are getting acquainted, we examine an instance of jocular language play which implements a mocking stance that is ostensibly directed at a third party. However, it appears to also be masking a tease that is directed at that recipient himself. The excerpt begins after Jake, an Australian student who is in his first year at university, and Zane, an American student in his mid-twenties, have just found out that an acquaintance of Jack’s comes from Stockton, California, which is very close to where Zane himself lives. (3) CAAT: AmAus18: 1:27 51 Z: how do you know someone that lives in Stockton.= 52 J: =ah do you know tumblr. 53 (0.8) 54 Z: yea:h. 55 J: °hh° 56 Z: alright 57 (0.4) 58 J: ↑through ↓that (.) 59 J: [through] a fandom [it’s d-] doesn’t really= [°.hhh° ] 60 Z: [alright] 61 J: =[ma]tter hhh= 62 Z: [hh] 63 Z: =and what do they do in Stockton °hh hh° 65 (.) 66 J: he lives the:re. 67 Z: yea:h? (0.4) and [>justjust< tumbles      71 (0.5) 72 J: °n(hh)o hhhheh heh° |((smiles)) 73 Z: |((smiles)) 74 OH no he’s actually American?= 75 Z: =yea:h.= 76 J: =but um: .hh (0.3) he: (.) doesn’t go to >high 77 school there he does< (.) he does it onli:ne?

←

154 

 Michael Haugh

At the point this excerpt begins, Zane is seeking an account as to how Jake knows someone in Stockton (line 51). Jake responds that it is through tumblr, a social networking website that allows microblogging (lines 52–61). However, Jake seems to presume that Zane may not be familiar with it, as he first asks if Zane knows what it is (line 52), and even when Zane indicates that he does (line 54), nevertheless goes on to initiate a subsequently abandoned explanation of what it involves (lines 59, 61). Zane next topicalizes Jake’s acquaintance in Stockton through a continuer question about what the latter does (line 63). However, Jake responds minimally with a hearably non-informative response (line 66), which while flouting the Gricean maxim of quantity (Grice 1975), does not appear to implicate anything in addition to what has been said. This seemingly vacuous response thus occasions an opportunity for a mockery (Haugh 2010), which Zane takes up in line 70 in suggesting, in the form of a candidate answer (Pomerantz 1988) that Jake’s acquaintance “just tumbles”, a playful allusion to the relationship having been formed through tumblr. This is interactionally recognized and ratified as non-serious by Jake, albeit after a brief pause (line 71), through smiling, breathy laughter (Drew 1987; Glenn 2003; Jefferson, Sacks and Schegloff 1987), and a “no”-prefaced return to a serious response to Zane’s question in line 63, which thereby treats Zane’s preceding candidate answer as non-serious (Schegloff 2001). Jake’s construal of the candidate answer as non-serious is ratified by Zane who mirrors Jake’s smile (line 73). However, while the jocular language play in line 70 implements a mocking stance that is ostensibly directed at a third party (i.e. Jake’s acquaintance in Stockton), it also hearably implements a mocking stance towards Jake himself with respect to his prior, seemingly vacuous response, and, perhaps, towards his presumed reliance on tumblr to make friends as well. In other words, while the jocular mockery is interactionally achieved as overtly directed at a third party, it licenses the inference that a tease has also been directed at the recipient himself, namely, a mild sanction of Jake’s lapse (Everts 2003; Haugh 2010; Norrick 1993) with respect to what is expected when getting acquainted, namely, providing responses that do not simply state the obvious. Through jocular language play, then, participants may also mask sensitive social actions within non-serious talk. In the final example that we consider in this section, we can observe how a stance implemented by a prior action may also be modulated through subsequent episode of jocular language play. The excerpt in question comes from a conversation between two male postgraduate students in their mid-twenties, Michael and Cameron, who are both tutors in the economics department at a university in Sydney. It begins with Cameron asking Michael whether he has been to any of the recent honors workshops that tutors are expected to regularly attend.



Jocular language play 

 155

(4) ICE-AUS: S1A-024: 8:26 229 C: °so:¿° (0.5) >did you< go: to >any< (0.4) of the 230 honours workshop? (.) [°talks°] 231 M: [  NO::  ] °cause:° 232 (0.5) 233 C: °you ↓busy to[day°] 234 M: [I'm ] gonna go (.) >this afternoon 235 thoughfor ↓lunch.< (0.4) 238 239 M: ye:ah >gonna go for lunch and I’ll go to< o:ne or two. 240 (.) 241 C: what’s on for lunch. (.) s- sandwiches ag[ain] 242 M: [it’]s just 243 sandwiches °ye:ah°. 244 (0.6) 245 M: ↑if they wanna provide lunch, I’ll ↑eat it. 246 (0.3) 247 C: well¿. 248 (0.5) 249 M: tha(hh)t’s- hh 250 C: £who said there is n(hh)o fr(hh)ee ↑lunches£   251 (.) 252 M: £supply creates its own de↑ma:nd.£ 253 C: £that’s ↑ri:ght.£ (.) [>heh heh heh ] heh< 254 M: [£Sa:y’s law.£] (.) 255 256 C: °he°

←

While Michael initially proffers the beginnings of an account for having not attended the workshops in line 231, this account trails off and is left incomplete. Cameron then proffers a candidate account for this lack of attendance through a polar question (Pomerantz 1988), namely, the presumption that he is too busy (line 233). Given this question presupposes that Michael could attend the workshops if he were not too busy, the pursuit of a response from Michael by Cameron is indicative of an agenda underpinning Cameron’s prior question in lines 229–230 (Pomerantz 1984). Michael’s subsequent indirect response orients to this agenda through a non-type-conforming response (Walker, Drew and Local 2011), in which he confirms that he will be attending that afternoon

156 

 Michael Haugh

(lines 234–235), rather than indicating whether or not he is too busy. Cameron’s subsequent question in line 237 is thus hearable as a teasing accusation (Partington 2008) that Michael is only attending the workshops in order to get lunch. This teasing accusation implicates the premise (Sperber and Wilson 1995) that going to the workshops in order to obtain a free lunch is not a good thing. Michael initially disattends this implicit accusation in responding only to the literal import of Cameron’s prior question (line 239), which touches off a question and answer sequence about what they will be getting for lunch (lines 241–243). However, he subsequently returns to address it through expressing a stance about his plan to go for lunch despite not attending all of the workshops, namely, that he is quite happy to take up an offer of free food (line 245). Notably, this stance is framed within a hearably defiant key through the markedly higher pitch and intonational emphasis with which it is delivered. Cameron then responds in line 250 with an instance of play on formulaic language (Bell 2012), specifically, a playful allusion to the well-known adage “there’s no such thing as a free lunch”, and through taking an evidently ironic footing (Clift 1999), implements a mocking stance towards this piece of conventional wisdom. The questioning of this wellknown expression is delivered with a smile voice and interpolated laughter particles (Jefferson 2004b; Holt 2010) that invites Michael to treat such conventional wisdom as laughable, because in this case Michael is indeed getting a free lunch. Michael then affiliates with this mocking stance through a playful invoking in line 252 of another well-known expression – at least to those who have studied economics – from the French economist Jean-Baptiste Say. Critically, both participants would be aware that this position has been subsequently discredited in mainstream economics, and so Michael is also taking a mocking stance towards conventional wisdom in economics, given the offer of a free lunch does appear, in this case, to be generating demand for free lunches. This mocking stance is also signaled through smile voice that invites Cameron to treat this conventional wisdom as a laughable. Cameron subsequently agrees and proffers laughter in response (line 253), while Michael explicitly names the source of the well-known claim in economics that was alluded to in his previous utterance. What is important to also note here, however, is that aside from jointly accomplishing a mocking stance through jocular play with language, Cameron is also modulating the grounds of his prior teasing accusation that was directed towards Michael. While the original tease presupposed that going to the workshops in order to get a free lunch was not a good thing, in the subsequent episode of jocular language play it is presumed that getting a free lunch is a good thing, and those who think such things cannot happen are being proven wrong by the fact Michael is able to get lunch for free in this case. In this way, the teasing accusation is recast as an appreciation for Michael’s plans to take advantage of



Jocular language play 

 157

­ pportunity to get a free lunch. Notably, it is Cameron who initiates the modulao tion of his own prior tease through the playful allusion, something which is signaled through well-prefacing (line 247). In this case, the well-prefacing arguably implements a “my side alert”, that is, where “a first speaker’s characterization of some state of affairs is matched by a second speaker’s corroborative description that is specifically founded in the second speaker’s experience, knowledge or umwelt” (Heritage 2015: 98). In other words, while going to workshops just to get a free lunch legitimately opens the target (here Michael) to criticism or derision, as per Cameron’s previous tease, Cameron subsequently takes a different stance which undermines the grounds for that teasing accusation. In this section, we have seen how through jocular play with language, participants may negotiate or modulate the degree of seriousness of a delicate or sensitive social action. It has also been suggested that this modulation or negotiation occurs in ways that allow participants to alter the grounds on which they are affiliating with a jointly accomplished action or prior accomplished action, or even to mask a sensitive social action. In the following section, we move to show how sensitive social actions can also be accomplished through jocular play in language.

5 (Dis)affiliating through jocular play in language Jocular language play also encompasses instances of play in language, or what is sometimes known as “fantasy layering” (Clark 1996), “collaborative fantasy” (Norrick 2000), “fantasy humor” (Hay 2001) or “joint fantasizing” (Kotthoff 2007). This kind of fantasy-based play involves the “the construction of humorous, imaginary scenarios or events” (Hay 2001: 62), including “narrative-like fantasies” (Norrick 2000: 126). Such instances of language play have generally been analyzed as being in the service of enjoyment or relaxation amongst adults (e.g. Béal and Mullan 2013; Kotthoff 2007), or as a means of establishing solidarity or rapport amongst those participants (e.g. Hay 2001; Norrick 2000). Dynel (2011), for instance, argues that “fantasy layering serves as the bedrock for pretence-based humor, where practically no informative content is conveyed, interpersonal effects aside…such as solidarity building or a testament to the speaker’s wit and intelligence” (p. 226). However, in this section we examine instances where participants appear to be engaging in this kind of jocular language play in order to negotiate, mask or modulate sensitive social actions through which participants are implementing some kind of affective or moral stance, in addition to the pleasure or reinforcement of social bonds they may also gain from engaging in such episodes of jocular language play.

158 

 Michael Haugh

In the following example, we examine how joint fantasizing may modulate the degree of seriousness of the stance that an action implements. The excerpt is taken from a conversation between four friends in their mid-twenties who are chatting at their tennis club, Carly, Julia, Sarah and Tony. They have been talking about other members of the club introducing themselves before games. The excerpt begins with Julia taking the stance that she thinks it is “nice” when people introduce themselves. (5) ICE-AUS: S1A-031: 8:11 305 J: but it’s ni:ce. it is nice [to know who ] you’re [↑ye:ah, it is.] 306 C: 307 J: playing before you got names and that you’re not 308 going ↑Hey yo:u↑ [( )] 309 T: [I dunno. I feel ] really 310 suspicious you know if someone does that? 311 C: yHa ha [ha ha ha] [°ha ha he he he hh.°     ] 312 T: [somebody] [goes ↑He:llo I’m George↑] 313 J: [ha ha ha] 314 (0.4) 315 T: what he’s after. 316 C: ha [ha] 317 T:     [he]’ll sell me some insurance after or 318 something      319 C: ha [ha ha [ha he 320 J:     [ha ha 321 S: [he he

←

←

While Carly affiliates with this stance (line 306), Tony takes a more disaffiliative stance in not agreeing but rather expressing doubt, and then going on to claim that he feels “suspicious” when people introduce themselves in such contexts (lines 309–310). Carly responds with laughter, thereby construing Tony’s disaffiliative stance within a jocular frame (Haugh 2010; Jefferson, Sacks and Schegloff 1987), while Tony goes on to offer an account as to why he feels suspicious in such situations through acting out an exaggerated scenario in which such a person has an underlying agenda, namely, to sell him insurance or the like (lines 312, 315, 317–318). This is also construed as jocular through laughing responses from the other participants, including Julia. Through this joint fantasizing episode, which builds off a possible although unlikely scenario, Tony modulates the degree of seriousness of his disagreement with Julia’s affective stance. In other words, while he appears to be disagreeing, the degree to which this response can be



Jocular language play 

 159

regarded as disaffiliative is modulated through constructing an exaggerated fantasy scenario, which is treated by the other participants as jocular, and so ostensibly at least, non-serious. Other sensitive social actions, such as complaints, may also be modulated through episodes of joint fantasizing. In the following excerpt from an initial interaction between Natalie, a postgraduate student in her mid-twenties, and Aaron, a lecturer in his late-thirties at an Australian university, they have been talking about staff and postgraduate students having to move out of a building at the center of the campus to one that is less conveniently located. Andy proposes one advantage of such a shift, namely, that she would get air-conditioning in a new office. (6) AGA: NJAB: 2:32 70 A: well at least there’s air conditioning. 71 (0.4) 72 N: mm 73 (0.3) 74 A: [yea:h ] 75 N: [°might] be something° 76 (0.2) 77 A: yes it looks ah: (0.8) it has been erratic when 78 they first insta:lled it (0.4) it got extremely 79 co:ld and you could feel >that< ˚sort of˚ (0.3) 80 filing cabinets °sort of° (1.0) felt like the 81 inside of a refrigerator and 82 (0.3) 83 N: £sounds ↑lovely£ ha= 84 A: =could’ve chilled £beer in your fi:ling cabinet.£  85 ha ha[>he he he he he he he he< he he he he he] 86 N:  [£perhaps that would’ve been a good idea£] .hh ha 87 A: ha .hh (0.3) well it didn’t didn’t sta:y that way 88 (0.2) completely sort of erratic but uh 89 N: mm 90 A: ↑seems to be (0.3) they seem to have it programmed 91 better (now but). 92 N: mm this feels fine

←

Natalie offers only a minimally affiliative response (line 72), and then hedges her agreement in suggesting it only “might” be an advantage (line 75). This mildly affiliative response from Natalie appears to occasion a subsequent complaint from Andy about the air-conditioning in the newly refurbished offices being far

160 

 Michael Haugh

too cold (lines 77–79), and likened to the inside of a refrigerator (lines 80–81). Natalie subsequently responds with an ironic positive assessment, which is framed as jocular through smile voice that treats the situation of the cold office as a laughable (Jefferson 2004b; Holt 2010), and turn-final laughter that invites Andy to laugh at the situation he found himself in (Glenn 2003; Jefferson 1979). In doing so, although Natalie treats this situation as a legitimate source of complaint (Drew 1998), she suggests at the same time that it also constitutes a laughable. Andy then proposes a mock scenario in which he could have kept beer chilled in the filing cabinet in his office (line 84). This is framed as non-serious through smile voice that treats the exaggerated degree of cold as a laughable, as well as the subsequent extended laughter whereby he responds to and thus treats his own mock suggestion as jocular (line 85). Natalie agrees with this proposal (line 86), although this agreement is implemented with a somewhat ironic footing (Clift 1999) that is indicated through smile voice and turn-final laughter. Andy responds with brief laughter (line 87), before going on to say the coldness subsequently abated (lines 87–88, 90–91). Notably, while Andy’s complaint prior to Natalie’s ironic stance is delivered as serious, through his subsequent mock suggestion he modulates the degree of seriousness of his complaint through exaggerating how cold it was. This enables him to affiliate with Natalie’s mocking stance.5 The degree of seriousness of a jointly accomplished complaint can also be modulated through jocular language play, as we can see in the following example from an initial interaction between Norma, an activities officer, and Toby, a university administration officer, both of whom are in their mid-thirties. Norma has been complaining that she is not able to work in an aged care facility or a pub in Australia despite having prior experience, because she lacks the appropriate certification. (7) CAAT: AusAus09: 9:25 367 N: it’s just like, come on the:se are jobs that 368 (0.3) 369 T:   >thatI mean okay fine the R S A’s< not much you know 380 you pa:y () sixty-five seventy bucks or something [so [uh 381 T: 382 takes a few hours or hh (.) [yep ] 383 N: [>buthaha[ha< and ] £tying yo[ur shoelaces£ 391 T: [>gettinghaha[hahayeah yeah you knowbut I mean< (0.2) I: dunno. (.) Australians 456 seem to love their bureaucracy 457 it’s just (0.2) [>°you know° (2) 7 David: = 8 All: = [@*@@ = 9 Marek: = [(xxx)* 10 Juan: = = = @@@@@ = 11 All: 12 David: = you’re having a heart attack now? (.) oh really? (.) OK (.) send me the figures first = 13 Others: = @@@ = 14 Juan: = (.) ((raises hands in placation)) please As with his previous utterance “mmm-hmm”, Marek’s “yeah yeah” (line 3) triggers laughter. However, on this occasion, Juan transforms Marek’s seeming indifference into cold-hearted callousness as he plays out a scenario where the latter presses for a favor while his interlocutor breathes his last. This is reinforced by a dismissive hand gesture (line 6). The absurdity of the created scenario is added to by Juan as he juxtaposes incongruous elements, imagining the need to remember the social niceties in such an exchange by adding “please” (lines 10 and 14). Interestingly, it is again Marek’s mistake which Juan is happy to exploit here. Marek laughs freely at these episodes rather than taking offence. In part, this seems to be because Juan’s contributions give the impression that he is humorously pointing to their common predicament, something that becomes explicit some hours later at the end of training, when the group is packing up and disbanding for the day: Episode 4: The Call Exercise was Very Fun 1 Juan: the (.) the (.) the call exercise was er was very fun @@@ (2) 2 David: the core exercise (.) did you say? (.)



Building rapport and a sense of communal identity 

 233

3 Juan: the? (.) 4 David: the what exercise? (.) the call (.) 5 Juan: 6 David: oh the call (.) yeah yeah = 7 Juan: = was very fun(.) for me (2) 8 David: yes I think we had a (2) a slight identity crisis but (1) 9 Marek: < @ yeah we had > (2) 10 Juan: < @ and you have the video > = 11 David: = < @ I have (.) I have all the evidence I need > = 12 Others: = @@@@@@ = 13 Juan: = Harry is dead (.) OK = 14 All: = @@@@@@ 15 Juan: ((over laughter)) < @ where is my figure (.) I have no computer (1) OK (2) it’s your problem It is noteworthy here that Juan explicitly says how much he enjoyed the telephone calls. Yet, as his comment about the video (line 10) suggests, he is aware of the way in which the simulations highlight the learners’ own shortcomings when communicating in English and, thus, have the potential to embarrass and even humiliate. Despite this, he takes pleasure in evoking the moment through reconstructing and voicing the conversations. Again, he utters something which was not said in the simulations themselves – “Harry is dead” (line 13). However, this reinforces a scenario which, of course, is of his own invention (see Episode 3). Furthermore, he evokes his own particular telephone call when Harry told him that the computer was down. It is noteworthy here that “OK” is beginning to assume a metonymic status, representing the seeming indifference to their interlocutor that both he and Marek display during their phone conversations. In this regard, it is useful to introduce a couple of useful concepts from linguistic anthropology. These are the related notions of entextualization and recontextualization. The first of these is defined by Bauman and Briggs (1990: 73) as “the process of rendering discourse extractable, of making a stretch of linguistic production into a unit – a text – that can be lifted out of its interactional setting”. In this case, the entextualized unit is the word “OK”, rendered memorable and extractable by the play around it in Episodes 1 and 3. This is then taken from its original context and recontextualized in yet another of Juan’s invented dialogues. The laughter that Juan triggers in Marek and David is precisely because of the significance it has accrued with each recontextualization. Furthermore, the utterance “it’s your problem”, like “OK”, explicitly embodies the attitude that he and Marek inadvertently create during the t­elephone conversations rather than echoing anything they actually say at the time of the calls.

234 

 David Hann

It is noteworthy that Juan seems to actively seek out moments where he can indulge in self-denigrating humor. This seems to be an important strategy in terms of what Brown and Levinson call “positive face” which requires that the individual’s “wants be desirable to at least some others” (Brown and Levinson 1987: 62). Although, in explaining their influential politeness framework, they say nothing about humorous language play, they do make a passing reference to joking as a positive politeness strategy (Brown and Levinson 1987: 102). It has been left to others (e.g. Boxer and Cortés-Conde 1997: 281; Norrick 1993: 47) to point out that such behavior can actually enhance the speaker’s positive face by showing him or her not to be a threat and to be approachable. In addition, admitting to such failures shows a certain composure and control in that it demonstrates self-awareness, even under stress (Dynel 2009: 1295). Furthermore, by sharing their weaknesses with others, speakers implicitly send out the message that they have trust in their audience (Holmes 2000: 170). The self-denigrating nature of much of the humor between this pair is further evident in the anecdotes they tell each other (Hann 2013: 273). Immediately following on from Episode 4 are a series of exchanges as the group takes leave of each other which shows Juan blending the simulated world and the here-and-now of “reality”. David refers to an important football game that evening between Chelsea and Barcelona which Juan, a keen supporter of the Spanish team, is planning to watch: Episode 5: Tomorrow Juan is Going with Harry 1 David: OK guys (.) 2 Juan: (.) 3 David: so have a good eveni- are you beginning to feel nervous? (1) is the stomach going? (2) 4 Juan: yes (2) [I’m concentrate*   [@@@* 5 Marek: 6 David: how long have you got (.) I’m concentrate (2) 7 Juan: 8 David: mm (.) it’ll all be over tomorrow (.) If er (.) we will know the result if- if Juan doesn’t turn up 9 tomorrow [morning* 10 Others: [@*@@ (2) I have butterfly in my [stomach* 11 Juan: [butterflies in* your stomach (3) 12 David: 13 Marek: @@[@* 14 David: [it’s a* serious matter (.)



Building rapport and a sense of communal identity 

 235

15 Marek: in every case (.) = 7 Juan: 8 Others: = @@[@* 9 Juan: [don’t worry* = 10 David: = < @don’t worry don’t worry (.) don’t take it personally (.) OK (.) 11 Juan: be careful with the OK = 12 Marek: = @@@ < @ OK > Both learners laugh at the prospect of David speaking about them to their respective future teachers. Juan pinpoints their use of “OK” as something that characterizes them both. He does this by voicing David in the imagined conversation to come. As in Episode 7, the use of the collective pronoun to associate them both with the expression is significant, as is the distinctive realization of “OK”. Unlike

238 

 David Hann

the previous examples of Juan re-enacting the telephone call, he does not reframe their use of “OK” as indicative of an indifferent attitude. Indeed, his reference to them being “good guys” implicitly concedes that their use of the term shows up their shortcomings when using English rather than revealing anything about their characters. Throughout the exchanges featuring the use of “OK”, Juan’s humor lies in pretending that he and Marek were intentionally rude and, in so doing, acknowledging that they were, in fact, unintentionally so. In the preceding episodes we see the gradual development of the significance of the word “OK”, until it becomes an integral part of the pair’s identity. In this regard, my BizLang colleague observes: there is that ‘running joke’ feel to this... there is a voicing, reliving of the earlier conversation, a reinvention.. and it has become a ‘humor touchpoint’… all they will need to say is ‘OK’ for the rest of the week, and they will have this release of laughter. There’s a joy in discovering these touch points, and playing on them.

Juan uses “OK” to represent their collective shortcomings in the telephone simulation. He evokes the conversation by re-enacting it. However, it has been noted before that reported speech in everyday conversation is, in effect, a construction (Tannen, 2007: 132). Juan’s re-enactments are not attempts at an accurate recreation of what was originally said but creative and symbolic constructions of his own making, for social and humorous ends. It is noteworthy that the errors that are playfully exploited in the data can be categorized as performance mistakes. They involve pragmatic shortcomings in the learners’ reactions to particular moments. They do not involve errors with formal properties of the language, such as tense, word order or semantic meaning (Corder 1981).4 Furthermore, they cast doubt on Aston’s (1993: 229) assertion that “[r]ole-played interactions are without effective social consequences, since the relationships between characters are, in the final analysis, fictional and temporary”. In one sense, role-plays are indeed fictional and temporary. However, the comment fails to take account of the lamination of frames (Goffman, 1974: 82) pertaining at any one time in a classroom simulation which allow real relationship work to be undertaken in imagined scenarios.

4 Corder (1981) differentiates “errors”, which he says as reflective of the learner’s present state of language knowledge, from “mistakes” which are simply indicative of moment-to-moment performance. I do not differentiate these terms in my discussion.



Building rapport and a sense of communal identity 

 239

6 Conclusion Some might question whether the utterances featured in this chapter are humorous or have very much to do with language. After all, on the face of it, Juan’s repetition of the word “OK” hardly seems linguistically inventive. However, such an observation would not take account of the way in which he uses the word metonymically. It is distilled to represent the learners’ shared experiences together and their shared limitations in the TL. As such, its significance can only be truly appreciated by those within his group. Furthermore, he re-creates dialogue for humorous (or perhaps I should say, recreational) purposes and nimbly moves between and exploits different frames. Through his recreations, Juan highlights something which William Hazlitt, the early 19th century writer and critic, sees as being at the heart of the nature of comedy. He describes humans as “the only animal that is struck with the difference between what things are and what they ought to be” (Hazlitt, from a lecture in 1818, cited in Morreall 1987: 56). Thus, Juan’s use of the word “OK” represents the difference between the English level he has reached and the level he strives for as he re-imagines the unintentional impression the word creates as an intentional selfcentered response to his interlocutor’s plight. Also, by turning this gap into comedy, he cathartically transforms the learners’ potentially negative experiences into positive ones which they can literally laugh off. Indeed, he is able to take his and Marek’s shortcomings in performance and make them part of their playful repertoire. In doing so, he takes control of a situation in which he might otherwise feel helpless. So, despite the learners’ pratfalls, or indeed, maybe because of them, Juan is able to declare in all sincerity that “the call exercise was very fun”. This psychological dimension to play helps explain why it can be found in the seemingly unpromising context of an intensive hot-housing English-for-Business course as well as the more informal, relaxed settings where it is usually thought to thrive (Carter 2004: 165). The data here, like Holmes’ (2000; 2007) investigations among L1 users in the workplace, seems to suggest that humorous play can relieve tension, indicating that work and play are perhaps not as mutually exclusive as some might think and, indeed, as Cook posits (2000: 150), may overlap. In terms of the group’s culture, Juan’s stylized realization of the word “OK” shows how allusions to shared experience can be ritualized with their repetition. As such, they begin to symbolize a common history and sense of community. Although it is primarily Juan who plays in the featured episodes, his play is validated by the other members of the group: David sometimes builds on his contributions, and both David and Marek demonstrate their appreciation of them through laughter. In the latter’s case, this appreciation is despite Juan’s potentially face-threatening act of making fun of Marek’s mistakes. Juan transforms his references to these mistakes into an act of solidarity with his fellow learner by making clear that Marek’s shortcomings

240 

 David Hann

are also his own. In addition, Juan’s reconstructed dialogues create a mythology rather than a history, one that the learners feel free to revisit. As Armstrong (2005: 111) comments when discussing the importance of myth in human history, “a myth (...) is an event that – in some sense – happened once, but which also happens all the time.” For the learners in this group, the myth of their indifferent attitude on the telephone is one which endures for the length of their stay together. In this mythologizing process, the word “OK” is an integral part: the event which “happens all the time” is one evoked by its frequent recontextualization, allowing the original context to be re-imagined every time it is referred to. Play, therefore, seems to be central to the process of making a community of practice with its own culture, one with an emergent history, language, mythology and ritual. It is noteworthy that the typical language class often has role-plays and revision sessions, making it an environment which is rich with possibilities to playfully recycle and recontextualize language items as part of the culture-building process. The humorous language play featured above is not simply of a type that L2 users fall back on because they lack the verbal dexterity of an L1 quipster. As Baynham (1996) comments on the native-speaking adult numeracy classes that he investigates, “[t]here are examples in the data of exchanges that clearly refer to on-going, in-group, joking, the full meaning of which it is hard for the analyst/ outsider to gain access to” (p. 194). Indeed, the incremental nature of much play can be evidenced not only in social groups such as language classes or family units, but also in contexts where the in-group may be far more ephemeral. Take, for example, a radio or TV audience. Tune into a comedy program in either of these media when it is halfway through and see how much of the humor you understand. You may well be surprised to find out that it often depends on preceding talk or events within the program which you won’t have access to unless you have watched or listened from the beginning. Needless to say, it is a risky business to make generalizations about humorous language play among L2 learners based on the evidence of only one pair of learners and their teacher. For a start, it is clear that there are differences in the play behavior exhibited by Marek and Juan: the former plays far less than the latter (although the “OK thread” does not do justice to his playful side, occurring elsewhere in the original collected data). Furthermore, in the broader research of which this is a part (Hann 2013), there are clear differences in the collective playful behavior of the four groups investigated. However, all of them exploit the frame-rich environment of the classroom for their own comedic ends and all make particular words and phrases into important collective reference points that they can return to and have fun with. The possible role of humorous language play in the language acquisition process has been beyond the scope of this chapter. However, even the limited



Building rapport and a sense of communal identity 

 241

evidence here suggests that playing in the TL can aid acquisition. Juan’s extending and enacting of particular scenarios indicates that he was able to extend his own repertoire through play. Hall (1995: 218) posits the notion that becoming ­competent in a language involves “ventriloquating” i.e. developing a range of voices. Furthermore, it could be posited that the heightened affective sense that seems to accompany play may well help in making language memorable, given the importance in SLA literature of such concepts as noticing (Tomlin and Villa 1994) and attention (Schmidt 1998). The role of humorous language play in the acquisition process is an avenue for further investigation. Suffice it to say for the moment that its role in making a classroom of learners into a cohesive social group must be of benefit: the members of such a group have a far greater chance of working together efficiently in the task of learning a TL than those who have not cemented their relationships with the social glue of humorous language play.

Transcription conventions rising intonation ? pause (shorter than a second) (.) pause (a second or longer, timed to the nearest second) (2) starting point of an overlap [ ending point of an overlap * turn-continuation or latching (no discernible gap between turns) = speaker's incomplete utterance teparalinguistic and non-verbal activities ((activity)) laughter (each '@' representing one 'syllable' of laughter) @@@

spoken while laughing unintelligible speech; x marking approximate syllable number (xxx) assumed utterance (text) heard through speakers (recorded or on phone) ((speaker's name)) text Note: Capitals are used for the first person subject, for proper nouns or to indicate the use of acronyms e.g. ‘IT department’

References Adolphs, Svenja & Ronald Carter. 2007. Beyond the word. European Journal of English Studies 11: 133–146.

242 

 David Hann

Appel, Joachim. 2007. Language teaching in performance. International Journal of Applied Linguistics 17: 277–293. Armstrong, Karen. 2005. A short history of myth. Edinburgh: Canongate. Aston, Guy. 1993. Notes on the interlanguage of comity. In Gabriele Kasper & Shoshana Blum-Kulka, (eds.), Interlanguage pragmatics, 224–250. New York: Oxford University Press. Attardo, Salvatore. 1994. Linguistic theories of humor. Berlin: Mouton de Gruyter. Bakhtin, Mikhail. [1935] 1981. The dialogic imagination. Austin: University of Texas Press. Bakhtin, Mikhail. [1929] 1984a. Problems of Dostoevsky’s poetics. Minneapolis, MN: University of Minnesota Press. Bakhtin, Mikhail. [1965] 1984b. Rabelais and his world. Bloomington, IN: Indiana University Press. Bateson, Gregory. [1955] 1972. A theory of play and fantasy in Steps to an ecology of mind. 177–193. New York: Ballantine. Bauman, Richard & Charles Briggs. 1990. Poetics and performance as critical perspectives on language and social Life. Annual Review of Anthropology 19: 59–88. Baynham, Mike. 1996. Humour as an interpersonal resource in adult numeracy classes. Language and Education 10: 187–200. Bell, Nancy. 2005. Exploring L2 language play as an aid to SLL: A case study of humour in NS–NNS interaction. Applied Linguistics 26: 192–218. Bell, Nancy. 2007. How native and non-native English speakers adapt to humor in intercultural interaction. Humor 20–21: 27–48. Block, David. 2007. Second language identities. London: Continuum. Boxer, Diana & Florencia Cortés-Conde. 1997. From bonding to biting: Conversational joking and identity display. Journal of Pragmatics 27: 275–294. Broner, Maggie & Elaine Tarone. 2001. Is it fun? Language play in a fifth-grade Spanish immersion classroom. The Modern Language Journal 85: 363–379. Brown, Penelope & Stephen Levinson. 1987. Politeness. Cambridge: Cambridge University Press. Carter, Ronald. 1999. Common language: Corpus, creativity and cognition. Language and Literature 8: 195–216. Carter, Ronald. 2004. Language and creativity: The art of common talk. London: Routledge. Carter, Ronald. 2006. Common talk [CD-ROM]. The art of English Milton Keynes: The Open University. Cekaite, Asta & Karin Aronsson. 2005. Language play: A collaborative resource in children’s L2 learning. Applied Linguistics 26: 169–191. Chandler, Daniel. 2002. Semiotics: The basics. London: Routledge. Coates, Jennifer. 2007. Talk in a play frame: More on laughter and intimacy. Journal of Pragmatics 39: 29–49. Cook, Guy. 2000. Language play, language learning. Oxford: Oxford University Press. Corder, Stephen Pit. 1981. Error analysis and interlanguage. Oxford: Oxford University Press. DaSilva Iddings, Ana & McCafferty, Steven. 2007. Carnival in a mainstream kindergarten classroom: A Bakhtinian analysis of second language learners’ off-task behaviors. The Modern Language Journal 91: 31–44. Du Bois, John. 1986. Self-evidence and ritual speech. In W. Chafe and J. Nichols (eds.), Evidentiality. 313–336. New Jersey: Ablex.



Building rapport and a sense of communal identity 

 243

Dynel, Marta. 2009. Beyond a Joke: Types of conversational humour. Language and Linguistics Compass. 3: 1284–1299. Eckert, Penelope & Sally McConnell-Ginet. 1992. Think practically and look locally: Language and gender as community-based practice. Annual Review of Anthropology 21: 461–490. Glenn, Phillip. 2003. Laughter in interaction. Cambridge: Cambridge University Press. Goffman, Erving. 1959. The presentation of self in everyday life. London: Penguin. Goffman, Erving. 1974. Frame analysis: An essay on the organization of experience. Boston: Northeastern University Press. Goffman, Erving. 1981. Forms of talk. Philadelphia: University of Pennsylvania Press. Gordon, Cynthia. 2008. A(p)parent play: Blending frames and reframing in family talk. Language in Society 37: 319–349. Gumperz, John. 1982. Discourse strategies. Cambridge: Cambridge University Press. Hall, Joan Kelly. 1995. (Re)creating our worlds with words: a sociohistorical perspective of face-to-face interactions. Applied Linguistics, 16: 206–32. Hann, David. 2013. A study of the playful use of English among low-proficiency language learners on an intensive business English group course. Milton Keynes: The Open University PhD thesis. Holmes, Janet. 2000. Politeness, power and provocation: How humour functions in the workplace. Discourse Studies 2: 159–185. Holmes, Janet. 2007. Making humour work: Creativity in the job. Applied Linguistics 28: 518–537. Hoyle, Susan. 1993. Participation frameworks in sportscasting play: Imaginary and literal footing. In Deborah Tannen (ed.), Framing in discourse, 114–145. Oxford: Oxford University Press. Huizinga, Johan. [1944] 1970. Homo ludens. London: Paladin. Janesick, Valerie. 1998. The dance of qualitative research design: Metaphor, methodolatry and meaning. In Norman Denzin & Yvonne Lincoln (eds.), Strategies of qualitative inquiry, 35–55. Thousand Oaks, CA: Sage Publications. Jenkins, Jennifer. 2000. The Phonology of English as an international language. New models, new norms, new goals. Oxford: Oxford University Press. Kotthoff, Helga. 2003. Responding to irony in different contexts: on cognition in conversation. Journal of Pragmatics 35: 1387–1411. Lantolf, James. 1997. The function of language play in the acquisition of L2 Spanish. In William Glass & Ana Perez-Leroux (eds.), Contemporary perspectives on the acquisition of Spanish, 3–24. Somerville, MA: Cascadilla Press. Lantolf, James. 2000. Introducing sociocultural theory. In James Lantolf (ed.), Sociocultural theory and second language learning, 1–26. Oxford: Oxford University Press. Lave, Jean & Etienne Wenger. 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press. Long, Michael & Graham Crookes. 1992. Three approaches to task-based syllabus design. TESOL Quarterly 26: 27–56. Maybin, Janet. 2006. Children’s voices: Talk, knowledge and identity. Basingstoke: Palgrave Macmillan. Morreall, John (ed.). 1987. The philosophy of laughter and humor. Albany, NY: State University of New York. Norrick, Neal. 1993. Conversational joking: Humor in everyday talk. Bloomington: Indiana University Press.

244 

 David Hann

Rampton, Ben. 2006a. Language in late modernity: Interaction in an urban school. Cambridge: Cambridge University Press. Rampton, Ben. 2006b. Language crossing. In Janet Maybin & Joan Swann (eds.), The art of English: Everyday creativity, 131–139. Basingstoke: Palgrave Macmillan. Rampton, Ben. 2007. Neo-Hymesian linguistic ethnography in the United Kingdom. Journal of Sociolinguistics, 11: 584–607. Raskin, Victor. 1985. Semantic mechanisms of humor. Dordrecht: D. Reidel Publishing Company. Saussure, Ferdinand de. [1916] 1959. Course in general linguistics. London: Peter Owen Ltd. Schmidt, Richard. 1998. The centrality of attention in SLA. University of Hawai’i Working Papers in ESL, 16: 1–34. Skehan, Peter. 1998. A cognitive approach to language learning. Oxford: Oxford University Press. Street, Brian. 1993. Culture is a verb: anthropological aspects of language and cultural process. In David Graddol, Linda Thompson & Mike Byram (eds.), Language and culture. British studies in applied linguistics, 7. Clevedon, UK: Multilingual Matters. Swann, Joan and Maybin, Janet. 2007. Introduction: Language creativity in everyday contexts. Applied Linguistics 28: 491–496. Symons, Donald. 1978. The question of function: Dominance and play. In Euclid Smith (ed.), Social play in primates, 193–230. New York: Academic Press. Tannen, Deborah. 2007. Talking voices: Repetition, dialogue, and imagery in conversational discourse. 2nd Edition. Cambridge: Cambridge University Press. Tomlin, Russell & Victor Villa. 1994. Attention in cognitive science and second language acquisition. Studies in Second Language Acquisition, 16: 183–203. van Lier, Leo. 1988. The classroom and the language learner. Harlow: Longman. Victoria, Mabel. 2011. Building common ground in intercultural encounters: A study of classroom interaction in an employment preparation programme for Canadian immigrants. Milton Keynes: The Open University PhD thesis.

Jet Van Dam and Anne Bannink

10 The first English (EFL) lesson: Initial settings or the emergence of a playful classroom culture Abstract: This discourse-based micro-ethnographic study scrutinizes the various forms of play that are a recurrent and salient feature of a first English (EFL) lesson in a Dutch secondary school context. In the first part of the lesson collective speaking formats associated with cultural games and routines in early L1-acquisition overwhelmingly have the floor. All-class chorusing, rhymes and chants create affordances for off-record and intermittent participation modes in peer-scaffolded speaking and whispering slots. They mediate the transition from (over)hearer to (co)speaker and allow learners to find their voice in the new language – on the basis of self-selection. Individual student voices (even disrespectful ones) are also dialogically modeled by the teacher for taskrelated ‘serious’ business in playful asides in the shared L1 metalanguage (duallanguage teaching strategies). Cued by prosodic shifts, rhythmic variation and other multimodal/semiotic behavioral features, these complex footing changes (Goffman 1979) and (re)framing practices are taken up by individual learners towards the end of the lesson. In the context of an IRF sequence that elicits formulaic phrases in the L2, they spontaneously create virtual identities and playful speaking roles for themselves. Instances of play may, however, present interpretive problems for next speakers. Since play is by definition parasitic on other templates or interactional norms, the challenge is to reflect online created ambiguities in the models we develop to articulate discourse complexity in multiparty classroom floors. 

1 Introduction The first day of school is an important event in the lives of students and teachers alike. Recently The New York Times featured extended coverage of the imminent return to school of primary schoolers. The accompanying website mentioned an effective and skillful teacher as the number one factor in determining learner success and failure. It also stressed the importance of effective classroom management skills with an emphasis on clear instructions, procedures and ­routines:  “What you do on the first day of school will determine your success for the rest of the year. You will either win or lose your class on the first days of DOI 10.1515/9781501503993-011

246 

 Jet Van Dam and Anne Bannink

school.” This quote resonates with a long-standing research interest in the mediating role of relevant others – parents, siblings, peers – in scaffolding early language acquisition and socialization (e.g. Duranti et al. 2011; Ochs and Schieffelin 1983; Schieffelin 1979). Investigations into the early language learning experiences of a community of learners are rare, however, with Cekaite’s (2006) in-depth study as a notable exception. In this chapter we present a detailed investigation of a very first English (EFL) lesson involving a class of 12- or 13-year-olds in the Netherlands. More in particular we zoom in on the many instances of play that were in evidence throughout this lesson. While in a general sense the interrelatedness of play and (language) learning is hardly contested (cf. Bateson 1972; Cook 2000; Huizinga 1951), its precise role in mediating and synchronizing the learning of many in institutional settings is, as yet, not well-understood. We hope our study contributes to the rich and growing body of research into forms of play encountered in classrooms in various parts of the world. In our case the teacher, as director of talk, seemed to have a crucial role in bootstrapping a variety of playful discursive practices. Some forms of play witnessed in this lesson are cultural play frames in and by themselves such as children’s games or collective verbal contests. Others emerge in the course of more narrowly institutional episodes, for instance turns that realize the consecutive moves of an Initiation-Response-Follow-up (IRF) sequence (e.g. Cazden 1988; Wells 1993) or are inserted in a formulaic sequence (cf. Bell 2012b). Instances of play also surface in off-record asides, both on the part of the teacher and the students, which are sneaked in at the interstices between more substantial classroom turns. The latter often contextualize what happens at the level of the pedagogical agenda in a different discourse world with different actors and/or voices (Bakhtin 1981, 1986; Goffman 1979, 1981; Volosinov 1973). Meta-communication in L1-Dutch, dialogic and hybrid forms of communication (cf. Kamberelis 2001), heteroglossia and (re)voicing practices (cf. Maybin 2008) turn out to be important features of this teacher’s classroom register that, we suggest, may trigger corresponding footing changes (Goffman 1979) or complex speaker/listener roles on the part of the learners. In order to identify instances of play we attended to verbal contributions that invited (shared) laughter, chuckles or smiles but also to interactional detail: timing, changes in tone of voice, speech rhythm or volume; shifts in bodily orientation or gaze direction; facial expressions and other nonverbal features of talk that guide the coordination and interpretation of participants’ behaviors in classroom multiparty settings. Tentatively – since our main focus is on a detailed description of instances of play in the course of this particular first lesson event – the analyses in this chapter also raise the issue of longitudinal dimensions of the role of play in foreign language classes. In the final section we present some brief observations and field



The first English (EFL) lesson 

 247

notes of what happened in the same class later in the year. To what extent can we capture patterns in the interactional behaviors in which the learners and the teacher mutually, repeatedly, engage each other? Do they interconnect with strategic, pedagogical choices, e.g. the use of L1-Dutch as a meta-language, collective output frames or other interactional formats that maximize learner practice, while minimizing anxiety and fear of loss of face?

2 Theoretical preliminaries Classroom research aimed at elucidating the discursive practices involved in teaching and learning another language in a multiparty institutional setting inevitably raises the issue of interpretive context. Ideally, we should be able to interpret what we hear and see (‘the data’) in any particular lesson event from a longitudinal insider perspective, i.e. in terms of a familiarity with the shared discourse history of participants which constrains, enables and colors developments over time. Such an emic, longitudinal perspective, however, is hard to get. What we do know, however, both intuitively, and from research in the field of applied linguistics, ethnography and complexity theory, is that initial conditions are important (e.g. Agar 2004; Finch 2010; Larsen-Freeman and Cameron 2008; Seedhouse 2010). Clearly the first lesson contextualizes all following lesson events and sets up expectations for what is to come. Zooming in on the very beginning of what will be the shared discourse history of a teacher and a particular group of learners may therefore enable us to identify specific features or patterns in the co-construction of a learning culture that informs modes of participation, learning and teaching.  An investigation into the everyday practices of an emergent learning community requires that we focus not only on discrete features of those practices but also on the interrelatedness between them. The scrutiny of any discrete parameter, however valuable in itself, is unlikely to shed light on the overall institutional, discursive and ideological matrices that insiders orient to when interpreting each other’s behaviors or planning an appropriate next move. We therefore opted for a holistic, micro-ethnographic approach: there was no clear plan for what we were looking for or should be looking at specifically – other than that we were interested in the moment-by-moment interactional details of participants’ behaviors that together constitute ‘the first lesson’. What stood out after detailed analyses of the primary classroom data was the predominance of forms of play, both planned and unplanned, both initiated by the teacher and by learners. In this chapter we will look at instances of (language) play and trace some of the interactional conditions that may have fostered them in the course of the

248 

 Jet Van Dam and Anne Bannink

lesson. Not surprisingly, multimodal dimensions of interactional behaviours (e.g. Mondada 2009, Norris 2004) like winks and laughter, prosodic changes including whisper voice (Hellermann 2003, 2005; Skidmore and Murakami 2010), changes in gaze direction, bodily orientation and facial expressions, turned out to be pivotal cues (cf. Kamberelis 2001) in identifying contextual changes or complexities that were oriented to by participants as instances of play. In its emphasis on fine-grained observation, interactional detail, and insider relevance (interview metadata) this study builds on seminal studies in the field of educational ethnography (e.g. Erickson and Schultz 1981; McDermott 1988; Mehan 1979, 1980, 1998; Varenne and McDermott 1998) as well as sociocultural approaches to language acquisition and language socialization (e.g. Heath 1983; Kramsch 1993, 2002; Lantolf 2000; Leather and Van Dam 2003a, 2003b; Ochs and Schieffelin 1983; Van Lier 1988). With respect to the role of play in classroom discourse and school-based L2 learning we draw on e.g. Bell 2012a, 2012b; Cekaite 2006; Cekaite and Aronsson 2004, 2005; Cook 2000; Lytra 2007, 2008; Pomerantz and Bell 2007, 2011; Rampton 1999; Sullivan 2000; Van Dam 2002, 2003; Waring 2012.  A distinctive feature of our analytic approach is that we aim to articulate the way in which “linear and nonlinear processes and procedures interrelate in spoken interaction” (Seedhouse 2010: 21). Contexts of talk in institutional multiparty settings are dynamic and emergent: i.e. constructed ‘online’, in discourses-in-progress, by the coordinated behaviors of participants. For a formal account of the way contexts of talk can be changed, updated, invalidated, (re)-embedded and stacked in the course of an interaction, we are indebted to the Dynamic Discourse Model first elaborated in Polanyi (1988) and Polanyi and Scha (1983). This model has been applied to classroom multiparty data: the ‘classroom machine’ or ‘pushdown stack automaton’ metaphor (Van Dam van Isselt 1993: 38–48; cf. Bannink and Van Dam 2006). It aims to establish a complex, technical notion of ‘current state of talk’ that mirrors the processing strategies or interpretive procedures that participants use to figure out ‘where we are’ in the discourse. It works incrementally on a move-by-move basis. Since a detailed description of the model is not currently available in print, we will briefly summarize its main features. The initial state of a classroom event is determined by the non-linguistic context of the interaction; it has to contain enough information to resolve problems of deixis and reference of the first incoming unit. The school bell (‘initial symbol’) calls the procedure for speech events of the type ‘lesson’. It generates a set of specific institutional conventions, procedural rules and constraints that guide participants’ behaviors and that partly overrule unmarked cultural (egalitarian) conventions for the coordination of utterances and interactional behaviors. The procedure is put ‘on the stack’ and from there on contextualizes



The first English (EFL) lesson 

 249

everything that happens in the classroom situation. For the first incoming move or event each participant (in this metaphor: ‘machine that reads input’) has to decide what type of unit (e.g. a monologue; an IRF, a story type unit, etc.) is currently on top of the stack, i.e. contextualizes the coordination of utterances and events. That unit generates a specific set of procedural rules and constraints for what would be a relevant next move (e.g. the answer to a class question in an IRF-in-progress; the next event in a story). A particular move may have to be interpreted in a different set of contextual parameters and values, e.g. an interruption or non-lesson event (PUSH); a return to a previously interrupted unit (POP). Verbal, nonverbal, paralinguistic and prosodic features of interactional behaviors function as discourse markers/operators that may change the parameter settings for the interpretation of an incoming next event (POP and PUSH markers). The model thus provides for complexity and allows the recursive embedding, nesting and stacking of discourse units and subunits. Sets of conventions that have been overruled at the start of the lesson (e.g. egalitarian conversational ones for instance) are saved and are still accessible at embedding levels of the interaction. They can be shifted back to, causing non-linear transitions from state to state. More than one transition from state to state is possible, which provides an analytical base for identifying and describing misunderstandings, jokes and parallel states of talk as well as classroom behaviors that are selectively ignored. An example: in our data a wasp flies into the classroom and the teacher interrupts her explanation about the notion of grammar. The ‘lesson’ context is temporarily overruled: wasps are not ratified participants in the institutional event ‘lesson’. There is a procedural vacuum and unmarked cultural conventions for the coordination of interactional behaviors are oriented back to (POP/PUSH; embedded discourse subdomain). Students get out of their seats to chase the wasp; classroom order is in jeopardy. But  institutional roles and tasks are still oriented to as the embedding ‘higher-order’ parameter setting that guides participants’ interactional behaviors. This allows the teacher to tell everybody to get back in their seats: she appoints ONE wasp killer to do the job. When the embedded sub-unit is over (i.e. when the wasp has been killed), the discourse POPs back to the embedding ‘lesson as task’ institutional event. A cluster of markers (verbal, non-verbal, prosodic) on the part of the teacher signals the return to the earlier interrupted state of talk at the level of the pedagogical agenda: “[raises voice; positions herself in front of the class; directs gaze back to class] WE:: uh had got to the point where we were discussing GRAMMAR ..”. In the analyses in this chapter we hope to show that the notion of emerging discourse complexity the model articulates has explanatory power with respect to instances of play themselves as well as the ambiguities and misunderstandings that may arise in multiparty discourses-in-progress as a result of them.

250 

 Jet Van Dam and Anne Bannink

3 About the data 3.1 Data selection The lesson we analyze in this chapter originates from a corpus of very first ­language lessons (mixed audio/video; N = 17). It has been selected since it prominently features laughter and various forms of by-play and side-play (Goffman 1979) that contextualize (and are contextualized by) what happens at the level of the lesson agenda. An additional reason for selection is that we observed two other lessons involving the same teacher and the same group of students later in the year. This allows us to venture some tentative remarks about longitudinal aspects of the interrelatedness of language play and L2 learning in this particular setting and trajectory at the end of the chapter. Participating in the 50-minute lesson were twenty-six 12- to 13-year-olds on their very first day of attending secondary school. The students, boys and girls, came from varying social backgrounds. Most of them were native speakers of Dutch; two were second-generation immigrants who were fluent in Dutch. The learners did not know each other since they came from different primary schools in the central Amsterdam region, some of which offered elementary English on a voluntary basis. The group as a whole thus started from scratch and had no shared history or specific (epistemic, school-based) knowledge to fall back on. However, in a globally oriented country like the Netherlands, some informal exposure to English words and phrases can be taken for granted for all adolescents that age. Like many other first lessons in our corpus the lesson under scrutiny here is of the teacher-fronted variety. It was taught on the basis of a simple agenda (see 3.2) that had been negotiated with a colleague who was to teach the same lesson on the same day to a parallel group of learners – but did so in English only (‘target language strategies’; cf. Auerbach 1993). Interestingly the two teachers had not discussed what language the lesson was to be in. The teacher whose lesson we analyze here chose to conduct it mostly in L1-Dutch (‘bilingual teaching’; Hall and Cook 2012) except, of course, in narrowly task-oriented activities and episodes that constitute the pedagogical agenda. Spontaneous utterances in L2-English were volunteered by individual students also during episodes where there was no obligation to speak in the target language.

3.2 Overview of the lesson After welcoming the students and settling issues about who is who and who sits where (the students had been given a seating plan in advance) the teacher uses



The first English (EFL) lesson 

 251

the first twenty-odd minutes of the lesson to explain routine lesson procedures and to introduce the course materials in Dutch. While doing so she continually checks whether everybody understands and urges the students to speak up if they do not (see section 5.1). She also reassures them about the different levels of proficiency in English that are bound to exist among them: some learners have had some EFL lessons at primary school; some have not. These differences will be accommodated soon, she assures them. Then a wasp flies into the classroom (see section 2 and 5.2). When this emergency has been dealt with, the teacher announces the ‛real’ beginning of the lesson: she will start by rehearsing some items and routines in English that they are probably already familiar with. Volunteers recite the English numbers and the whole class joins in, saying them up to a hundred. Then the names of the letters of the English alphabet are introduced and repeated in chorus (see section 5.3), the teacher enthusiastically taking the lead. In the third round (“een mooi VOL koor nu alsjeblieft”/“let me hear a nice FULL choir now please”) nearly everyone participates. The class continues with a spelling activity: the teacher spells a word from the syllabus and the students write it down in their notebooks. Without any in­troduction the teacher then suddenly, at great speed and in a clear voice, begins to chant a rhythmic tongue twister that contains some letters that are difficult to spell for Dutch learners (section 5.4). Some students join in spontaneously and together they practice saying “Ottiwell Wood”. This becomes a kind of competition: they all join in and the teacher challenges individual students to try and emulate her performance – which is reframed as homework. She then proceeds to introduce the last topic on the lesson agenda: ritual phrases associated with greetings and saying one’s name in English (section 4). Chunks containing the personal and possessive pronouns in English are practiced in fast collective metadialogues, e.g. ‘if I say [you are called X] I can also say [your name is X]; if you want to say [L1 phrase] you can say [L2 phrase] but you could also say [equivalent L2 phrase]’. She is very strict in conducting this fast classroom catechism: the students should ‘recognize contexts for utterances’ (Erickson and Schultz 1981) and produce the correct formulaic phrase in the immediately adjacent slot. The lesson is supposed to end with an English song but the technology is not working properly and the sound quality is so bad that she decides against it. Instead they round off the lesson with a variation on the classic “I spy with my little eye” game (see section 5.5).

3.3 A note on the transcriptions Though fully aware of the explanatory value of detailed transcriptions as used in conversation-analytic studies of interactional data, we have chosen to include

252 

 Jet Van Dam and Anne Bannink

here only those elements in the production of talk that (a part of) the participants demonstrably orient to. Transcripts of multiparty interactions are almost inevitably selective in what they do or do not record (even if musical score type transcription methods are used). For the sake of readability we have just included (marked) pauses, changes in body movement, gaze direction or bodily orientation; prosodic changes like whisper voice, changes in rhythm, pitch, intonation or volume; laughter, giggling and other appreciative or evaluative noises and facial expressions that signal amusement or embarrassment. Pauses are not indicated in seconds but should be read as context-sensitive: an unmarked pause in the institutional register is longer than one in most conversational registers (cf. ‘gap’; McHoul 1978); a conversational aside following  an institutional turn is marked as immediately adjacent (see e.g. data 1C).

4 Playing classroom games: discourse complexity In this section we describe instances of spontaneous play on classroom and everyday names and identities that occur quite late in the lesson, in the pre-final lesson episode. They are initiated by three individual students in the context of a series of conventional IRF sequences. The pedagogical focus is on learning to say formulaic utterances in the L2 (cf. Bell 2012b). Prompted by the teacher elicitation frame “what’s your name?” the learners are to say their names in the appropriate lesson format. But in order to clarify the discourse issues involved and contextualize what happens later on in this episode, we briefly discuss a  meta-comment in L1-Dutch with which the teacher prefaces the first class question now on the lesson agenda: Data 1A Discourse complexity T NOU – ik vind het een beetje RAAR om te vragen ‘what’s MY name’ want dat weet je meestal wel – maar ik kan bijvoorbeeld WEL zeggen [keert zich naar SF] –  what’s ↑YOUR name T WELL – I think it is a bit ODD to ask ‘what’s MY name’ because one normally knows that – but what I CAN say for instance [turns to SF] is – what’s ↑YOUR name Why did the teacher, before making the first class question of this new episode operational, insert a meta-comment on a question she decides not to ask? She publicly shares her concern with the students that starting with the first



The first English (EFL) lesson 

 253

person singular of the grammatical paradigm, i.e. asking the question “what’s MY name”, would be odd or infelicitous “since one normally knows that”. That remark clearly refers to pragmatic constraints (cf. Austin 1962) on asking questions to which the questioner already knows the answer that generally apply in informal conversational contexts of talk. In task-oriented settings like the (language) classroom, however, asking such display questions is quite a normal procedure: a widely accepted template to implement learning. It provides the teacher with opportunities to act on individual students’ performances and can be a tool to orchestrate the learning of many (Cazden 1988; Hall 1997; Hellerman 2003, 2005; Lee 2007; Nassaji and Wells 2000; Seedhouse 2010; Waring 2009; Wells 1993). The fact that the teacher mentions these constraints does not mean that they are also currently in use (Levinson 1983; Sperber and Wilson 1981). For her comment to make sense, however, we must assume that they are still oriented to at embedded levels of the institutional interaction as a set of normative judgments about the appropriateness of a particular classroom move. Apparently different interactional systems (‘interaction grammars’; ‘language games’; cf. Wittgenstein 1953) can be oriented to simultaneously. Such an observation is to the point if there is empirical evidence that this double orientation, and the potential for discourse complexity it implies, is meaningful for insiders: consequential for what they do, say and understand in language lessons and classroom situations. We will revisit these observations at the end of this section when, it turns out, the teacher decides to ask this question anyway.

4.1 Learner-initiated play We now turn to the class question that is actually on the floor: ‘what’s YOUR name?’ It is the first move in an evolving IRF sequence that solicits a student answer in next slot:

Data 1B Learner-initiated play 1 T NOU – ik vind het een beetje RAAR om te vragen ‘what’s MY name’ want dat weet je meestal wel – maar ik kan bijvoorbeeld WEL zeggen [keert zich naar SF] – what’s YOUR↑ name SF→ MY name is  – uh FREDERIC [luid en duidelijk; overdreven Brits-Engels accent en intonatie]

254 

 Jet Van Dam and Anne Bannink

T WELL – I think it is a bit ODD to ask ‘what’s MY name’ because one normally knows that – but what I CAN say for instance [turns to SF] is – what’s YOUR↑ name SF→ MY name is – uh FREDERIC [loud, clear voice; emphatic hypercorrect ­British-English prosody; low fall]  The student who is allocated the first turn is called ‘Freek’ (pronounced / freɪk/in L1-Dutch) which is short for the rather formal Dutch name ‘Frederik’. His answer displays a “normative orientation to the pedagogical focus” (Seedhouse 2010: 12) of the current interaction. Without being prompted to do so by the teacher he spontaneously provides his name in the full-sentence classroom format – rather than just saying his Dutch name, which is the unmarked form of providing this information outside the classroom. But that is not all. On the spot he constructs a  new L2 name and identity for himself that is a parody on authentic ‘typical’ British-English speakers. In a clear voice he animates a posh British version of himself with exaggerated intonation contours  – a persona that he speaks for rather than the self (cf. Goffman 1979). This is both a correct lesson answer and a classroom act that plays upon informal and classroom conventions for disclosing one’s name or identity (cf. Waring 2012). In principle the teacher’s third-place follow-up or evaluation move is now due. But in the gap or transitional space between consecutive institutional turns at talk (cf. McHoul 1978) the other students show their appreciation of Frederic’s performance through subdued giggling. Almost simultaneously there is an unsolicited contribution by an anonymous student in a dramatic whisper voice that is picked up by the microphone: Data 1C Learner-initiated play 2 SF [loud, clear voice; emphatic hypercorrect British-English prosody; low fall] MY name is – uh FREDERIC= =[soft giggles] = SSS → Sx  →  = mis-ter X [English phonology; slow, sotto voce; clearly audible dramatic whisper]= = [amused looks; chuckles] SSS In the language classroom overhearer backchannels like chuckles and subdued laughter, amused looks and whispered comments are the norm rather than the exception. These unsolicited spontaneous student contributions “do[es] not



The first English (EFL) lesson 

 255

appear to be intended as disruptive but rather as a playful interlude alongside the main classroom business” (Lytra 2007: 169). Note that the play on frames implied in the anonymous student’s whispered, conspiratorial “Mr. X” addresses discourse complexities and potential dilemmas that are inherent in formulaic sequences in the L2 classroom (cf. Bell 2012b). This episode is not about the exchange of information: it does not really matter what one’s name is. What is at issue is the performance of a classroom task: displaying that one can correctly say an L2-English sentence frame. The collusive ‘Mr.X’, while violating that constraint, is an instance of creative play at multiple levels of the interaction. It contains the English word ‘mister’ that unambiguously indexes it as a target language utterance. It successfully evades the pragmatic incongruity associated with having to insert one’s L1-Dutch name in an L2-English sentence frame (for evidence that this can indeed be a dilemma for learners, see section 4.2). An illegitimate contribution to the discourse, it topicalizes the ritual nature of the current interaction – as well as, possibly, expressing resistance to it (cf. Bell 2012b; Pomerantz and Bell 2007, 2011; Rampton 2006). Why reveal one’s real name in public when in fact any name will do? In the meantime the slot for the institutional third-place teacher evaluation move in the IRF-in-progress is still open, and a filler for that slot is now sequentially due: Data 1D Collaborative play SF [loud, clear voice; emphatic hypercorrect British-English prosody; low fall] MY name is – uh FREDERIC= =[soft giggles]= SSS Sx =mis-ter X [English phonology; slow, sotto voce; clearly audible dramatic stage-whisper]=  =[amused looks; chuckles] SSS T → [turns away; imitates, recitation mode] ‘MY name is – FREderic’ – [dreamily; ‘middle distance’ look] MOO:I hè↑ ((BEAU:tiful, isn’t it↑)) – [echoes again] ‘MY name is FREderic’ [turns towards a different student] The teacher echoes Freek’s answer in recitation mode. At first sight this looks like a purely conventional third-place teacher follow-up move that relays what one student has said for all to hear as a signal that the answer is correct. But that is the case only if we ignore the cluster of changes in bodily orientation, gaze direction and tone of voice that accompany its production, as well as the emphatic evaluative comment in L1-Dutch that immediately follows it.

256 

 Jet Van Dam and Anne Bannink

While echoing Freek’s words the teacher physically orients away from him and also avoids eye contact with any of the other students in class. She adopts a middle-distance look and a soft dreamy tone of voice quite unlike her normal clear ‘classroom’ voice. All of these features, in combination with her brief switch to L1-Dutch, suggest that what we witness here is a brief interlude of self-talk (Goffman 1981; Hall and Smotrova 2013; Skidmore and Murakami 2010) that she collusively shares with the students and that is meant to be overheard – rather than an institutional next move that is addressed to them. This footing change (Goffman 1979) in the middle of an IRF-in-progress is a nonlinear move that briefly embeds a more informal conversational stance and speaking voice within her sustained institutional one (‘PUSH’; see section 2). A personal expression of delight in the aesthetic and performative elements in Freek’s play upon L1 and L2 names and identities (“MOO:I hè”/ “BEAU:tiful, isn’t it”) delays the institutional follow-up moves in which the full-sentence format (“my name is Frederic”) is relayed and modelled for the whole class. The Dutch verbal tag “hè” (approximately English: “isn’t it”) suggests that her appreciation of Frederic’s classroom performance is shared by all co-present. While remaining ‘the teacher’ evaluating a student answer, here is a glimpse of ‘me-as-a-person’ showing my appreciation of a classroom performance. It is only when the teacher re-engages the class with her gaze, repeats Freek’s answer in her ‘normal’ classroom voice and allocates the turn to a different student that we are firmly back on institutional grounds. The complexity we have unraveled here is consistent with what many authors have observed with respect to the third-place teacher F-move. Sinclair and Coulthard (1975) first referred to the many different functions it may have and tasks it may perform. Lee (2007: 1205) notes that “what teachers do in the third turn position is not predictable, in principle, because its character is contingent upon the prior turns”. Indeed – the teacher’s footing shifts are responsive to the creative and playful stances that Freek and the anonymous student who whispered “Mr. X” have initiated. The authenticity on display here involves recognizing affordances for play in the moment-by-moment unfolding of ritual task-oriented discourses. Within a single speaking-slot-in-progress formal and informal speaking roles can briefly be shifted into or out of – or stacked. Embedding a theatrical stance or speaking voice as an overlay over institutional roles may simultaneously display an orientation on task – and one away from overly rigid institutional parameters. We have seen how both the teacher and the students exploit the interactional niches of classroom floors to realize their own agendas (cf. Mehan 1980). They literally collude or ‘play together’ and, mutually sensitive to what is going on, combine business and fun. But emergent frame complexity as a result of playful moves may also cause participants to be confounded about what is the current



The first English (EFL) lesson 

 257

state of talk: where are we? what is serious, i.e. task-related, and what can be ignored? The status of collusive contributions and instances of play – if not explicitly taken up by the teacher – may be unclear: do they update the current state of talk? Sometimes the rules of a particular classroom game and the notion of what is a legitimate next move can become ambiguous or opaque – as we will see in the next section.

4.2 Learner-initiated play 3 In the preceding section we saw how the first student addressed in the ritual ‘what’s your name’ sequence introduced himself by creating a locally relevant ‘English lesson’ persona for himself. His playful move was greatly appreciated both by the teacher and the overhearing peers. The teacher then proceeds by asking a different student the same question. His answer is marked by multiple signs of trouble:

Data 1E Attempt at play 1 T and uh - [turns to Hans] what’s YOUR name↑ [looks worried, shrugs shoulders; blushes, pulls face] ja - uh God SH →  [L1 phonology] - u::h - nou ((yes - uh - my God - u::h - well)) - [grunts] [‘low Dutch’ pronunciation] HANS= =[chuckles; soft laughter] Class A first move in a sequence discursively sets up expectations for what is to come (Schegloff 1968; Seedhouse 2010). Hans’ evident embarrassment and delay in producing an answer can only be explained with reference to what happened in the immediately preceding first IRF of the episode. Taking his cue both from Freek’s response itself and from the teacher’s emphatic praise of it, Hans must have assumed that inventing an L2 name or identity for oneself was an integral part of what he was expected to do in next turn. Just saying his Dutch name ‘Hans’ – which has no obvious English-sounding equivalent – in an L2 formulaic sentence frame apparently would not do. His body language (worried look; shrugging of shoulders), pauses, verbal hesitation markers and response cries (Goffman 1981) all testify to his predicament. After considerable delay he finally produces his name with a marked fall in tone and ‘lowDutch’ phonology. Hans’ exaggerated performance of defeat is in itself a form of play. It gives rise to yet another discreet round of laughter on the part of his peers, clearly of

258 

 Jet Van Dam and Anne Bannink

the affiliative kind: by over-acting Hans has succeeded in saving his face vis-à-vis his peers – has maybe even enhanced it. The teacher, uncharacteristically, has not recognized his failed attempt at playing along and responds with a public, unmitigated other-correction – the only one in the entire lesson:

Data 1F Attempt at play 2 T and uh - [turns to Hans] what’s YOUR name↑ SH ­[looks worried, shrugs shoulders; blushes, pulls face] ja - uh God [L1 ­phonology] u::h - nou ((yes- uh - my God - u::h - well)) - [grunts]uh ['low Dutch' pronunciation] - HANS= Class  =[chuckles; soft laughter] T → I don’t want to hear ‘HANS’ – I want to hear ‘my NAME is Hans’ SH → [high-fall ‘relief tone’] OO:H – [echoes] my name is Hans Hans’ answer is classified as an error since it does not exhibit the formulaic full-sentence format. Apparently the teacher was not aware of Hans’ structural reading of the situation: that her emphatic praise of Freek’s playful answer could be taken as updating discourse expectations with respect to what he was supposed to do in next turn. She just models the full sentence lesson format in which his L1-based Dutch name is apparently unproblematic. Hans obediently echoes her correction, his update-marker “OO:H” (Heritage 1984) – that we gloss as ‘is that all?!’ – expressing both surprise and relief. Clearly he had set himself a different, more difficult task: that of playing along with his peers. Discussion. The data and analyses above illustrate the ambiguous status of instances of play in the margins of the pedagogical agenda with respect to what counts as data: what relevantly happens in class. What we have described is emergent discourse complexity: evidence of the different readings that participants may bring to bear on multiparty classroom situations. Blurted vocalizations, instances of play and response cries are “creatures of social situations, not states of talk” (Goffman 1981: 121) and may or may not be ignored by next speaker. According to Markee and Seo (2009) misunderstandings are most likely to occur when participants switch from one speech exchange system (e.g. a formal register associated with pedagogical activity) to another (e.g. informal conversation). Our data show that this is a fortiori the case when two or more interactional templates are simultaneously in force (cf. ‘hybridity’; Kamberelis 2001). The rich answer format that Hans apparently wished to emulate was introduced as an element of play overlaid upon an otherwise formulaic institutional sequence – and it was introduced by a fellow student, not by the teacher herself. Apparently she was not aware of the possibility that a nonlinear reading of the



The first English (EFL) lesson 

 259

situation might include the play upon L1/L2 names and identities as setting the norm for what the following speaker should do in next turn. The double bind (Bateson 1972; Watzlawick et al. 1967) such complex discourse parameters may create for subsequent speakers is, however, recognized by overhearing peers – as may be inferred from the soft chuckles and affiliative laughter in class. Discourse complexity revisited. Later on in the same lesson episode the teacher solicited other names from individual students with ‘what’s his/her name’ prompts. This resulted in utter confusion owing to a different kind of discourse complexity: lack of shared real-world knowledge. The learners did not yet know each other’s names. They came up with code-switched responses like: “her name is – uh dat weet ik niet” (“her name is – uh I don’t know”). Also, the teacher positively evaluated a learner answer which, it turned out, contained the wrong embedded information: one student had been referred to by the name of his peer. There was great hilarity and noise: orientation on task was in jeopardy. Raising her voice the teacher resorted to the very question she had overruled as ‘odd’ or marked (‘raar’ in L1-Dutch) at the beginning of the episode (see data 4.1) “because one normally knows that”. It is precisely the type of question she now needs to bootstrap ‘common ground’ or shared knowledge in the newly-created learning community. Even in a first lesson with a new group of learners, the name of the teacher is an item the learners all know – if only from their lesson schedule. In a loud voice, emphatically, without allocating the turn to anyone in particular, she asks: “AND – what’s MY name?” In response the students all shout in chorus: “YOUR name is Mrs. Van Polanen”. Class order is restored – the lesson can go on. The teacher now introduces fixed referents for each of the personal pronouns and practices L2 formulaic phrases in fast all-class meta-dialogues, varying the prompts. In conclusion. In this section we zoomed in on instances of spontaneous play on the part of individual learners – and the misunderstandings they may give rise to as a result of online-created, emerging discourse complexity. We also showed that what happened in the first subunit in an IRF sequence-in-progress generated normative expectations for what was to come. We will now go back to the very beginning of the lesson and examine to what extent we can identify playful moves and stances that, together, constitute patterns of participation modes (cf. Agar 2004: 16) which feed into the joint construction of a ludic classroom culture.

5 Dialogism or the discursive emergence of a culture of play In the previous section we described a variety of sophisticated playful moves initiated by individual learners. We may well wonder at these students who seem

260 

 Jet Van Dam and Anne Bannink

so creative and uninhibited in their very first English lesson in a new school and a new peer group. Could this be a particularly gifted or sophisticated and cooperative group of learners? It may be so. But there is also a different angle. The data analyzed above occur approximately ten minutes before the end of the lesson. In the preceding forty minutes the teacher has consistently embedded a range of playful or dramatic speaking roles and voices both for herself and for co-present or virtual students (cf. ‘double-voicing’; Bakhtin 1981, 1986; ‘revoicing’; Maybin 2008). She has also involved the learners repeatedly in collective output frames like chanting or chorusing (cf. Cook 2000; Huizinga 1949; Heath 1983) that she leads with great gusto. Has she somehow managed to engage the students in an emerging culture of play (cf. Sullivan 2000)? To make a case for that possibility we will go back to the very beginning of the lesson. We will systematically trace how affordances for playful participation modes and speaker-hearer roles are created and sustained throughout the lesson (cf. also Bannink and Van Dam 2013).

5.1 Dialogic play and complex speaking roles: animating a student voice In a first lesson it is essential that students feel at ease and understand what is going on, that common ground is established. Everything is new and the learners have to figure out what are practical routines and accepted practices. There is the risk of information overload – and in a new class it is especially daunting for learners to own up to not-understanding in the public domain (cf. “Kids spend most of their time in class trying not to be caught not-knowing something”; Varenne and McDermott 1998). Since teachers cannot look into individual learners’ heads, they are dependent upon the information that is conveyed through their students’ words, body language and interactional behaviors. The sooner this process of mutual monitoring (‘feedback system’) is in working order, the better. From the very first moment in class, when the teacher greets and welcomes the students, she adopts a light conversational tone and playful manner with little jokes and side remarks when answering student questions. She reassures one student who worries about not yet having received his English books by playfully exaggerating the alleged offense: “maar dat is een RAMP”/“but that’s a ­disASTER”, accompanied by a wink and a grin, before embarking on a detailed helpful explanation about where copies of the book are probably available. She also puts the students at ease about the differences in prior knowledge or ­exposure to English that are bound to exist among them: these are sure to be negotiated in time and should not be a problem – provided they all speak up as



The first English (EFL) lesson 

 261

soon as there is anything they do not understand. She borrows a virtual student’s voice to bring that point home: Data 2 T is dat duidelijk voor iedereen↑ - [kijkt rond] - eerlijk VRAGEN hoor als je iets niet begrijpt- want ik sta hier ook maar een heel VERHAAL te vertellen maar als ik iets onduidelijk ZEG - moet je het even VRAGEN hè dat je niet naar huis gaat met het idee van – [andere stem, sneller] NOU –wat dat mens heeft staan uitkramen mag JOOST weten – maar IK begrijp het in elk geval NIET = Cl =[gegrinnik, zacht gelach] T [normale stem] VRAAG het alsjeblieft – ik ben DOL op kinderen die vragen stellen T   is that clear to everyone↑ - [looks around] – DO ASK if there is anything you don’t understand - ‘coz this is just me telling you a long STORY but if I am not SAYING things clearly - JUST ASK - so you don’t go home thinking – [different tone of voice; faster] GOODNESS KNOWS what that woman has been gibbering about but I really haven’t a CLUE= CL       =[grins; soft laughter] T [back to previous voice] PLEASE ASK - I just LOVE children who ask questions This teacher monologue occurs some five minutes after the beginning of the lesson. A sudden change in tone of voice, pitch, speech rhythm and pace signals a change in footing (Goffman 1979): the teacher appropriates the voice of a virtual student who is completely at a loss about what on earth could be going on in class – and is being very open and vocal about it. He or she (there is no hint as to gender) is presented as thinking aloud. In the embedded imaginary discourse world the teacher herself is framed as the culprit: she is to blame because she is the one who has been gibbering all this nonsense. Note that the teacher refers to herself in slightly derogatory terms (“dat mens”/“that woman”) of the kind that students may use amongst themselves during breaks in order to impress their peers by feigned or real toughness and which may raise their status in the peer group (cf. Rampton 2006). The teacher’s dramatic performance provokes appreciative chuckles and soft laughter in class. What is the difference between saying in class: “Please tell me when you don’t understand” or “Just ask me if anything is unclear” and adopting a virtual student’s voice (‘say-foring’ or ‘speaking for’; Goffman 1981) in order to convey the ‘same’ message? The point is, of course, that it is not the same message, precisely because there is this element of play and juggling of

262 

 Jet Van Dam and Anne Bannink

identities. Embedding the speaking (or thinking) voice of a different persona in the course of one’s monologue is a quite normal feature of everyday talk-in-interaction (cf. Bakhtin 1981, 1986; Kamberelis 2001; Maybin 2008; Volosinov 1973). That is precisely why this footing change (Goffman 1979, 1981) on the part of the teacher, early in the discourse history of this class, is worth noting. Her dialogic impersonation of a grumpy non-understanding student semiotically models what they will all – crucially – have to learn how to do: find a new speaking voice, a new identity for themselves in the L2; re-invent themselves in order to acquire a new life in the new language (cf.  Kramsch 2009; Leather and Van Dam 2003a, 2003b). We are reminded of Goffman’s observations about initial speaking roles in early L1 acquisition. Children learn to speak, he says, by having parents and educators speak for them rather than being addressed as conversational partners: “so even as the child learns to speak it learns to speak for, to speak in the name of figures who […] are not yet the self” (1981: 151; note). The ‘laminative features’ of these interactions, he adds, are anything but childlike. Heteroglossia, double voicing and complex speaking roles are in evidence throughout this lesson.

5.2 Interlude. Playing with cultural frames 1: Wasp hunt In the course of her explanations about the course book, classroom routines and lesson materials the teacher asks the students whether they know what ‘grammar’ is. A wasp is circling around her head. One of the students suggests “taalgebruik” /“language use” and the teacher enthusiastically responds while at the same time trying to evade the wasp: Data 3 T Ja dat is HEEL knap gezegd - taalgebruik - hoe ZEGGEN ze het - WIJ zeggen bijvoorbeeld uh u:h ik HEET zo-en-zo - zij zeggen niet ik HEET – ZIJ zeggen­mijn NAAM is -  dat is een - [wesp vliegt langs, lerares bukt, stapt opzij] [roept] OO:::H – [luid] ANDERE MANIER [kijkt angstig naar wesp]= Class     =[gelach] T  [herstelt zich, normale toon] op een - ANDERE manier zeggen ze het - en dat HELE systeem van regels waar je als buitenlander een hulp­middel aan hebt om te leren hoe DOEN zij dat nou [wesp vliegt om haar hoofd]=    =daar heb je ‘m weer Sx



The first English (EFL) lesson 

 263

T [kijkt naar wesp, boos, luide stem, nadrukkelijk] NOU ZEG GA NAAR HUIS JOH T Yes that’s VERY well said - language use - how do they SAY it - WE say for instance uh - u:h ‘I am called so-and-so’ - they don’t say ‘I am called’ they say ‘my name is’ - that’s a - [wasp flies along; teacher ducks away, steps aside] [cries] OO:::H – [loud] DIFFERENT WAY [looks at wasp ­anxiously]= Class   =[laughter] T they say it in a DIFFERENT way - [regains herself; normal tone] and the WHOLE of that system of rules that is a help to you as a foreigner to learn how do they DO it [wasp flies around teacher’s head]= Sx   = there it is again T [looks at wasp; angrily, raises voice emphatically] HEY COME ON - GO HOME YOU Some students get up and start chasing the wasp across the classroom. Others join in – classroom order now really seems to be in jeopardy. The teacher, however, solves the problem by exploiting her institutional right to allocate turns and tasks while at the same time importing a new cultural scenario that is laminated onto the current classroom one. A dramatically-intoned “we hebben EEN moordenaar onder ons”/“we have ONE murderer among us” sets the stage for a collective wasp hunt. Note that the embedded play template is itself used as a resource to re-establish classroom order. Roles and scenarios that obtain within the story world warrant the teacher’s decision to appoint one contract killer: the two frames are collapsed. The other students are assigned the temporary role of supporters. Data 4 SSS [hilariteit; staan op het punt om op te staan] T HIER - wacht even tot-ie zit - [verheft stem] OK - ALLEMAAL BLIJVEN ZITTEN - we hebben EEN moordenaar onder ons! - wie wil dat op zich nemen↑=      =IK- IK - IKKE SSS [geeft S1 rol van moordenaar] T S1 [staat op, jaagt op wesp] Class [commentaar en adviezen] S1 [vangt wesp, slaat wesp dood] Class [geluid neemt af] T [zachtjes] dankjewel - [richt zich tot de klas; verheft stem] WE:: uh waren bij de GRAMMATICA aangekomen [stem weer zachter, terug naar onderbroken les]

264 

 Jet Van Dam and Anne Bannink

SSS [hilarity; prepare to get up from their seats] T HERE - wait a second till it sits down - OK [raises voice] IN YOUR SEATS EVERYBODY – we have ONE murderer among us - who wants to do the job↑=     =ME ME - I DO SSS T [allocates role of ‘murderer’ to S1] S1 [gets up, chases wasp] Class [comments and advice from the class] [gets the wasp, squashes it] S1 Class [noise dies down] T [low voice] thank you - [directs gaze to the class; slightly raised voice] WE:: uh had got to the point where we were discussing GRAMMAR [back to interrupted lesson topic; lowers voice] The teacher re-establishes order by locally assigning one of the students a new role in a now current cultural scenario or embedded play frame: chasing the enemy. The contract killer is cheered on by all, the teacher included. When the embedded business has been dealt with the teacher employs a cluster of verbal, prosodic and non-verbal contextualization cues (POP-markers; Polanyi 1988; Polanyi and Scha 1983) to effect a return to the embedding task-related state of talk that was interrupted (cf. ‘classroom machine’; section 2).

5.3 Playing with cultural frames 2: chanting and chorusing After the English numbers learning to say the letters of the English alphabet is on the agenda. In the transcript below we have included the transition to this episode because it shows that the teacher anticipates difficulty: the next task, she publicly announces, might well be more challenging for the learners than the previous one. Data 5 T NOU dat gaat allemaal PRIMA jongens - dan kunnen we meteen doorstoten naar een moeilijker onderwerp - - SPELLEN Class (….) T kan iemand dat↑ [kijkt rond] Class (….) [enkele verontruste blikken] ja da’s lastig -- da’s HEEL lastig [trekt vies gezicht]= T SS =[zacht gelach]



The first English (EFL) lesson 

 265

T

ik ga even het alfabet opzeggen - als iemand het weet dat-ie denkt van ‘AH dat weet ik OOK’ - dan val - dan valt-ie bij - OK↑ [kijkt rond] - - IK begin en als je wat WEET dan ga je mee en dan hou je weer ‘s je mond dicht en dan doe je weer mee en dan hou je weer ‘s je mond - al naar gelang je kunt T [per letter ongeveer 1,0 sec.] /ei/ /bi:/ /si:/ /di:/ /i:/ /ef/ /dzji:/ GOED zo [gaat door] Class [SSS in koor met teacher] - / bi:/ /si:/ /di:/ /i:/ /ef/ /dzji:/ - - [etc.] T

WELL that’s going REALLY well folks - so we can push on at once to a more difficult topic - - SPELLING Class (…) T can anyone do that↑ [looks around] Class (….) [some worried looks] T yes that’s tricky - that’s REALLY quite tricky [mock worried look]=       = [soft laughter] SS T I’m going to say the alphabet out loud - if anyone KNOWS it and thinks ‘AH I can do that too’ - then you - then that person joins in - OK↑ [looks around] -- I begin and if you know something you join in and then you shut up again and then you join in again and shut up for a while - just see what you can do T [approx. 1 sec. per letter] /ei/ /bi:/ /si:/ /di:/ /i:/ /ef/ /dzji:/ [fast, low pitch] well done [continues] Class [in synchrony with teacher] - /bi:/ /si:/ /di:/ /i:/ /ef/ /dzji:/ - - [etc.] The teacher first compliments the class on their performance in the previous episode (counting up to a hundred in English). She not only mentions that they have done very well but also makes it consequential for what happens next: constructs it as the reason she can now move on to a more difficult topic: spelling. Progress can be fast (L1-Dutch ‘doorstoten’ in the first line of the transcript has connotations with fast-moving troops in pursuit of strategic targets). The teacher asks for a volunteer to say the alphabet out loud but no hands are raised. Once again she qualifies the task as – indeed – more difficult than the previous one, both verbally (“ja da’s lastig - da’s HEEL lastig”/“yes that’s tricky – that’s REALLY tricky”) and nonverbally, by putting on a facial expression that is a parody on anxiety. The class reacts with laughter. Since it has become clear that none of the students is willing to volunteer, the teacher decides on a different participation format. Instead of soliciting an individual turn at talk she implements an all-class speaking slot. Detailed  instructions are given as to what everyone should do.

266 

 Jet Van Dam and Anne Bannink

What everyone should do, it turns out, is up to themselves completely. A col­ lective speaker floor is constructed that is based on the (conversational) principle of self-selection. Individual students are free to decide whether, when and how to participate, on a local, moment-by-moment basis. The teacher takes the lead. She explicitly spells out the array of (verbal, nonverbal, partial; intermittent) participation modes that this collective output frame allows. For each incoming letter or subunit of the evolving recitation slot-in-progress, each student can either join in and repeat it out loud; repeat it sotto voce  or in a whisper voice (cf. private speech; Lantolf and Appel 1994; Vygotsky 1986); stay mum, listen to and notice (cf. Swain 1995) what their neighbor is saying before joining in (and possibly selfcorrect in same turn); or just move their lips as a sign of ritual participation. In fact each learner may adopt precisely the interactional role that suits his or her current level of competence and confidence, on a local, item-by-item basis (cf. Bannink and Van Dam 2006). Note that there are no individual teacher or peer evaluation slots in this collective format. In the public domain everyone is the speaker. There is collective scaffolding (Donato 1994; Vygotsky 1978), however, in collusive, screened-off interactional peer sub-domains. With a maximum of practice for all, the risk of public exposure or loss of face is minimal. There is no overhearing audience; pronunciation errors are drowned or made anonymous in the chorus. The teacher’s detailed instructions explicitly cue this array of possible participation modes along the axis of speaker-hearer roles. Individual learner behaviors can only be inferred. They are hardly observable or transcribable – even in musical-score-type transcriptions. After the first round the class is collectively praised. But the teacher requests a re-run so that “people who could not do it before” can also join in. Note that interactional roles are updated here – not cognitive tasks: non-speakers are invited to become speakers or at least whisperers. The third time (T.: “een lekker VOL koor graag”/“a nice FULL choir now please”) the volume has indeed increased considerably: nearly everyone has apparently joined in. After each round the teacher briefly provides feedback on the correct pronunciation of some letters that turned out to be particularly troublesome – in the way a conductor would practice some notes that are persistently out of tune. Feedback and treatment of error are postponed until she knows exactly which letters are persistent trouble sources for several students (she would not have heard otherwise). These are efficient teaching strategies since individual errors and errors that disappear spontaneously need not be dealt with. Yet again a different cultural frame is borrowed to organize classroom tasks. Situated learning is reframed as choir practice which provides equal opportunities for all, finely-tuned to their current L2 proficiency and confidence (cf. ‘zone of proximal development’ or ZPD; Vygotsky 1978). From the increased volume



The first English (EFL) lesson 

 267

and sonority of the choir we may infer that by the third round many students had joined in for a first try or had self-corrected their initial pronunciation errors. We do not know exactly who learned what, how much progress any individual student made in acquiring articulatory skills or (the beginning of) a ‘voice’ in the L2. But does that matter? It is the ecology, the ‘overall participation framework as an emergent process that ‘does the learning’ just as [...] it is the participation framework in which an utterance acquires meaning that ‘does the talking’ (Hanks 1996; Leather and Van Dam 2003b: 13). The teacher rounds off the episode by spelling a few simple English words like ‘body’ that the students individually have to write down in their exercise books (introduced as ‘don’t worry – it’s just a game’). When they have done so she provides the correct answer on the blackboard and has each student self-correct: ‘this is what should have been there’. Students spontaneously, report their errors. The letter ‘y’ is unanimously identified as a trouble source which evokes affiliative comments in the peer group. Several other words are spelled, culminating in a particularly difficult one that is explicitly announced as a challenge for exceptionally bright students: the word ‘exaggerate’. The teacher adds that she does not expect anyone to ‘get it’. Students who get it right, however, are emphatically praised. They are framed as brilliant. Getting it wrong is ‘normal’. Students publicly report their errors with enthusiasm. Guttural noises, i.e. phonological play on ‘g’, in Dutch pronounced as /x/, echo through the classroom and provoke hilarity. Interview meta-data. In one of the interviews we conducted with this teacher we asked her whether her promise that existing differences in proficiency would soon be negotiated was realistic. The answer was “yes – by Christmas in most cases – yes”. She spontaneously added that all-class co-speaking and chorusing were a crucial tool in that respect: “They all get lots of practice and nobody fails”. She told us she initially only allocated individual turns to students (e.g. to Frederic in 4.1) when she had noticed that they were already confident in English. When asked whether the brighter students did not get bored by repetitive classroom routines like ‘saying’ the numbers or the alphabet, she said that she usually included more challenging extra tasks for the more advanced students. They were invited, for instance, to say the letters of the English alphabet while simultaneously reading their phonetic notations. For the same reason, she occasionally solicited the spelling of English words like ‘exaggerate’ which are particularly difficult for Dutch learners. What transpired was that in these collective participation modes the smart can be smart without the less smart being framed as dumb (cf. ‘smart-dumb continuum’; McDermott and Tylbor 1986: 132). From the second week onwards theatrical performances in class provided ample opportunities for bright students to further develop their oral skills – and be applauded by their peers (see section 6.1).

268 

 Jet Van Dam and Anne Bannink

5.4 Playing with cultural frames 3: spelling contest The collective choir practice is followed by a spelling game that gives the students another opportunity to practice their newly-acquired skills. The class is taken by complete surprise when the teacher suddenly, at great speed, gives a virtuoso performance of spelling the name “Ottiwell Wood”. She recites it as a two line chant, with a marked melodic line and a rhythmic break and change of tempo in the middle, where the series of double ‘l’s and double ‘u’s starts:

The learners all sit up and there are exclamations of admiration and appreciation all around; some students even applaud the teacher’s performance. This is how the episode is introduced: Data 6 T OK dan ga ik je nou een NAAM laten spellen - die je HEE:L snel kan uitspreken – ik zal hem EERST opschrijven - [draait naar het bord] - de naam – van de man is -  [schrijft naam op bord] Ot-ti-well - ↓idiote naam – ↑Wood [wijst naar bord] DIE kun je spellen heel snel [andere toon; presto, ritmisch] ‘o’double ‘t’ - ‘i’ - ‘double u’ - ‘e’ [korte pauze; prestissimo] double ‘l’ double ‘u’ double ‘o’ /di:/ = SSS =[gelach, applaus] SSS = OH – JEE – JA: T JA↑[kijkt rond] – daar gaan we T OK then I’ll have you spell a NAME now - that you can say very FAST - I’ll write it down FIRST - [turns to blackboard] - the name - of the guy is - [writes on blackboard] ‘Ottiwell - [sotto voce] ↓ idiotic name – ↑Wood’ [points at blackboard] THOSE you can spell – very FAST [spells very fast in rhythmic cadence] ‘o’-double ‘t’ - ‘i’ - ‘double u’ ‘e’ [pause; prestissimo] double ‘l’ - ‘double u’ - ‘double o’ - ‘d’= SSS =[laughter, applause] SSS = OH- WOW - YES T OK↑ [looks around] – here we go



The first English (EFL) lesson 

 269

The chant is practiced four times. The teacher varies her instruction in successive rounds: “I’ll do it SLOWly”; “NOW we do it just a LITTLE faster - you join in”; “the rhythm is beginning to come - I hear”; “ONE MORE time: YOU start and then I’ll join in”. Again the students participate with gusto, the sound becoming more uniform and robust with each repetition, as more and more people join in. The task/chant provides opportunities for all to practice articulatory skills of considerable sophistication. It also highlights some characteristic differences between Dutch and English spelling. It contains the notoriously difficult vowels ‘e’ and ‘i’ that are often confused because the Dutch letter ‘i’ (/i:/) is pronounced almost like the name of the English letter ‘e’; and English ‘w’ is referred to as ‘twice another letter’ rather than having a name of its own, as in Dutch. Since the chant features phonological repetition as well as rhythmic musical structure and rhyme, it is easy to commit to memory (cf. Cook 2000; Schieffelin 1983; Sullivan 2000) and brings to mind the sound play that is characteristic of early stages of L1 acquisition. Ochs observes that discourse formats acquired in early childhood are not replaced by later more sophisticated ones but are retained to be shifted back to when the situation so demands (Ochs and Schieffelin 1983; Cook 2000). The tongue twister is set for homework: the students are to practice pronouncing it at great speed. Casually the teacher adds: “ÉÉN keer heeft een leerling het van me gewonnen – dat vond ik natuurlijk helemaal niet leuk ..”/“just ONCE there was a student who beat me at speed – of course I did not like that at all so I stopped doing it for a while. But now I trust I will win again, as usual”. The game is on: a competition involving a virtual student has begun.

5.5 Playing with cultural frames 4: a children’s game The last item on the lesson agenda was listening to a song and co-singing its refrain. That plan has to be abandoned, however, because it turns out the audio device in the classroom is not working properly. A cacophony of sounds elicits spontaneous comments from the students, in different languages: “kaput” (German for “out of order”; cf. Rampton 2006; Van Dam 2003) and “it’s ONE blur” can be heard on the tape. The teacher calls in the help of an expert student – to no avail. She then re-sets the agenda: they will do a guessing-game instead. It is a variation on ‘I spy with my little eye’, the conventional children’s game that, with minor adaptations, is imported as a collective learning frame in the classroom. One student leaves the classroom and on re-entering has to guess which object the class has chosen to be ‘it’. The teacher is very strict now and insists they follow the rules of the game: “no” and “yes” are incorrect answers; it has to be “no, it isn’t” or “yes, it is”.

270 

 Jet Van Dam and Anne Bannink

As in some of the earlier episodes the students seem to be totally involved. There is evidence of great energy, focus and flow (Csikzentmihalyi 1996) in their spontaneous contributions: “WARM! – WARMER! – [loud, presto] HOT-HOTHOT – [slow and low-intoned] CO::LD – CO::LDER – [loud and clear] ICECREAM – [laughter] – ICE ICE!” [overtaken by the bell]. The collective game format allows for linguistic creativity and play on the part of individual students from the side line while maintaining a conventional all-class participation frame (cf. Sullivan 2000).

6 Observations later in the year In this section we will share some of the observations and field notes of lessons we attended later in the year, involving this teacher and the same group of students. 

6.1 Three weeks later We returned to the same class three weeks later. The lessons had settled into a routine by then. The first half of the lesson the students worked from a conventional course book with explanations of elementary points of grammar and lists of vocabulary items they needed for what turned out to be the key episode in the lesson: the dramatic performance of the set lesson dialogue. Running through the course book was a continuing story of an accident-prone teenager with an ill-fated love affair and a pompous rival. Dialogues connected to this storyline opened each chapter and were the focus of each lesson. The students listened to the audiotape of each dialogue a number of times and rehearsed difficult words and parts in chorus. Now the stage was set for the highlight of the class: a role play performed by the students on the basis of these dialogues. Students ­enthusiastically bid for the roles; boys were allowed to play the female roles and vice versa – which added to the general mirth. Both the actors and the audience clearly enjoyed this lesson episode: they all stepped into the fictional world of the unfortunate hero with gusto. After three weeks the students had already become quite proficient in imitating and exaggerating the intonation patterns and tone of voice of the characters in the story – which frequently provoked appreciative laughter in the audience as well as online comments on the performance of the actors. When the student playing the love interest of the main character successfully imitated her high-pitched voice, there was a delighted: “Miss Piggy, Miss Piggy!” from the audience. The students were so familiar with the language of the dialogues that, ­overwhelmingly,



The first English (EFL) lesson 

 271

they were able to attend to the coordination of verbal and nonverbal behaviors. Just once there was a hitch which caused widespread hilarity: one of the actors thanked the waiter for a plate that had not yet been placed in front of him. The actors occasionally improvised or made small changes in their English lines – variations on familiar, simple words and constructions. The performance was repeated several times so the majority of the learners had the opportunity to participate. Weaker students had shorter parts. When, after the third performance, the class rhythmically shouted “an-o-ther-time, an-o-ther-time!”, the teacher granted their request – but bribed them into practicing vocabulary items for the next lesson and doing a grammar exercise first. After that the play was repeated yet another three times. By the end of the lesson nearly everyone had played a part – with amazing fluency.

6.2 Eight months later In the class we observed at the end of the same school year the teacher had organized the lesson activities around a language task called ‘Alibi’. She instructed the students to draw the outlines of three human figures on a piece of paper according to the descriptions she read out to them in the role of the presenter of a TV crime show. She told the class that the police appealed to the public to solve a robbery that had taken place in the very neighborhood of the school. Witnesses had come up with detailed descriptions of three suspects. She then described each suspect in detail (“The first suspect is Al Jones. He is six foot tall and according to witnesses he is wearing a striped sweater, rubber boots and he has a moustache”). When she had finished she invited the students to compare their drawings in groups of three and had them describe one of them back to her, using the progressive or the 3rd person present tense in the appropriate slots. There was no need to allocate turns: there were lots of spontaneous student contributions. She then presented them with a new situation. They had to imagine they were on a school trip to London. In the morning the students are surprised that she (the teacher) has not come down for breakfast. They go up to her room and find her bed not slept in. They decide to go to the police and report her missing. The policeman in charge requires an exact description of the lost person. What would they say? The teacher joked: “No comments about my weight, please.” They set to work, again in groups of three. The teacher walked around the class, answered questions about vocabulary items and was occasionally asked to come and stand close so the students could check e.g. the exact color of her eyes. The teacher then invited the groups to take turns and name one feature of the description at a time, which she would then write down on the blackboard. There was lots of

272 

 Jet Van Dam and Anne Bannink

laughter and teasing during this episode (“she is wearing old lady shoes”; “she giggles all the time”); the students had great difficulty curbing their enthusiasm and waiting for their turn. When the class ended there was a long list of 15 items on the blackboard. Again, institutional and default cultural roles are dialogically merged. The teacher is both a character in the embedded story-world and the director of talk in the embedding classroom world. As a figure in the story the students may mildly mock her while respecting her in her institutional role.

7 Conclusions, discussion and research issues In this chapter we have presented detailed analyses of what happened, almost on a moment-by-moment basis, in the course of a first EFL lesson in a secondary school in the Netherlands. It involved a group of twenty-six 12 or 13-year-old learners on their first day in a new school. We zoomed in on the various forms and modalities of play (phonological; semantic, pragmatic; semiotic) that were in evidence throughout the lesson. They were initiated both by individual students and by the teacher, in traditional IRFs and in collective participation formats like allclass chanting and chorusing, co-speaking and say-after-me episodes. We traced the affordances of these robust cultural frames in synchronizing and scaffolding the participation and learning of many – while simultaneously allowing for individual variation and play in off-record niches of institutional speaking slots. Tentatively we also addressed longitudinal dimensions of the role of instances of play in mediating an emerging ‘ludic’ classroom culture. Play is by definition parasitic on other templates or interactional norms. Instances of play resonate against normative expectations for next moves that are non-playful. In that respect the ‘reciprocal’ relation between ritualformulaic output frames and instances of L2 language play suggested by Bell (2012b) is ­corroborated in our data. Its further exploration indeed calls for a ‘unified model’ (2012b: 119) to which we hope our proposals in this chapter may contribute. In section 4 we describe how the IRF that is often regarded as a prototypical institutional format can itself be reframed as play. One student’s spontaneous play on L1 and L2 names and identities was taken up by similarly playful stances on the part of the teacher and two peers – making the notion of what counts as a relevant next move at the level of the pedagogical agenda ­ambiguous or opaque. Complex speaking roles and voices are all-pervasive in this lesson, both on the part of the teacher and the learners. Bakhtin’s much-quoted adage that one’s word or speaking voice is half someone else’s almost literally applies in the choir



The first English (EFL) lesson 

 273

episodes (section 5.3). The marked increase in volume and sonority of the choir in the third round can only be accounted for if we assume that peer-scaffolded self-corrections or first tries have massively taken place in off-record niches of institutional all-class speaking slots. These provide evidence of learning but we do not know what progress any individual learner has made along the continuum of non-vocal (lip movements); half-vocal (whisper voice), intermittent (co-speaking on an item-by-item basis) to fully-vocal participation modes. We noted also how the teacher herself frequently shifted out of and back into her institutional role and voice, often within a single speaker-slot-in-progress. She dialogically modelled the voice of a student who is at a loss, the cheerleader of a choir, the organizer of a wasp hunt or contest with a virtual peer – as well as informal versions of herself, e.g. as a person who loves children or hates to lose. These playful or theatrical stances are ‘serious’ in so far as they feed into classroom tasks such as constructing feedback, minimizing fear of failure or loss of face, restoring classroom order, challenging individual students and team-building. They may have invited, we hypothesize, the similarly playful footings (Goffman 1979) on the part of learners that were in evidence throughout the lesson. In ESL settings (or monolingual EFL classes for that matter) where the cognitive load of figuring out what is being said in class is already high (cf. Duff 2004; Lytra 2008) such playful framing practices may be more difficult to implement owing to the absence of a shared L1 ­meta-language. Confusions or misunderstandings may arise as a result of the ambiguous status of playful contributions to institutional multiparty discourses-in-progress (section 4.3). If not taken up by the teacher these may be interpreted as structural moves that do not update the state of talk at the level of the lesson agenda – and which for that reason may be ignored. This is a key issue, we argue, in the study of the role of play in classroom settings; it should be made analytically more transparent. The theoretical formalism we used to analyze the data (‘classroom machine’; section 2) aims to model the interpretive strategies that participants – in our metaphor: machines that simultaneously read multimodal input on a move-bymove basis – bring to bear on complex classroom situations. It allows linear as well as structural transitions from state to state (cf. Larsen-Freeman and Cameron 2008; Seedhouse 2010): the latter signal the context changes that participants demonstrably orient to when making sense of classroom talk and events. When institutional templates for the organization of talk and interactional behaviors are temporarily overruled (as in the wasp hunt; section 5.2) they are still oriented to at embedded levels of the interaction and can be re-embedded at a moment’s notice. Since our analytic tool introduces a dynamic complex notion of what could be a relevant next move in a given state of talk, it is, in principle, able to articulate the discourse parameters of e.g. parallel states of talk and behaviors

274 

 Jet Van Dam and Anne Bannink

that are selectively ignored in class – typically loci where instances of situated play may occur. A rich notion of what are the data to be addressed and close attention to interactional detail are essential in investigating the role of play in L2-classroom multiparty settings. Code-switches and changes in tempo or tone of voice (cf. the chant in section 5.4); laughter and chuckling; shifts in gaze direction or bodily orientation and other semiotic features accompanying talk-in-interaction are crucial cues in monitoring the frame ambiguities or frame complexities that instances of play often address – as well as create. The rehearsal-type all-class speaking slots that prominently feature in this lesson are less vulnerable in that respect. Since they do not acknowledge conversational adjacency they constitute what Goffman has called “a teething ring for utterances rather than a ball game” (1981: 151 note). The affordances of these ritual participation formats for early language acquisition lie precisely in the fact that the notion of a discrete individual turn at talk may be temporarily overruled. In the final sections we tentatively address the issue whether this teacher’s teaching strategies and playful classroom style might feed into the long-term L2 learning trajectories of many. Assuming that finding a (new) voice or identity for oneself in the new language is a key element in acquiring L2 competence, it seemed that after three weeks the students were already well on their way to doing so – in play: symbolically modelling an L2 character in a story world rather than being ‘themselves’ (whatever that may mean in classroom settings..). Their performances were truly stunning: fluent and authentic-sounding in spite of occasional slips and hitches. There was great flow and motivation was high: everybody volunteered for a part. The classroom game also involves calculation: ‘if you play our game, we play yours.’ When soliciting yet another performance, the class was bribed into doing grammar/vocabulary exercises first. In the lesson we observed at the end of the year we witnessed the same mix of situated play and task-orientedness. What the teacher and the students were doing together, it seemed, was making fun of the lesson frame while being fully involved in it. Discussion. Why did we select this lesson for analysis? Many of the practices described in this chapter are firmly located in the predigital age. Well – apart from the fact that it is hard to find a lesson which is clearly experienced by the learners as ‘fun’ while they all get so much practice, learn articulatory skills fast and are ‘on task’ nearly all of the time, in practical terms many classrooms all over the world still live in the predigital age. Moreover, a better understanding of the ways language learning trajectories in face-to-face settings crucially involve ‘laminative discourse features’ – such as “learning to speak for, learning to speak in the name of figures that will never be, or at least aren’t yet the self”



The first English (EFL) lesson 

 275

(Goffman 1981: 151 note) – may feed into thinking about the affordances of computer-mediated communication and community-based digital engagements. Detailed descriptions of traditional face-to-face classroom practices (and learner perspectives on those practices; cf. Van Dam 2003) may lead us to re-assess what relevantly happens in the language classroom – and to what effect. We tentatively suggest that there may be a historical dimension to instances of play: participants may develop a memory for it, finding new opportunities to implement their symbolic freedom across classroom situations. Research issues. In section 4.1 we described a particularly sophisticated example of ‘spoken artistry’ (Sullivan 2000) on the part of a beginning L2 learner: the dramatically-whispered ‘Mr. X’ (see data 1C). It may have been designed for the overhearer floor (Cekaite and Aronson 2005) and illustrates the difficulty of making forms of play in classroom talk analytically transparent. While being a perfectly correct, coherent and cooperative English lesson answer it was either not-noticed by the teacher or intentionally ignored. It both plays upon and mocks and iconically represents (and expresses resistance to?; cf. Pomerantz and Bell 2011) the ritual-formulaic nature of classroom question-answer sequences. It conveys all of these possible pragmatic meanings both verbally, nonverbally and semiotically through the phonological and prosodic features of its production – and, of course, its timing. Are we meant to hear it? Who is ‘we’? Does it count as data? If so, all features of its production must be included as data. What triggers it and what is its effect? Can it be reduplicated under experimental conditions? These are really complex questions that as yet our models – however sophisticated – cannot address. For the moment we just allude to them with a wink: Can we teach a robot to play? Or a trainee teacher?

Transcription conventions T teacher S student Sx unidentified student SF identified student, initial indicating first name SSS several students simultaneously Class (nearly) all students NEE capital letters, indicating strong emphasis and/or louder voice ↑ marked rise in pitch ↓ marked fall in pitch – unmarked pause = immediately adjacent utterances

276 

 Jet Van Dam and Anne Bannink

[ ]

c ontextual information, prosodic features, events in the situation, nonverbal features lengthening of preceding sound unintelligible vocalizations translation

: (::) (...) ((.. ))

Original transcript in italics; translation in Roman.

References Agar, Michael. 2004. We have met the other and we’re all nonlinear: Ethnography as a nonlinear dynamic system. Complexity 10(2). 16–24. Auerbach, Elsa.1993. Re-examining English-only in the ESL classroom. TESOL Quarterly 27. 9–32. Austin, John. 1962. How to do things with words. Oxford: Oxford University Press. Bakhtin, Mikhail. 1981. The dialogic imagination: Four essays by M. M. Bakhtin. (C. Emerson & M. Holquist, Trans.). Austin, TX: University of Texas Press. Bakhtin, Mikhail. 1986. Speech genres and other late essays. Austin: University of Texas Press. Bannink, Anne & Jet Van Dam. 2006. A dynamic discourse approach to classroom research. Linguistics & Education 17(3). 283–301. Bannink, Anne & Jet Van Dam. 2013. The first lecture: playing upon identities and modeling academic roles. Linguistics & Education 24(4). 556–571. Bateson, Gregory. 1972. Steps to an ecology of mind. New York: Ballantine Books. Bell, Nancy. 2012a. Comparing playful and nonplayful incidental attention to form. Language learning 62(1). 236–265. Bell, Nancy. 2012b. Formulaic language, creativity and language play in a second language. Annual Review of Applied Linguistics 32. 189–205. Cazden, Courtney. 1988. Classroom discourse: the language of teaching and learning. Portsmouth, New Jersey: Heinemann. Cekaite, Asta & Karin Aronsson. 2004. Repetition and joking in children’s second language conversations: playful recyclings in an immersion classroom. Discourse Studies 6. 373–392. Cekaite, Asta & Karin Aronsson. 2005. Language play: A collaborative resource in children’s L2 learning. Applied Linguistics 26. 169–191. Cekaite, Asta. 2006. Getting started: children’s participation and language learning in an L2 classroom. Linköping: Linköping Studies in Arts and Science (35). Csikszentmihalyi, Mihaly. 1996. Flow and the psychology of discovery and invention. New York: Harper Perennial. Cook, Guy. 2000. Language play, language learning. Oxford: Oxford University Press. Donato, Richard. 1994. Collective scaffolding in second language learning. In James Lantolf & Gabriela Appel (eds.), Vygotskian approaches to second language research, 33–56. Norwood, NJ: Ablex. Duff, Patricia. 2004. Intertextuality and hybrid discourses: The infusion of pop culture in educational discourse. Linguistics and Education 14(3). 231–276.



The first English (EFL) lesson 

 277

Duranti, Alessandro, Elinor Ochs & Bambi Schieffelin (eds.). 2011. The handbook of language socialization (Vol. 72). Hoboken, NJ: John Wiley & Sons. Erickson, Frederick & Jeffrey Shultz. 1981. When is a context? Some issues and methods in the analysis of social competence. In Judith Green & Cynthia Wallat (eds.), Ethnography and language in educational settings, 147–160. Norwood, NJ: Ablex. Finch, Andrew. 2010. Critical incidents and language learning: Sensitivity to initial conditions. System 38(3). 422–431. Goffman, Erving. 1979. Footing. Semiotica, 25. 1–29. Goffman, Erving. 1981. Forms of talk. Oxford: Basil Blackwell. Hall, Graham & Guy Cook. 2012. Own-language use in language teaching and learning. Language Teaching 45(03). 271–308. Hall, Joan Kelly. 1997. A consideration of SLA as a theory of practice: A response to Firth and Wagner. Modern Language Journal 81. 301–306. Hall, Joan Kelly & Teyana Smotrova. 2013. Teacher self-talk: interactional resource for managing instruction and eliciting empathy. Journal of Pragmatics 47(1). 75–92. Hanks, William F. 1996. Language and communicative practices. Boulder, CO: Westview Press. Heath, Shirley Brice. 1983. Ways with words: Language, life, and work in communities and Classrooms. Cambridge: Cambridge University Press. Hellermann, John. 2003. The interactive work of prosody in the IRF exchange: teacher repetition in feedback moves. Language in Society 32. 79–104. Hellermann, John. 2005. Syntactic and prosodic practices for cohesion in series of three-part sequences in classroom talk. Research on Language and Social Interaction 38(1). 105–130. Heritage, John. 1984. A change-of-state token and aspects of its sequential placement. In John Atkinson & John Heritage (eds.), Structure of social action: Studies in conversation analysis, 299–345. Cambridge: Cambridge University Press. Huizinga, Johan. 1949. Homo Ludens. London: Routledge and Kegan Paul. Kamberelis, George. 2001. Producing heteroglossic classroom (micro)cultures through hybrid discourse practices. Linguistics and Education 12(1). 85–125. Kramsch, Claire. 1993. Context and culture in language teaching. Oxford: Oxford University Press. Kramsch, Claire (ed.). 2002. Language acquisition and language socialization. London: Continuum. Kramsch, Claire. 2009. The multilingual subject. Oxford: Oxford University Press. Lantolf, James (ed.). 2000. Sociocultural theory and second language learning. Oxford: Oxford University Press. Lantolf, James & Gabriela Appel. 1994.Vygotskian approaches to second language research. Norwood, NJ: Ablex. Larsen-Freeman, Diane & Lynne Cameron. 2008. Complex systems and applied linguistics. Oxford: Oxford University Press. Leather, Jonathan & Jet Van Dam (eds.). 2003a. Ecology of language acquisition. Dordrecht: Kluwer Academic Publishers. Leather, Jonathan & Jet Van Dam. 2003b. Ecology of language acquisition, 1–29. Dordrecht: Kluwer Academic Publishers. Lee, Yo-An. 2007. Third turn position in teacher talk: Contingency and the work of teaching. Journal of Pragmatics 39. 1204–1230. Levinson, Stephen. 1983. Pragmatics. Cambridge: Cambridge University Press.

278 

 Jet Van Dam and Anne Bannink

Lytra, Vally. 2007. Play frames and social identities: Contact encounters in a Greek primary school. Amsterdam: John Benjamins Publishing. Lytra, Vally. 2008. Playful talk, learners’ play frames and the construction of identities. In Marilyn Martin-Jones, Ana Maria de Mejia & Nancy Hornberger (eds.), Encyclopedia of Language and Education, Volume 3, 185–197. New York: Springer. Markee, Numa & Mi-Suk Seo. 2009. Learning talk analysis. International Review of Applied Linguistics in Language Teaching 47. 37–63. Maybin, Janet (2008). Revoicing across learning spaces. In Marilyn Martin-Jones, Ana Maria de Mejia & Nancy Hornberger (eds.), Encyclopedia of Language and Education, Volume 3, 81–92. New York: Springer. McDermott, Ray. 1988. Inarticulateness. In Deborah Tannen (ed.), Linguistics in Context: Connecting Observation and Understanding, 37–68. Norwood: NJ: Ablex. McDermott, Ray & Henry Tylbor. 1986. On the necessity of collusion in conversation. In Leo Kedar (ed.), Power through Discourse, 153–170. Norwood, NJ: Ablex. McHoul, Alexander. 1978. The organization of turns at formal talk in the classroom. Language in society 7(2). 183–213. Mehan, Hugh. 1979. Learning lessons. Cambridge, MA: Harvard University Press. Mehan, Hugh. 1980. The competent student. Anthropology and education quarterly 11. 131–152. Mehan, Hugh. 1998. The study of social interaction in educational s ettings: Accomplishments and Unresolved Issues 1. Human development 41(4), 245–269. Mondada, Lorenza. 2009. Emergent focused interactions in public places: A systematic analysis of the multimodal achievement of a common interactional space. Journal of Pragmatics 41(10). 1977–1997. Nassaji, Hossein & Gordon Wells. 2000. What’s the use of ‘‘triadic dialogue’’?: An investigation of teacher–student interaction, Applied Linguistics 21(3). 376–406. Norris, Sigrid. (2004). Analyzing multimodal interaction: A methodological framework. New York: Routledge. Ochs, Elinor & Bambi B. Schieffelin. 1983. Acquiring conversational competence. London: Routledge & Kegan Paul Books. Polanyi, Livia. 1988. A formal model of the structure of discourse. Journal of Pragmatics 12(5). 601–638. Polanyi, Livia & Remko Scha. (1983). On the recursive structure of discourse. In Kurt Ehlich, & Henk van Riemsdijk (eds.), Connectedness in sentence, discourse and text, 141–178. Tilburg: Tilburg University. Pomerantz, Anne & Nancy Bell. 2007. Learning to play, playing to learn: FL learners as multicompetent language users. Applied Linguistics 28(4). 556–78. Pomerantz, Anne & Nancy Bell. 2011. Humor as pedagogical safe house in the foreign language classroom. The Modern Language Journal 95. 148–161. Rampton, Ben. 1999. Dichotomies, difference, and ritual in second language learning and teaching. Applied Linguistics 20(3). 315–340. Rampton, Ben. 2006. Language in Late Modernity: Interaction in an Urban School. Cambridge: Cambridge University Press. Schegloff, Emanuel. 1968. Sequencing in conversational openings. American Anthropologist 70. 1075–95. Schieffelin, Bambi B. 1979. Getting it together: an ethnographic approach to the study of the development of communicative competence. In Bambi B. Schieffelin & Elinor Ochs (eds.), Developmental Pragmatics, 73–108. New York: Academic Press.



The first English (EFL) lesson 

 279

Schieffelin, Bambi. 1983. Talking like birds: sound play in a cultural perspective. In Elinor Ochs & Bambi B. Schieffelin, Acquiring Conversational Competence, 177–84. London: Routledge & Kegan Paul Books. Seedhouse, Paul. 2010. Locusts, snowflakes and recasts: complexity theory and spoken interaction. Classroom Discourse 1(1). 4–24. Sinclair, John & Malcolm Coulthard. 1975. Towards an analysis of discourse: the English used by teachers and pupils. Oxford: Oxford University Press. Skidmore, David & Kyoko Murakami. 2010. How prosody marks shifts in footing in classroom discourse. International Journal of Educational Research 49(2). 69–77. Sperber, Dan & Deirdre Wilson. 1981. Irony and the use-mention distinction. In Peter Cole (ed.) Radical Pragmatics, 295–318. New York: Academic Press. Sullivan, Patricia. 2000. Playfulness as mediation in communicative language teaching in a Vietnamese classroom. In James Lantolf (ed.), Sociocultural theory and second language learning 115–132. Oxford: Oxford University Press. Swain, Merrill. 1995. Three functions of output in second language learning. Principle and practice in applied linguistics: Studies in honour of HG Widdowson, 2(3). 125–144. Van Dam van Isselt, Jet. 1993. “Her name is - uh dat weet ik niet” - Authenticity in the L2 classroom. Amsterdam: University of Amsterdam dissertation. Van Dam, Jet. 2002. Ritual, face and play in a first English lesson: bootstrapping a classroom culture. In Claire Kramsch (ed.), Language acquisition and language socialization, 237–265. London: Continuum. Van Dam, Jet. 2003. Language acquisition behind the scenes: collusion and play in educational settings. In Jonathan Leather and Jet Van Dam (eds.), Ecology of Language Acquisition, 203–223. Dordrecht: Kluwer Academic Publishers. Van Lier, Leo. 1988. The classroom and the language learner. New York: Longman Group Ltd. Varenne, Hervé & Ray McDermott. 1998. Successful failure. Boulder, CO: Westview. Volosinov, Valentin. 1973. Marxism and the philosophy of language (L. Matejka & I. R. Titunik, Trans.). Cambridge, MA: Harvard University Press. Vygotsky, Lev. 1978. Mind in society: the development of higher psychological processes. Cambridge, MA: Harvard University Press. Vygotsky, Lev. 1986. Thought and language (A. Kozulin, trans.). Cambridge, MA: MIT Press. Waring, Hansun Zhang. 2009. Moving out of IRF (initiation-response-feedback): A single case analysis. Language Learning 59. 796–824. Waring, Hansun Zhang. 2012. Doing being playful in the second language classroom. Applied Linguistics 34(2). 191–210. Watzlawick, Paul, Janet Beavin & Don Jackson. 1967. Pragmatics of human communication. New York: Norton. Wells, Gordon. 1993. Reevaluating the IRF sequence: a proposal for the articulation of theories of activity and discourse for the analysis of teaching and learning in the classroom. Linguistics and Education 5(1). 1–37. Wittgenstein, Ludwig. [1953] 1968. Philosophical investigations. Oxford: Blackwell.

Søren W. Eskildsen

11 The emergence of creativity in L2 English: A usage-based case-study Abstract: This chapter investigates emergent creativity in L2 English from a usage-based perspective (for a recent overview of usage-based L2 research, see Cadierno & Eskildsen 2015). The chapter adopts the trace-back methodology from child language studies (Lieven et al. 2009; Vogt and Lieven 2010), a method which allows researchers to tease out and explore the constructional fabric and thereby emergent creativity evidenced in language users’ linguistic repertoires. The method reveals the ways in which constructions in particular speakers’ language use are related to other constructions in the same L2 speaker’s repertoire at particular points in time and across time by way of tracing actual utterances backwards in longitudinal developmental data. The child language studies showed that children’s emergent repertoires consisted of formulas and utterance schemas, and that only a few basic syntactic operations accounted for early syntactic creativity. Applying the trace-back procedure to my L2 data revealed that the syntactic operations identified by Lieven and colleagues, especially substitution of semantically similar items in open slots in recurring constructions, also apply to beginning adult L2 learning (Eskildsen 2014). Recycled formulas and novel utterances based on utterance schemas and the substitution operation accounted for more than 85% of my focal student’s production, indicating that recycling is important in L2 learning and that emergent creativity is exemplar-based. Eskildsen (2014) also indicated that more complex processes are involved when L2 users combine existing constructions in novel ways. This chapter explores these complex processes further as it maps out emergent creativity in the same learner of L2 English at both beginning and more advanced stages. It corroborates the findings from the first study but also refines the methodology by reducing the number of syntactic operations needed to account for the emergent constructions. Finally, the chapter demonstrates, through local analyses of the usage-based learning of one particular construction, centered on the verb need, that language play in the classroom can be conducive to L2 learning (Bell 2012b, Waring 2013).

DOI 10.1515/9781501503993-012

282 

 Søren W. Eskildsen

1 Introduction: L1 and L2 creativity from a usage-based perspective Drawing on usage-based linguistics, i.e. a construction-based view of language and a view of language learning as an exemplar-based process of inferring schematic similarities among encountered formulas and utterance schemas (i.e. formulaic frames) (Ellis 2002; Tomasello 2003), this chapter reports on a larger project which investigates emergent creativity in L2 English. By creativity here is meant a person’s ability to expand existing L2 patterns in and through talk. The project draws on the usage-based traceback methodology from child language studies (Lieven, Salomo, and Tomasello 2009; Vogt and Lieven 2010), a method which allows researchers to tease out and explore the constructional fabric of the emergent linguistic repertoire in terms of the ways in which constructions put to use are related to other constructions in the repertoire across time by way of tracing actual utterances backwards in longitudinal developmental data (Eskildsen 2014). This chapter adds to the project by also investigating in more detail a locally contextualized example of L2 construction development through the lens of language play (Bell 2012a). The operationalization of creativity in this study and the possiblity and relevance of applying a trace-back methodology to investigate it derives from my theoretical positioning in usage-based linguistics (UBL). UBL is a cover-term for a range of models within functional-cognitive linguistics and social-constructivist child language research and – recently – second language acquisition (SLA) research (Ambridge and Lieven 2011; Barlow and Kemmer 2000; Cadierno and Eskildsen 2015; Ellis 2002, 2015; Ellis, O’Donnel, and Römer 2013; Eskildsen 2009; Langacker 1987; Tomasello 2003; Tummers, Heylen, and Geeraerts 2005). The core principle uniting these models is the fundamental importance ascribed to language use. Language development, phyologenetic and ontogenetic, is shaped by language use, and linguistic structure emerges in and from language use as pairings of form and meaning in usage events. What may be described as a human creative language ability emerges hand-in-hand with the general emergence of structure as people learn to master and manipulate the emerged form-meaning pairings. Such learning is also usage-driven and builds on countless occasions of experiencing and using language within different domains – for example, learning how to pluralize nouns is crucially dependent on rote-learning exemplars of the plural and, on the basis of this experience, extracting the generalized plural schema allowing for novel pluralizations. UBL proponents have investigated how a creative linguistic inventory comes into being on the basis of concrete recurring linguistic material in use. Such research (e.g. Brandt et al. 2011; Dabrowska and Lieven 2005; Eskildsen 2014; Lieven 2010; Lieven et al. 2003; Lieven, Salomo, and Tomasello 2009;



The emergence of creativity in L2 English: A usage-based case-study 

 283

­ acWhinney 1975, 1982; Tomasello 2000, 2003) has found language learning to M be concrete, exemplar-based, and rooted in usage, following a trajectory from specific recurring multi-word expressions to partially fixed, partially schematic utterance schemas to increasingly schematic constructions based on systematic commonalities among patterns. The commonalities, derived by the language user through social interaction, come to be represented in the mind, at the most advanced levels of learning, as schemas sanctioning the use, understanding, and learning of novel expressions of the same kind. This is the key issue in understanding creativity from a UBL vantage point. UBL, then, rejects the compartmentalization of language into (i) something that is listed in the mental lexicon, swiftly processed, and stable; and (ii) something that is rule-governed, cognitively demanding, and flexible (Eskildsen and Cadierno 2007). Instead, language knowledge is shown to be a structured inventory of symbolic units, i.e. form-meaning pairings (Langacker 1987) of varying complexity and abstraction. The units vary in size from morphemes to full expressions and range from the totally specific to the maximally general, allowing for the co-habitation in the grammar of abstract schematized representations and their concrete instantiations (e.g. Langacker 2000; Achard 2007). Usage-based perspectives have been gaining rapid attention within SLA research (Eskildsen and Cadierno 2015; Ortega 2014). The ontological status of constructions as form-meaning pairings in L2 learning has been empirically supported (e.g. Bartning and Hammarberg 2007; Collins and Ellis 2009; Ellis and Ferreira-Junior 2009a, 2009b; Ellis, O’Donnel, and Römer 2013; Goldberg and Casenhiser 2008; Gries and Wulff 2005, 2009; Robinson and Ellis 2008; Waara 2004), and a growing body of research (Ellis and Ferreira-Junior 2009a; Eskildsen 2009, 2011, 2012a, 2014, 2015, in press; Eskildsen and Cadierno 2007; Eskildsen, Cadierno, and Li 2015; Li, Eskildsen, and Cadierno 2014; Mellow 2006; RoehrBrackin 2014; Yuldashev, Fernandez, and Thorne 2013) is documenting and discussing L2 learning over time in terms of an exemplar-based process where the L2 user is constantly developing a repertoire of interrelated constructions on the basis of recurring exemplars. The present chapter expands on the body of longitudinal usage-based L2 research by exploring the specific ways in which linguistic creativity emerges in L2 learning. To achieve this, I draw heavily on Eskildsen (2014) in which I adopted the traceback methodology from child language studies (Lieven et al. 2003; Dabrowska and Lieven 2005; Lieven, Salomo, and Tomasello 2009; Vogt and Lieven, 2010; cf. also Yuldashev, Fernandez, and Thorne 2013), a method which allows researchers to tease out and explore the constructional fabric of emergent grammars in terms of the ways in which constructions put to use are related to other constructions in the inventory across time. This method thus presents a window onto the anatomy of the linguistic inventory at various points in

284 

 Søren W. Eskildsen

time and investigates the relationship between creative and routinized aspects of it. As will be shown, emergent creativity is rooted in experience, it grows slowly from affordances and from the speakerʼs amassed repertoire of constructions through use. One of the central dogmatic tenets of linguistics in the 20th century, namely that language is creative to such an extent that everytime we encounter an utterance or sentence it will be something we have never encountered before is simply wrong; instead, speakers reuse what is around them, and everytime they speak, they say something that has been said in exactly the same or a slightly different way before (Hopper 1998).

2 Formula, creativity, and language play Central to the present exploration is the discussion of stable and creative aspects of linguistic production, a discussion which has been undertaken before in SLA in terms of the antagonistic pair ‘formulaic expressions’ and ‘creative grammar’ (e.g. Ellis 2012; Myles and colleagues 1998, 1999; Nattinger and De Carrico 1992; Pawley and Syder 1983; Weinert 1995). Although nativist accounts of language acquisition have largely ignored formulaic chunks on the grounds that they are peripheral to the developing generative capacity (e.g. Pinker 1999), the discussions in non-nativist work are not so much about whether or not children rely on chunks in acquisition – this is an empirical finding for both L1 and L2 studies going back to at least Clark (1974) and Hakuta (1974) – but about whether these chunks are psycholinguistically different from the rest of the emergent grammar, and the extent to which they drive language learning. Instead of compartmentalizing language into what is ‘formulaic’ and what is ‘creative’, UBL assumes that all linguistic units are fundamentally identical, psycholinguistically (e.g. Croft and Cruse 2004; Goldberg 2003). What is formulaic and what is not is, in UBL, a matter of degree of abstraction and creativity rather than a question of eitheror. Empirical L2 research in this vein has accounted for the ways in which such chunks feed into the rest of the emergent L2 on the basis of frequency (e.g. Ellis 2002; Ellis and Ferreira-Junior 2009a, b; Eskildsen 2012b; Eskildsen and Cadierno 2007), the learner’s ability to break down chunks and derive the schematic characteristics that allow for the insertion of new lexical material in the created slots and form the backbone of schematization (Eskildsen 2009, 2014, in press; for L1, see Lieven, Salomo, and Tomasello 2009), as well as the role of interactional context in the process (Eskildsen 2011, 2012a, 2015, in press; Eskildsen, Cadierno, and Li 2015). In order not to make a sharp distinction between “formulaic chunks” and the rest of the linguistic inventory, I use the term “recurring multi-word expression” (MWE).



The emergence of creativity in L2 English: A usage-based case-study 

 285

As pointed out by Ellis (2002) the process of constructing a linguistic inventory is something learners actively do as they are figuring out the language, but showing this empirically in interactional data is difficult. In this chapter, I draw on insights from research on language play (Crystal 1998) to find evidence for the on-going process of expansion of patterns thought to be central in the schematization process. Language play, i.e. interactional practices in which language is being played with or is being used to make playful comments, may be crucial in the very process where learners are breaking down and manipulating MWEs (Carter 2004; Bushnell 2009; Bell 2012a, b), partly because it increases awareness of the manipulable aspects of linguistic constructions (Cook 1997; Cekaite and Aronsson 2005), often evidenced in private speech (Lantolf 1997; Ohta 2001). Although there are resemblances, the concepts of language play and private speech are different from language-related episodes in which learners engage in focus on form by talking about the language (Swain and Lapkin 1998). Such episodes, of course, may be situated in language play, but the main vehicle of language play and private speech is not meta-talk, nor is the main function to be explicitly focusing on form. Rather, language play is less formalized and a less restricted category of language use and may include many kinds of interaction, including joking. As such, engaging in language play adds to people’s interactional competence because it allows them to participate in a more varied set of communicative environments (Broner and Tarone 2001; Pomerantz and Bell 2007), which again – from a usage-based perspective – yields further opportunities for language learning (Waring 2013). The chapter is organized as follows: After an introduction to the database, I outline the traceback method. I then present the results from the traceback procedure in order to show what emergent L2 creativity looks like from a usage-based perspective. I then take a closer look at some instances from the data where language play seems to have played an important role in the development of creative linguistic constructions. In this part of the exploration I follow Waring (2013) and Huth (this volume) and draw on conversation analysis. Due to space considerations I will limit myself to one construction, centered on the verb need, in this part. It is my hope that the insights from the analyses of the instances of this construction will instigate a more principled method of investigating the role of language play as a social activity in the process of long-term L2 construction learning.

3 Data The data source for the present study is the Multimedia Adult English Learner Corpus (MAELC), which consists of audio-visual recordings of classroom interaction in an English as a Second Language (ESL) context. The classrooms, in which

286 

 Søren W. Eskildsen

the recordings were made, were equipped with ceiling-mounted video cameras and microphones, and students wore wireless microphones on a rotational basis; the teacher also wore a microphone (Reder, Harris, and Setzler 2003; Reder 2005). Consisting of recordings from September 2001 through June 2002, this is a longitudinal study of Carlos (pseudonym), an adult Mexican-Spanish speaking male learner of English, who was judged to be a successful learner (by standardized assessments and progress through the language school program). The final database of the inquiry consists of transcripts from sessions in which Carlos is either wearing a microphone or sitting next to someone wearing a microphone. In addition, there are transcripts from other sessions where Carlos’ talk has been picked up by the fixed ceiling-mounted microphones (cf. Eskildsen 2015).

4 Method 4.1 Traceback procedure in child language studies Methodologically, the modus operandi is inspired by Lieven et al. (2003), Dabrowska and Lieven (2005), Lieven, Salomo, and Tomasello (2009), and Vogt and Lieven (2010) in which the researchers reported on usage-based investigations of syntactic creativity in early child language. Referring to their method as ‘traceback’, the researchers used densely sampled longitudinal data sets of recordings of four two-year-olds to investigate how closely related the children’s utterances at the end of the recording period were to their previous utterances. Their basic finding for the four children was that 78–92% of their target utterances (utterances at the end of the recording period) were either verbatim repetitions or traceable to previous utterances by recourse to one specific syntactic operation, called substitution. If for example a target utterance was I got the butter and the closest match was recognized as I got the door, the syntactic operation identified to be required for producing the novel utterance was substitution (of the object). The underlying construction sanctioning this operation was thought of as consisting of a fixed part and a flexible part: I got the + an open slot for the insertion of lexically apt material. The majority of slots identified in such partially fixed, partially open constructions were for referents, typically nouns or pronouns, or – especially with longer utterances – slots for processes, typically verbs. The authors thus found a high degree of linguistic routinisation in their focal children and further concluded for the longitudinal data that tracing becomes increasingly difficult with increasing utterance complexity and mean length of utterance. These traceback studies thus confirmed the results from



The emergence of creativity in L2 English: A usage-based case-study 

 287

MacWhinney (1975), the primordial empirical work showing that children’s evolving grammars are item-based.

4.2 Traceback in the present study The methodology has been modified to fit the present data because these data were collected over a longer time span (Eskildsen 2014). I divided my focal student’s time in the ESL classroom into five recording periods (henceforth RPs) that constitute the traceable databases, the main corpora. At the end of each period I made a baseline point consisting of three to six classroom sessions from which I drew the target utterances (i.e. those to be traced), the test corpora. The data collection periods are outlined in Tab. 1. The reason for the difference in amounts of sessions (dates) for collecting the test corpora is that I aimed for roughly the same number of target multi-word utterances (MWUs) to be traced. Obviously, the student employed single word utterances (yes, no, exactly, alright etc.) and tokens of recipiency and change-of-state (Heritage 1984) (uhuh, mhm, oh etc.) which have been left out here because the focus of the investigation is on MWUs. Utterance end-points were defined interactionally, i.e. as points of speaker-transition where the next speaker takes the floor. Utterances with embodied completions1 (Olsher 2004) were also included. Tab. 1: Data collection periods, main corpora and test corpora Period

Main corpus

Test corpus

RP1 RP2–3 RP4 RP5

September 27 – November 8, 2001 January 8 – June 7, 2002 Sep. 23, 2003 – Mar. 12, 2004 Sep. 30, 2004 – Mar. 3, 2005

November 19, 26 and 29, 2001 May 17, 21, 24, 28 and June 7, 2002 February 10, 24, 28 and March 2 and 12 February 1, 15, 17, 22 and March 1 and 3

I then took the MWUs identified at each baseline point and traced them backwards in development in order to discern the interplay between routines and creativity in L2 development. All MWUs in the test corpora were thus traced to identify their closest matching utterances. A closest match was either the exact same string – an MWE – or a string that differed from the target utterance in

1 An embodied completion is an turn-at-talk ending in a bodily action, e.g. he went to the [deictic gesture].

288 

 Søren W. Eskildsen

one or more slots in the construction, which was then labelled an utterance schema; a mixed construction, consisting of a fixed and a variable part. The variable part is an open slot for the insertion of semantically relevant material and can be established if slots are filled by semantically similar slots in the schema over time (Lieven et al. 2009). The more slots that need filling to arrive at a target utterance from the closest match, the more creative the construction can be said to be. Once the MWUs to be traced had been identified in the test corpora, the first step in the tracing process, following Lieven, Salomo, and Tomasello (2009), was to identify recurring chunks, i.e. fixed MWEs, in the data and then to identify and count the utterances that varied in one or more pattern slots from their closest matches and to categorize the slots semantically. Whereas only verbatim repetitions count as recurring fixed MWEs, schemas are less fixed constructions (i.e. they have open slots) but still share lexical material with the closest matches. Fixed strings and schemas with open slots are the component units available to the child language user; the present investigation brings this concept of linguistic knowledge to bear on adult L2 learning.

5 Results This section presents the results of the traceback procedure in Carlos’ first year in class, from September 27, 2001 thru June 7, 2002. In a recent publication (Eskildsen 2015) this period was divided into three separate recording periods, hence I refer to this period as recording periods 1, 2, and 3. A total of 274 intelligible utterances were traced in RP1 and 295 in RPs 2 + 3. The first step in the tracing procedure is to match the target utterances with previous utterances. To fully comprehend the data it was necessary to make an initial distinction between simple and complex traces. Simple traces were operationalized as utterances that could be traced to one previously occurring singular utterance, either a fixed string, i.e. a MWE, or a schema with one or more open slots, i.e. an utterance schema. It was necessary to carry out this preliminary tracing for two reasons: 1) not all utterances could be traced in a simple straightforward manner to either an utterance schema or a MWE; and 2) utterances that could not be traced in such a straightforward manner could still be traced to some previously used material. In the latter cases, however, the tracing had to be carried out two, three or four times, depending on the number of utterance schemas the target utterance was traced to. There are no untraceable utterances in RP1 (cf. Eskildsen 2014) and only one in RPs 2 + 3.

The emergence of creativity in L2 English: A usage-based case-study 



 289

All other items in Carlos’ early linguistic inventory can be described with recourse to a relatively simple and highly concrete grammar of recycled patterns, as will be shown. Tab. 2: Simple and complex traces, RPs 1, 2, 3 Simple trace (trace to an utterance schema, MWE or affordance)

RP1

Complex trace (trace to several sources, either multiple schemas, multiple MWEs, (a) schema(s) and a MWE(s) in combination, or a schema(s) or (a) MWE(s) in combination with recycled single items)

RPs 2 + 3

RP1

RPs 2 + 3

Type

Absolute/relative no.

Type

Absolute/relative no.

MWEs

87 31.8%

109 36.9%

7 2.6%

Affordances

54 19.7%

33 11.2%

One utterance schema + MWE(s)/ recycled items

Full verbatim repetitions, total:

141 51.5%

142 48.1%

One utterance schema

102 37.2%

114 38.6%

Two utterance schemas

14 5.1%

9 3.1%

Three or more 10 utterance schemas + 3.6% MWE(s)

15 5.1%

Untraceable

0

1 0.3%

Total:

31 11.3%

39 13.3%

Total:

243 88.7%

256 86.7%

4 1.4% 10 3.4%

As Tab. 2 indicates, one-to-one traces and simple traces far outnumber complex traces; in RP1, 243 out of 274 target utterances (88.7%) could be traced to either a MWE, an affordance or an utterance schema, whereas 31 utterances (11.3%) could not be traced to one specific previous occurrence but needed to be traced to more than one match. For RPs 2 + 3, the numbers are slightly different with 86.7% constituting simple traces and 13.3% complex traces. This is in itself a remarkable finding that supports a notion of language as a very tangible and concrete tool for communication (cf. MacWhinney 1975), that language in the

290 

 Søren W. Eskildsen

learning process is well thought of as a new machine built out of old parts (Bates and MacWhinney 1988), and that the emergence of creative language is crucially dependent on the recycling of something that has been said in identical or similar ways before (Hopper 1998) – in this case by the language learner or his interlocutors in the on-going interaction. Language emergence is, in other words, usage-based. This preliminary count also concerns the ratio of repetitions to novel constructions. For RP1 and RPs 2 + 3, respectively, 47.2% and 48.1% of Carlos’ utterances were full verbatim repetitions. These are the MWEs and the affordances in Tab. 2. Whereas MWEs have been said by Carlos in their entirety before, affordances are found in the immediate interaction leading up to Carlos’ use by either a peer or a student, or they had been introduced by the teacher as part of the current task and/or written on the classroom whiteboard. These kinds of uses are referred to here as examples of affordances with a term borrowed from van Lier (2000). Afforded uses may range from short expressions like hot tea to longer ones such as do you like ice cream, where did you go and but not this time. Another kind of affordance comes about as private speech (Lantolf 2000). These are utterances that derive from the environment but which Carlos repeats to himself in lower voice. Such private speech, it has been argued in socio-cultural theory, is crucial in internalization processes in L2 development (e.g. Lantolf 2006, Ohta 2001) but may also have a relation to language play (Lantolf 1997) to be explored in Section 6 below. The remaining portion of utterances in the test corpora could not be traced to exactly identical utterances; i.e. they were novel and thus creative. These are the utterances traced to one utterance schema and all the complex traces. As Tab. 2 reveals, such utterances constitute just over 50% of all Carlos’ traced utterances, and the lion’s share of these novel, creative utterances consisted of one utterance schema; i.e. utterances that were traceable to previous uses by way of one simple syntactic operation, substitution, or in a few cases, two or three substitutions. In the case of the former, the target utterances differed from previous uses in only one open slot (e.g. are you sick? → are you sure?), and in the case of the latter, in two or three open slots (e.g. I no cook nothing → I no remember my zip code). The high percentage of utterances traceable to a particular utterance schema tallies well with previous usage-based findings that indicated the importance of utterance schemas in L2 development (Eskildsen 2009). Table 3 displays examples of utterances brought about by one, two and three substitutions, respectively. The tracing is done in the most simple way possible and the substitutions always operate on slots in constructions, i.e. it is possible for a string of elements to substitute another if it recurs as a string in the data.



The emergence of creativity in L2 English: A usage-based case-study 

 291

Tab. 3: Examples of utterances traced to one utterance schema and derived from the substitution operation. Substituted elements emphasised Utterance

Closest match

Substitutions

Number of examples, RP1

Let me check I have only twenty-six I have good memory

Let me see We have only one class We have the same teacher

1 (see → check) 64 2 (we → I; one 28 class → twenty-six) 3 (We → I; the 3 same → good; teacher → memory)

Number of examples, RPs 2+3 76 28 10

Having traced and outlined the basic constructional material of Carlos’ utterances during the first year in class, the next step was to investigate the nature of the open slots in the creative utterances. The slots identified in RPs 1, 2, and 3 are displayed in Tab. 4 and exemplified. Tab. 4: Types of slots in the recurring constructions in the data Semantic type of slot

Example utterances

Schema with slot(s)

Referent (REF)

I have only 26 We have only one class You no have address? You no like ice cream? Are you sick? Are you sure? You think so much I worked too much It’s on the 36 picture It’s in the first avenue I’m writing I was dancing More difficult More time / more + gesture The picture My picture How is your son? How are you?

REF have only REF

Process (PRO) Attribute (ATTadj) Attribute (ATTadv) Location (LOC) Progressive (PROing) Utterance (UTT) Determiner (DET) Copula (COP)

You no PRO REF? Are you ATTadj REF PRO ATTadv It’s LOC I am/was PRO-ing More UTT DET picture How COP REF

The slots are the same as in Lieven, Salomo, and Tomasello (2009) with four additional slots, adverbs as verbal attributes (ATTadv), the progressive verb form

292 

 Søren W. Eskildsen

(PROing), a determiner slot (DET), and copula (COP). Whereas PROing and DET are straightforwardly derived from the data and coded, ATTadv is more taxonomically challenging. In Lieven et al.’s work adverbials were usually seen as concrete lexical items (e.g. here) being added on to existing utterance schemas, but in my data there is too much variety in adverbial use to be captured by a simple syntactic operation based on the occurrence of specific lexical items. The difference between ATTadv and LOC is not always clear-cut, as LOCs typically have adverbial function, and thus single-word adverbs of place (e.g. here) maybe coded as either LOC or ATTadv depending on the actual utterance; e.g. Carlos’ utterances it’s here and it’s in the first avenue are seen as instantiations of it’s LOC, whereas only sometimes and only here are seen as instantiations of sometimes ATTadv. This dual coding, as it were, is in perfect alignment with the usage-based dictum that linguistic items may be stored simultaneously at different levels of complexity, abstraction and formulaicity in the brain (e.g. Achard 2007). Apart from UTT all slots must have shared semantic features in their instantiations. This also means that the UTT category applies in cases where Carlos seems to be working on an item-based pattern (e.g. more UTT) but where the open slot may be occupied by material that does not necessarily share semantic features. In the case of more UTT, the utterance more difficult? could be traced to more time and more + embodied completion which indicates that the open slot is, in fact, of a quite versatile nature, making it reasonable to assume that for Carlos it is so, too. The reasoning behind positing an UTT slot is thus very much data-driven; the empirical data simply lead me to suggest that Carlos is working on a more UTTschema. Another such example is the schema you no like UTT as the utterance you no like play football could be traced to you no like ice cream. The relationship between utterances was investigated to the furthest extent possible on the basis of one simple syntactic operation, namely substitution.2 The ADD-operation, as used by Lieven and her colleagues, plays a very minor role, in their child language data and here; an operation was identified as ADD if one item, typically vocatives or adverbials, had been added to either end of a recurring string or schema – and only if this dual placement was in principle possible (Vogt and Lieven 2010). In my data, vocatives – e.g. when Carlos summons the teacher or other students – do a distinct job of getting the attention of the person summoned and do not seem to be a part of the following construction. They are typically uttered in a way that sets them apart prosodically from what follows, and they

2  Dabrowska and Lieven (2005) also used the operation superimpose to account for cases where the substitution between the target utterance and the closest match shares lexical material and thus fills more than the open slot. I use the operation substitution for cases like this.



The emergence of creativity in L2 English: A usage-based case-study 

 293

are predominantly used as first pair parts in adjacency pairs requiring a response, oftentimes carried out as an embodied action, e.g. a shift in posture and / or eye gaze on the part of the recipient. Therefore, they are treated here as one-word utterances and not counted for the present purposes. Adverbials, on the other hand, abound – they may be single words (e.g. here and now) or strings or schemas in their own right (for me/you, in Portland and in the REF) and do not seem to be merely ‘added on’ to existing patterns. Instead, as indicated by the existence of the slot ATTadv, when such adverbials are used they are typically integrated into the schema (e.g. I spell for you traced to (can) you spell for me; only here traced to only sometimes). The only case that seems befitting of the ADD operation in RP1 is can you write for me please (see Tab. 3) in which please seems to be added on to an existing MWE. Moreover, Carlos uses please in both turn-initial and turn-final position, which further establishes it as an ‘addable’ item as mentioned above. As shown until now, simple operations can quite straightforwardly account for the vast majority of the utterances made by Carlos in his first year in class – either they were verbatim repetitions of novel utterances brought about by substitution of, in most cases, one item in an open slot in an utterance schema. The challenge still remains to deal with cases where multiple schemas need to be invoked to arrive at the target utterances. In total, there are 31 such utterances in Carlos’ RP1 and 39 in RPs 2 + 3. In Eskildsen (2014) they were grouped into subtypes on the basis of the tracing. A basic distinction was made between utterances that consisted of one utterance schema in combination with a variety of recycled material and utterances that consisted of multiple utterance schemas. Examples of the former share the feature of being a partial or full repetition of something Carlos has said before with the addition of other material in a way that has not been done by Carlos before. So although the utterances themselves consist of recycled and thus traceable material, they are novel and creative uses. There were 8 and 13 such uses in RP1 and RPs 2 + 3, respectively. Table 5 gives two examples from RP1. Tab. 5: Partial or full repetitions of something Carlos has said before with the addition of other material in a way that has not been done by Carlos before Utterance

Closest matches

Type

MWEs / Schemas

Operations

1

Can you write for me please

1 MWE + 1 recycled item

In the night for the dance

Can you write for me Please In the night For REF

ADD (please)

2

Can you write for me Please In the night For the write

1 MWE + 1 utterance schema

Repetition + 1 substitution (write - dance)

294 

 Søren W. Eskildsen

The first example, can you write for me please consists of a MWE, can you write for me, with the addition of please, an item Carlos has used once before, in a situation where it was prompted by the teacher. The target expression may thus be traced to two previous occasions of use, one that is the prior use of the MWE and one that is the prior use of please. The second example is a MWE and an utterance schema in combination: in the night for the dance. The two components are not intertwined in this use; instead they seem to be merely produced sequentially as separate units: in the night + for the PRO (dance). The remaining complex traces consist of two or more utterance schemas; there were 19 examples in RP1 and 25 in RPs 2 + 3. The use of two or more schemas in one utterance implies the necessity of introducing additional kinds of operations; the simple and otherwise frequent substitution operation no longer suffices, although it is also being used for the production of these utterances, and the ADD operation is obviated. Instead it will be hypothesized that Carlos constructs novel utterances by embedding utterance schemas in each other. The operation of embedding is similar to what Vogt and Lieven (2010) called a two-step substitution and will now be explained in detail using the examples in Tab. 6, which displays two of 21 utterances consisting of two utterance schemas found in RP1 and RPs 2 + 3. Tab. 6: Utterances based on two utterance schemas Utterance

Closest matches

Schemas

Operations

1

It’s an action right now

UTT right now It’s REF

2 substitutions (it’s an, action) + embedding

2

Oh for the next year

He’s not here right now It’s a woman For the test Next week

For UTT Next REF

2 substitutions (next year, year) + embedding

Example 1, it’s an action right now, is the result of two substitutions and ­embedding. When it is traced, two matches appear, he’s not here right now and it’s a woman3. It’s an action right now and he’s not here right now share the lexical feature of having right now as an utterance-final temporal adverbial. There is one more example of this use of right now in Carlos’ RP1, two teachers right now, which leads to the positing of an utterance schema in Carlos’ ­inventory denoted as “UTT right now”.

3 Please note that it’s an action and he’s not here could be thought of as one schema based on the copula, but the pattern-based approach to tracing development reveals that the negated variety of the former pattern does not emerge until later in Carlos’ development.



The emergence of creativity in L2 English: A usage-based case-study 

 295

This means that what takes the place of UTT is substitutable, in this case he’s not here / two teachers was substituted with it’s an action. Instead of proposing that Carlos is working on an abstract copula representation, shared by the two examples, plus a structurally manipulable, or even removable, negation particle, which would ultimately imply that this was a much more abstract schema, cf. note above, I assume that it’s an action is derived from a previous use of the it’s REF schema, it’s a woman. This schema is, then, embedded in the “UTT right now”-schema. The notion of embedding is to be thought of as a metaphor for the process that allows Carlos to combine his way to it’s an action right now on the basis of the two previously used patterns identified as the closest matches. Schematically it looks like this: It’s an action right now: UTT right now ↕ It’s a REF ↕ action Using that template it is possible to derive the same process for for the next year and for other examples of the same kind: For the next year: For UTT       ↕ The next REF       ↕ year

I don’t know write snore: I don’t know UTT

Maybe only here: Maybe UTT

Maybe you go the level_: Maybe you UTT ↕       Go REF        ↕ The level_

     ↕

Only ATTadv

        ↕

here

        ↕

Write REF



     

↕

snore

296 

 Søren W. Eskildsen

In Eskildsen (2014) I argued that for the next year was brought about by ­substitutions and a process referred to as merge that seemed to be different from embedding. It was hypothesized that Carlos had internalized a for REF-schema rather than a for UTT-schema and a next REF-schema. I suggested that Carlos was merging these two existing utterance schemas. The difference between embedding and merging as syntactic operations is that in the former, new linguistic items are straightforwardly incorporated into pattern slots but only by means of an existing utterance schema sanctioning the recruitment of another utterance schema. In it’s an action right now, for example, it’s REF is thought of as embedded in the UTT slot in the UTT right now schema. In the latter, i.e. when the operation merge was found to apply, two existing utterance schemas were analyzed as merging into each other to result in a new emergent schema: for REF and next REF merge to become for ATTadj REF. In the words of Ellis and Ferreira-Junior (2009a), the UTT right now schema is capable of recruiting the schema it’s REF in the UTT position; for REF, on the other hand, does not, in and of itself, allow for recruiting next REF, because there is no slot for ‘next’, and hence the merge operation was posited. Schematically, merging looked like this in Eskildsen (2014): For the next year: For the students → For REF  For ATTadj REF Next week → ATTadj REF However, as was shown above, if it is asserted that Carlos is working on a more generic for UTT-schema, then it is possible to account for the novel construction by recourse to substitution and embed alone. As the embed operation is essentially a substitution operation, i.e. an operation working on open slots in constructions, it would seem more empirically valid and thus more theoretically coherent than the merge operation and its principle of a seamless, almost magical combination of two existing constructions. It is possible to perform the same reanalysis of other examples of the same kind where the merge operation was posited in Eskildsen (2014): I’m very sad: I’m good → I’m ATTadj  I’m very ATTadj Very good → very ATTadj

=>

I‘m very sad: I’m UTT       ↕ Very ATTadj     ↕ sad



The emergence of creativity in L2 English: A usage-based case-study 

They live in different house: I live in Portland → I live LOC (= in REF)  REF live in ­different REF Different one → different REF

 297

They live in different house: REF live in UTT ↕ => ↕    I Different REF ↕ house

The complex traces accounted for so far have been found to consist of a variety of combinations of MWEs or affordances or other repeated items and a singular utterance schema or two utterance schemas. The nature of the operations needed to produce the utterances when compared with closest matches in the tracing procedure was most profitably analyzed in terms of embedding and substitution operations. Even in the case of these more complex traces, the constructional lineage of the utterances can be found in the data by way of the tracing methodology, showing an astounding degree of recycling in L2 learning and accounting for creativity – i.e. the ability to construct novel utterances – in terms of operations on the basis of open slots in constructions. Utterances that were found to consist of more than two schemas presented further challenges to the tracing methodology. Table 7 shows two examples of utterances that draw on three and four utterance schemas, respectively. Both are from Eskildsen (2014). Tab. 7: Utterances based on three or more utterance schemas. 1 I no wanna go the level B

I no remember I want correct Go the bank

3 schemas I no UTT I want UTT Go REF

2 You know if the people have one P or two P

You know what is ah 4 schemas address If you can read Three people They have vacation One pee or two pees

You know UTT If UTT DET people REF have REF

3 substitutions (wanna go) + (go) + (level b) + 2 embedding 5 substitutions (if the people have one P or two P) + (the people have one P or two P) + (the people, one p or two p) + (people) + 3 embedding

In Eskildsen (2014) I no wanna go the level b was analyzed as three merged ­utterance schemas: I no remember → I no PRO I want correct → I want PRO I no want PRO LOC (REF) Go the bank → go LOC (REF)

298 

 Søren W. Eskildsen

In this analysis the construction is a result of the posited merging of three schemas, but that analysis does not really explain how the three schemas would join like a zipper to become one. The explanatory power of the embed operation is much greater as it builds directly on usage-based assumptions of how the linguistic inventory of constructions emerges. By positing a more general UTT slot in the positions where a more specific semantic slot was proposed in the original analysis of I no wanna go the level b, it can be asserted that Carlos works on an I no UTT schema that sanctions the recruitment of the rest of the utterance, which then comes to be seen as embedded in the UTT position. The same procedure applies to the Want UTT schema: I no wanna go the level b: I no UTT        ↕ Want UTT       ↕ go LOC (REF)

The second utterance, you know if the people have one P or two P, is the most complex in terms of tracing in RP1. Consisting of four schemas and brought about by three embeddings, it may be represented schematically in the following way: You know if the people have one P or two P: You know UTT ↕ If UTT ↕ DET REF have REF ↕ ↕ The people one P or two P

In RP2 there were an additional three utterances that required tracing to four utterance schemas. One, can you write my name in Chinese but in the letters big, will serve as example. The basic schema is UTT but UTT. There are quite a few examples in the data where Carlos is using but as a conjunction between clauses and/or phrases. In this case, the first UTT slot is filled by can you write my name in



The emergence of creativity in L2 English: A usage-based case-study 

 299

Chinese¸ whose closest lexical match is can you you write her name in your [gest]. Two slots in this utterance schema has been substituted by her and Chinese, ­respectively. The second UTT slot is filled by another schema, in UTT whose UTT slot is filled by yet another schema, the non-standard DET REF ATTadj, lexically filled by the letters big. The entire utterance can be represented using the same template as above: Can you write my name in Chinese but in the letters big: UTT but UTT   ↕ ↕ Can you write DET name in REF in UTT ↕ ↕ ↕ Her Chinese DET REF ATTadj ↕ ↕ ↕ The letters big

So far, the application of the traceback methodology from child language studies to longitudinal adult L2 data has shown a case of early L2 learning as rooted in recurring MWEs and utterance schemas with creativity evolving in a slow and piecemeal fashion that in many ways resembles the processes involved in child language learning. The substitution operation, as was the case in the child language data, was found to be the most frequently employed operation to produce new, creative utterances. However, it was also necessary to introduce the embed operation for utterances that consisted of more than one utterance schema. The merge operation (Eskildsen 2014) can be done away with which makes for a more empirically valid, theoretically coherent, and powerful methodology to investigate longitudinal L2 learning research in terms of a usage-based, emergent inventory of lexically specific constructions. This results in a conceptualisation of emergent creativity as building on recycled linguistic matter in the form of MWEs and utterance schemas. Development is thus described and analyzed as the emergence of new MWEs and utterance schemas, some of which may be combined in different ways, with a slight tendency for the number of schematically sanctioned lexical options and intra-turn schematic operations to be on the increase. The next questions concern what tracing looks like when applied to data further down the developmental path and the extent to which the identified operations can account for more advanced L2 learning. As Tab. 8 shows, Carlos’ inventory changes after the first year in class. In the next two recording periods,

300 

 Søren W. Eskildsen

simple traces, while still accounting for the majority of his utterances, are on the decrease, while the use of utterances based on two or more utterance schemas is increasing. Tab. 8: simple and complex traces, RPs 4 and 5 Simple trace (trace to an utterance schema, MWE or affordance)

RP4

Complex trace (trace to several sources, either ­ ultiple schemas, multiple MWEs, (a) schema(s) m and a MWE(s) in combination, or a schema(s) or (a) MWE(s) in combination with recycled single items)

RP5

RP4

RP5

Type

Absolute/relative no.

Type

Absolute/relative no.

MWEs

61 23.6%

58 26.2%

1 MWE + recycled item / several MWEs

3 1.2%

4 1.8%

Affordances

24 9.3%

8 3.6%

9 3.5%

7 3.2%

Full verbatim repetitions, total:

85 32.9%

66 29.8%

One utterance schema + MWE(s) / recycled items

One utterance schema

90 34.9%

68 30.8%

Two utterance schemas

25 9.7%

27 12.2%

Three or more utterance schemas + MWE(s)

44 17%

46 20.8%

Untraceable

2 0.8%

3 1.4%

Total:

83 32.2%

87 39.4%

Total:

175 67.8%

134 60.6%

Especially affordances are on the decrease, while MWEs and utterances based on one utterance schema display less dramatic decreases in use. Fig. 1 ­captures this development while also showing the increase in the share of utterances based on multiple utterance schemas and the constant and low ratio of utterances based on multiple MWEs and other recycled items and utterances based on one utterance schema in combination with one or more MWEs and/or other ­recycled items. The data thus suggest that an increased creative ability rests on building an inventory of utterance schemas that may then be combined in new ways in usage events.

The emergence of creativity in L2 English: A usage-based case-study 



 301

45 40 35 30 25

RP1 RP2-3

20

RP4 RP5

15 10 5 0 mwes

affordances 1 utt schema 2 utt schemas

3+ utt schemas

mult mwe/rec 1 utt sch + items mwe(s)/rec items

Fig. 1: Overview of the data

Apart from the decrease of afforded uses, the most dramatic change from early to later RPs concerns the ratio of utterances based on two or more utterance schemas. These multiply approx. 4 times. As Carlos is developing his linguistic inventory he also becomes increasingly capable of sustaining longer t­ urns-at-talk consisting of an increasing number of utterance schemas. At more advanced levels in the classroom it becomes more relevant for example to engage in storytellings which again both displays and feeds into his developing interactional competence (Eskildsen in press). However, the tracing methodology still allows me to search out the roots of the components of these utterances, no matter how many utterance schemas enter into the construction of Carlos’ turns-attalk (Eskildsen forthc.). Moreover, only 12 utterances take the methodology into uncharted territory. These are the utterances in RPs 4–5 that consist of more than four utterance schemas; in RPs 1–3 there were none of these. They can be traced in the same manner as other complex traces in the early RPs, but they require more embed ­operations. In the interest of space, one utterance, based on eight utterance schemas, will serve as example.

302 

 Søren W. Eskildsen

I no decide yet but I I need go to Mexico for visit my family now because I don‘t see them for four years. UTT but UTT ↕ ↕ UTT yet UTT because UTT ↕ ↕ ↕ I no PRO UTT for UTT I don’t see REF UTT ↕ ↕ ↕ ↕ ↕ decide I need go to REF PRO REF now them for four years ↕ ↕ Mexico visit my family As one final point before looking closer at language play, utterances in RPs 4–5 that are traceable to one singular utterance schema are still predominantly brought about by one substitution operation and a few by two or three ­substitution operations, but in addition to these, it was necessary to posit four substitution operations in six cases, which is another minor way in which Carlos’ inventory and his abilities to bring about novel constructions are developing: the schemas are becoming increasingly abstract in their representation. Table 9 gives one example. Tab. 9: One-schema uttterances requiring four substitutions Utterance

Closest match

Substitutions

Number of examples, RP4

Number of examples, RP5

I saw that on the TV

She sees the wallet on the ground

4 (she → I; sees → saw; the wallet → that; the ground → TV)

2

4

6 Situated emergence of creativity – the case of language play In recent publications I have argued in favour of a locally contextualized approach to longitudinal L2 research and developed a method drawing on UBL and ­conversation analysis (CA) (Eskildsen 2009, 2011, 2012, 2015, in press; Eskildsen, Cadierno and Li 2015; Theodórsdóttir and Eskildsen forthc.). This research has shown how linguistic constructions emerge in an exemplar-based fashion in use and has elaborated on the specifics of the situations in which exemplars



The emergence of creativity in L2 English: A usage-based case-study 

 303

of the constructions are first put to use. In this section, again drawing on CA, I further substantiate these findings with examples where language play seems to have played an important role in L2 emergence. The particular construction investigated is based on the verb “need”. As mentioned above, this exploration may serve as exemplary for further usage- and CA-based studies investigating the role of language play in the emergence of L2 constructions. In the first example (Extract 1) we see Carlos using the expression you need the glass to make a playful comment on a fellow student’s (Martina) tripping over a cord in class. Prior to the extract the students have filled in a sheet with questions about themselves (their families, their homes, life in the US etc.). They were also instructed to write more questions than those already on the sheet. The next part of the task for the students is then to find a partner and ask him/her the same questions. Just before line 1 in Extract 1, Martina gets up and starts walking to Carlos’ desk. Extract 1, “you need the glass” (Oct. 4, 2001)4 01 MAR: trips over cord, shuts off OHP incidentally ↑↑ . huh: 02 CAR: ↑↑o : : h 03 MAR: (    ) UKR; fiddles with cord 04 TEA:   it’s okay↓ we don’t need it any more 05 MAR:   hehhehheh  . hh heh heh tsoh heh= 06 =heh*heh . hh *sits down at desk with Carlos 07 CAR: you need the glass heh heh 08 MAR:  yeahehhah * hah hah hah .hh hehhehhehheh=     *touches temples, as if indicating glasses 09 CAR: hehhehhehheh 10 MAR: =.hhh hehhehheh . hh hehheh ˥ ˩

˥ ˩

˩

˥ ˩

˥

In line 1, Martina then trips over the cord to the overhead projector. She makes an exclamation of surprise and Carlos reacts to the scenario with a similar sound (line 2). Martina then starts fiddling with the cord while speaking Ukrainian in overlap with which the teacher tells her not to worry about it (line 4). Martina then bursts out in laughter and sits down with Carlos (lines 5–6). Carlos then makes his remark you need the glass in response

4 The data can be viewed here: http://www.labschool.pdx.edu/Viewer/viewer.php?languageplay. Requires Internet Explorer.

304 

 Søren W. Eskildsen

to which Martina begins laughing again, perhaps with an agreement token before the laughter, and she then touches her temples as if indicating her missing glasses, which works as a further display of alignment with Carlos’ joke. They then both laugh (lines 8–10). This is Carlos’ second use of a you need-construction. The first one was you need put Monday and Tuesday directed at the teacher approx. 90 min. earlier. Seen from a constructional perspective this use is not only creative in the sense that Carlos was using it to make a playful comment, it is also creative in introducing a noun where there was a verb in the previous instantiation. Carlos’ need-inventory spreads from these you need-instantiations (you need put (2), you need the glass, and you need (the) paper (2)) to accommodate other pronouns (he, I, we), other nouns (help, umbrella), and other verbs (study, work) in the first month. There then seems to be a dawning recognition on Carlos’ part that something is missing from the instantiations with verbs – the infinitive-marker to – and this recognition can be traced to the following situation, recorded approx. one month after Extract 1. The teacher is writing examples of I need to VERB-constructions on the whiteboard, the first example being I need to study English. She is not doing this primarily to teach this construction but to lead the students’ attention to the notion of “place” and the WH-question marker that goes with it, namely “where” (at this point she has reviewed “who”, “what” and “when”). This becomes evident in lines 11–14. At first, though, the teacher introduces the I need-construction and begins writing it on the board (lines 1–2). Some students display attention to the teacher’s actions throughout (lines 4, 6, 8), but not Carlos. He is still busy taking notes of what the teacher has been saying so far. Extract 2: “I need to study English” (Nov. 8, 2001) 01 TEA: your friend says to you I need to study english↓ 02 (4.0) teacher is writing on the board 03 TEA: I nee:d↑ 04 MUL: (° °) 05 TEA: to::: writing 06 UNI: stu dy 07 TEA: study::= writing 08 UNI: =english 09 TEA: english↓ *writing   *carlos begins writing “I need study” 10 (5.0) 11 TEA: where↑ 12 (1.5) carlos looks up, then adds "to" between “need” and “study” ˥ ˩



The emergence of creativity in L2 English: A usage-based case-study 

 305

13 TEA: knocks three times on whiteboard where does the-this person go↓ 14 where↓ 15 THA: P C C 16 TEA: P C C↓ okay However, during the teacher’s writing at line 9, he begins a new line in his notebook where he writes I need study. At this point the teacher is moving on with what transpires as her present purpose, namely to get the students to attend to the notion of place and the interrogative pronoun “where”. At her first where (line 11) the teacher does not get a response, so, following a pause (line 12), she knocks three times on the whiteboard and rephrases her action as a compete “where”-question (line 13). During the pause at line 12, Carlos looks up at the whiteboard and, apparently noticing the to in the teacher’s writing on the board, he then refocuses his attention on his own writing, adding to between need and study. Meanwhile, the teacher elicits an acceptable answer to her “where”-question which closes down the sequence (lines 15–16). So Carlos is not concerned with the teacher’s agenda here; he is not orienting to “place” and “where” at this moment. Instead he is noting down what he presumably considers a useful expression, I need (to) study. From the perspective of sociocultural theory Carlos is here engaged in a form of l­ anguage play as he enables himself to look more closely at his own cognitive processes (Lantolf 1997) that become visible in his writing and his adding to to the construction after having finished it already. Carlos’ actions indicate three things: 1) he seemingly had not noticed the to before this moment, which explains the two prior instances of you need put; 2) the MWE you need put forms the exemplar-based backdrop against which he is making generalizations as he expands the pattern to I need study by way of two substitutions; and 3) there was a potential learning moment right there and then as he was attending to the details of the needconstruction. However, the need put-construction is less than permeable which is evident in what happens next (Extract 3). The teacher is referring to a situation “the other day” when one student had asked for help to send a postcard (lines 1–5). Extract 3 “I need put stamp” (Nov. 8, 2001) 01 TEA: the other day rica says↓ (2.2) I nee::d↑ (1.8) I nee:d to se:nd (1.8) a 02 postcard↓ I need to mail a letter (.) I need to send a postcard↓ that’s 03 what you said the other day↓ you said= 04 RIC: m: : nods 05 TEA:     ↑↑ooh please help me: I nee::d→ ˥ ˩

306 

 Søren W. Eskildsen

06 (1.8) 07 UNI: I need ( ) 08 (0.8) 09 CAR: .hh I need put stamp 10 TEA: stamps uhuh stamps so where do you go: when you need stamps↓ The teacher ends her turn at line 5 with a continuing intonation pattern indicating that a response is relevant. Following a pause, one student continues with a partially unintelligible turn and following another pause, Carlos volunteers his candidate response I need put stamp (lines 7–9). The teacher acknowledges this and uses it as a springboard to ask her main where-question, where do you go when you need stamps. They eventually agree that the post office is the place for this (not included in transcript). Thus, it is evident that although Carlos did have some time to focus his attention on the presence of to in the need-construction, it has not yet sedimented in his experience to an extent that it can help him break an existing, routinized pattern. This indicates the importance of frequency; in his experience, Carlos has not yet perceived the “to” often enough, whereas he has used the construction without the “to” more often, resulting in a need-inventory based on instances without “to”. This, in effect, is the frequency-biased build of the linguistic inventory in action (Ellis 2002; Eskildsen 2012). There are more opportunities for learning coming up, however, as the teacher continues writing examples on the board. The next two examples which Carlos repeats as private speech are I need to buy food and I need to wash my clothes. These examples of private speech as repetitions are identical to the empirical examples from L2 children shown in Saville-Troike (1988). Private speech can also be considered language play because its rehearsal-like aspects serve a cognitive function (Lantolf 1997) and allows the speaker to listen to, manipulate, and play with what is being spoken. Especially in the case of Carlos’ repetition of I need to buy food it becomes evident that not only is he engaged in private speech as language play, he uses the situation as a springboard to making a playful comment that, while also emerging as a laughable, displays his understanding of the linguistic expressions currently at the center of attention in the classroom. The interaction is shown in Extract 4 which starts in line 1 with Carlos’ reading of the teacher’s writing. Extract 4, “I need to buy food” (Nov. 8, 2001) 01 CAR: °I need to buy food° 02 (2.1) 03 UNI: restaurant↑



The emergence of creativity in L2 English: A usage-based case-study 

 307

04 THA: no= 05 UNI: =supermar ket 06 THA:            ?café? 07 CAR: supermarket nods 08 TEA: nods supermarket↑ 09 (0.8) 10 TEA: oka:y↑ 11 (4.9) TEA writes “supermarket” on board 12 CAR: oh in the fred meyer 13 TEA:   ?oh but? here we go ehehhehh fred meyer’s okay↓ 14 CAR:      hhh heh heh heh heh HEH HEH heh heh 15 heh heh .hh ˥ ˩

˥ ˩

˥ ˩

At this stage the students are familiar with the purpose of the teacher’s examples and they start proposing candidates for where to go when in need of buying food. The teacher eventually accepts supermarket (lines 3–10). After a pause, Carlos elaborates on the supermarket by suggesting a specific local store fred meyer (line 12). This happens in partial overlap with the teacher’s next action (line 13) but she abandons and instead shows alignment with Carlos’ remark by laughing and repeating (perhaps doing an embedded repair) fred meyer’s. During this turn Carlos is laughing but his laughter becomes louder after the teacher has begun laughing (line 14) indicating the social nature of laughables. In the examples shown here Carlos has had the opportunity to focus his attention on a particular linguistic construction through note-taking and private speech, both considered a form of language play in a Vygotskian framework (Lantolf 1997). It was also shown here that Carlos is producing laughables by using a variety of the same construction (Extract 1) or by building on discourse in which a variety of the same construction plays a crucial role (Extract 4). The language play Carlos engages in thus serves two purposes that are conducive to L2 learning: it enables him to use language for a real, social purpose and it allows him to focus his attention on the language he is using. As such, language play should arguably be thought of as a more serious matter in L2 learning than its name implies. It would seem that the to-infinitive marker is beginning to sediment in Carlos’ experience in the time following the situations above. This is indicated in the next extract (5), recorded half an hour later, in which the class are doing a cartoon description exercise. The teacher is showing a strip where a man is running to catch a bus. Prior to the extract the class have agreed that “wait for me” is a good expression for what the man might have yelled out to the bus driver. Then Carlos adds I need a take the bus (line 1).

308 

 Søren W. Eskildsen

Extract 5, “I need a take this bus” (Nov. 8, 2001) 01 CAR: I need a take the bus↓ 02 TEA: e:hheh heh 03 CAR:       I go home↓ 04 TEA: uhhehwaheh .hh ↑↑*wa::it for* me:: 05 (2.8) 06 CAR: I need a- I need a take this bus 07 TEA: I need to take this bus↓ (.) okay wait for me↓ I need to take this bus↓ ↑↑okay↓ 08 ˥ ˩

In response the teacher laughs (line 2) and in overlap Carlos continues I go home (line 3). This also makes the teacher laugh but in-between laughter tokens she also utters wa, suggesting that she is still attending to the expression wait for me. This becomes clear as she finishes her turn by repeating wait for me in full and in a distorted voice (line 4). Following a pause Carlos repeats a slightly modifed version of his first contribution, I need a take this bus which gets a positive acknowledgment from the teacher (line 7). The interesting thing to note here, of course, is the little a between need and take both in line 1 and line 6. It is transcribed as “a” although “uh” would probably be a closer rendition of what he is saying, but because “uh” usually indicates a speech perturbation I have opted for “a”. In any case it ­indicates that Carlos is now clearly using a lexical element of some sort, perhaps a reduced form of to, in the position where to is supposed to be. This development, traceable to the situations in the extracts above5, continues with the next two examples in his data containing what sounds like the same element (you need a put the paper…, and she need a buy new shoes). However, Carlos’ need-inventory is not fully grown yet, as it were. In the period that follows there are other instantiations without the infinitive-marker, so it seems that there is some competition between the two variants following the episodes displayed here. It is not until two years later that Carlos begins using the construction with to ­consistently.

7 Summary and conclusions The tracing of the emergence of utterance schemas and more abstract constructions has shed new light on how L2 learning takes place along a usage-based path

5  In addition to the extracts in this chapter, there is an instance approx. 33 mins. prior to Extract 5 where Carlos reads “to take the bus” from the teacher‘s writing on the board. The instance of the need-construction in Extract 5, then, has a lexically specific origin, too.



The emergence of creativity in L2 English: A usage-based case-study 

 309

of routinisation and increased creativity. Routines are both of a highly frequent kind, recurring expressions that become entrenched as such because they are widely applicable, as well as of a less frequent kind, transitory MWEs that lose their relevance when not brought about by specific interactional circumstances (Eskildsen 2009, 2011). But the notion of routine itself is also gradable; routines may be partially lexically specific items, utterance schemas, that consist of a fixed and a creative part. This study has suggested that together, these three aspects of the linguistic inventory, the frequent and not-so-frequent formulas and the utterance schemas, constitute the bulk of individual portability in L2 learning. The implication of this is that MWEs and utterance schemas seem to be at the very heart of language learning; it is now possible with technologically advanced databases and the traceback methodology to demonstrate that abstract rules of syntax do not seem to be what people operate on to produce creative language. Rather, creativity is a matter of de-routinizing routines, i.e. breaking up fixed chunks into semi-fixed utterance schemas, and language play may contribute to these processes. But it is probably also a matter of building increasingly versatile utterance schemas; while it has not been thoroughly investigated here and thus remains a point for future research, it would seem that schemas with UTT-slots, which allow for greater constructional variance than, say, the more restricted REF- and PRO-slots, become increasingly important over time. In both respects, the usage-based conception of language knowledge – an inventory of constructions along a continuum of formulaicity-schematicity – is particularly apt at capturing the interrelationship between routines and creativity in investigations of how people experientially derive the schematic properties of recurring MWEs to enable them to produce novel, i.e. creative language. Such a view of language knowledge – and the notion that language use is its necessary prerequisite – makes for a very concrete and usage-rooted concept of our linguistic abilities and eliminates the validity of the Chomskyan argument from the poverty of the stimulus (MacWhinney 2004; Mellow 2006); language is learnable as an experiential interplay of abstracted utterance schemas, based on exemplar MWEs and single lexical items encountered in local environments. Nick Ellis’ (2002) suggestion that the UBL exemplar-based path of language learning might serve well as a default investigative line of enquiry into questions of L2 development has been receiving some attention (Ellis and Ferreira-Junior 2009a; Eskildsen inter alia 2009, 2014, 2015; Eskildsen, Cadierno, and Li 2015; Li, Eskildsen, and Cadierno 2014; Roehr-Brackin 2014) but much more research is needed to bring about evidence of language learning as a trajectory from formula to full schematicity; it may not always be that clear-cut, but that does not undermine the point of using the proposed trajectory as a guideline (Eskildsen 2009). Language learning is probably too unpredictable and messy to allow for brief definitions, but the conceptualization of the emergent properties of language as the linking of

310 

 Søren W. Eskildsen

singular expressions as utterance schemas, and routinization as entrenchment of such singular expressions as MWEs seems to be an empirically valid frame within which to investigate the acquisition of L2 material. The focus here has been on an individual, portable linguistic inventory, the questions asked until now pertaining primarily to notions of such an individual inventory. In this respect, the epistemological raison d’être of the chapter has been placed in a long history of SLA research whose focus rests with issues relating to individual cognition (cf. e.g. Doughty and Long 2003). In fact, even studies of the development of a communicative competence (Canale and Swain 1980), including Schmidt’s famous case study of Wes (Schmidt 1983), which typically set out to investigate a broader notion of language abilities than implied by traditional (Chomskyan) notions of a fairly restricted linguistic competence, have ultimately investigated and been interested in an individualized competence (Hall and Pekarek Doehler 2011). Recently, however, the field of L2 studies has seen an interest in competing views of cognition and linguistic abilities framed in more social terms. Cognition and linguistic abilities are argued to be distributed across participants and form part of larger structures of social interaction (e.g. Firth and Wagner 1997; Hall, Hellermann, and Pekarek Doehler 2011; Kasper 2009; Markee 2000; Young 2008). The implication of such socially anchored views of language and cognition for language learning is that everything speakers do is inherently and fundamentally contextual; nothing can be defined in a priori individualized terms. So far, the data has presented considerable evidence that Carlos is, in fact, drawing on portable resources that can be called his own; if not, we probably could not trace his utterances to h is own previous uses. Of course, how something became his own is a different matter; it could easily be that most of what he does is an appropriation of what he has participated in (Eskildsen and Wagner 2015), which would render his individualized inventory a socially shared resource. In fact, precisely this relationship between Carlos and the contexts in which he navigates was investigated through the concept of language play and showed the interactionally rooted and locally contextualized development of Carlos’ inventory; by engaging in language play he was appropriating others’ actions for his own interactional purposes. His portable language had a social source; and so his portable language can become somebody else’s source of learning next. What needs more attention in the future is the question of what other interactional resources Carlos is learning as part of a developing interactional competence (Pekarek Doehler and Pochon-Berger 2015). This will cast further light on changes in Carlos’ abilities to perform certain actions in interaction, such as doing co-constructions, carrying out repair, initiating turns, initiating actions such as story-telling, employing gestures, making playful comments etc. These are issues currently being dealt with (Eskildsen in press).

The emergence of creativity in L2 English: A usage-based case-study 



 311

Transcription conventions CAR: TEA: Participants wei rd wo˥ rd Beginning and end of overlapping talk yeah ˩ * Marks beginning of gesture *Word word Description of gesture. In cases with no *, the gesture follows preceding word. (1.0) Pause/gap in seconds and tenth of seconds Micro pause (< 0.2 seconds) (.) word= =word Multi-line turn word Prosodic emphasis wo:rd Prolongation of preceding sound Falling, rising, continuing intonation word↑,↓,→ Shift to high pitch ↑↑word Louder than sorrounding talk WORD Softer than surrounding talk °word° wo- Cut-off (e.g. glottal stop) Non-audible speech ( ) (?word?) Uncertain transcription .hh Hearable in-breath ˥ ˩

References Achard, Michel. 2007. Usage-based semantics: Meaning and distribution of three French ‘breaking’ verbs. In Marja Nenonen & Sinikka Niemi (eds.), Collocations and idioms 1: Papers from the first Nordic conference on syntactic freezes, Joensuu, May 19–20, 2007 (Studies in Languages, University of Joensuu, Vol. 41), 1–13. Joensuu: Joensuu University Press. Ambridge, Ben & Elena Lieven. 2011. Child language acquisition: Contrasting theoretical approaches. Cambridge: Cambridge University Press. Bartning, Inge & Bjørn Hammarberg. 2007. The functions of a high-frequency collocation in native and learner discourse: The case of French c’est and Swedish det är. International Review of Applied Linguistics 45(1). 1–43. Bates, Elizabeth & Brian MacWhinney. 1988. What is functionalism? Papers and Reports on Child Language Development 27. 137–152. Bell, Nancy. 2012a. Formulaic language, creativitity, and language play in a second language. Annual Review of Applied Linguistics 32. 189–205. Bell, Nancy. 2012b. Comparing playful and nonplayful indicental attention to form. Language Learning 62(1). 236–265.

312 

 Søren W. Eskildsen

Brandt, Silke, Arie Verhagen, Elena Lieven & Michael Tomasello. 2011. German children’s productivity with simple transitive and complement-clause constructions: Testing the effects of frequency and variability. Cognitive Linguistics 22(2). 325–357. Broner, Maggie A. & Elaine E. Tarone. 2001. Is it fun? Language play in a fifth-grade Spanish immersion classroom. The Modern Language Journal 85(3). 363–79. Brouwer, Catherine. E. 2003. Word searches in NS-NNS interaction: Opportunities for learning? The Modern Language Journal 87(4). 534–545. Bushnell, Cade. 2008. ‘Lego my keego!’: An analysis of language play in a beginning Japanese as a foreign language classroom. Applied Linguistics 30(1). 49–69. Bybee, Joan. 2008. Usage-based grammar and second language acquisition. In P. Robinson & N. C. Ellis (eds.), Handbook of cognitive linguistics and second language acquisition, 216–236. New York: Routledge. Canale, Michael & Merrill Swain. 1980. Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics 1(1). 1–47. Cadierno, Teresa & Søren W. Eskildsen (eds.). 2015. Usage-based perspectives on second language learning. Berlin: De Gruyter Mouton. Carter, Ronald. 2004. Language and creativity: The art of common talk. London: Routledge. Cekaite, Asta & Karin Aronsson. 2005. Language play, a collaborative resource in children’s L2 learning. Applied Linguistics 26(2). 169–91. Clark, Ruth. 1974. Performing without competence. Journal of Child Language 1. 1–10. Collins, Laura & Nick C. Ellis. (eds.). 2009. Input and second language acquisition: The roles of frequency, form, and function [special issue]. The Modern Language Journal 93(3). Cook, Guy. (1997). Language play, language learning. English Language Teaching Journal 51(3). 224–231. Croft, William & David A. Cruse 2004. Cognitive linguistics Cambridge: Cambridge University Press. Crystal, David. (1998). Language play. Chicago, IL: University of Chicago Press. Dabrowska, Ewa & Elena Lieven. 2005. Towards a lexically specific grammar of children’s question constructions. Cognitive Linguistics 16(3). 437–474. Doughty, Catherine E. & Michael Long. 2003. Handbook of second language acquisition. Malden, MA: Blackwell. Ellis, Nick C. 2002. Frequency effects in language processing – a review with implications for theories of implicit and explicit language acquisition. Studies in Second Language Acquisition 24(2). 143–188. Ellis, Nick C. 2012. Formulaic language and second language acquisition: Zipf and the phrasal Teddy bear. Annual Review of Applied Linguistics 32. 17–44. Ellis, Nick C. 2015. Cognitive and social aspects of learning from usage. In Teresa Cadierno & Søren W. Eskildsen (eds.), Usage-based perspectives on second language learning, 49–74. De Gruyter Mouton. Ellis, Nick C. & Fernando Ferreira-Junior. 2009a. Construction learning as a function of frequency, frequency distribution, and function. The Modern Language Journal 93(3). 370–385. Ellis, Nick C. & Fernando Ferreira-Junior. 2009b. Constructions and their acquisition. Islands and the distinctiveness of their occupancy. Annual Review of Cognitive Linguistics 7. 187–220.



The emergence of creativity in L2 English: A usage-based case-study 

 313

Ellis, Nick C. & Teresa Cadierno. 2009. Constructing a second language. Introduction to the special section. Annual Review of Cognitive Linguistics 7. 111–139. Ellis, Nick. C., Matthew B. O’Donnell & Ute Römer. 2013. Usage-based language: Investigating the latent structures that underpin acquisition. Language Learning 63 (Supp. 1). 25–51. Eskildsen, Søren W. 2009. Constructing another language – usage-based linguistics in second language acquisition. Applied Linguistics 30(3). 335–357. Eskildsen, Søren W. 2011. The L2 Inventory in Action: Conversation analysis and usage-based linguistics in SLA. In Gabriele Pallotti & Johannes Wagner (eds.), L2 learning as social practice: Conversation-analytic perspectives, 337–373. Honolulu, HI: University of Hawai‘i, National Foreign Language Resource Center. Eskildsen, Søren W. 2012a. Negation constructions at work. Language Learning 62(2). 335–372. Eskildsen, Søren W. 2012b. Type and token frequency in SLA. In Peter Robinson (ed), The Routledge encyclopedia of second language acquisition, 660–662. London and New York: Routledge. Eskildsen, Søren W. 2014. What’s new?: A usage-based classroom study of linguistic routines and creativity in L2 learning. International Review of Applied Linguistics 52(1). 1–30. Eskildsen, Søren W. 2015. What counts as a developmental sequence?: Exemplar-based L2 learning. Language Learning 65(1). 33–62. Eskildsen, Søren W. In press. L2 constructions and interactional competence: Subordination and coordination in English L2 learning. In Andrea Tyler & Carol Moder (eds.), What does cognitive linguistics look like? Answers from the L2 classroom and SLA studies. Berlin: De Gruyter Mouton. Eskildsen, Søren W. forthcoming. Usage-Based L2 learning: The construction of linguistic and interactional resources in a L2. New York: Routledge. Eskildsen, Søren W. & Teresa Cadierno. 2007. Are recurring multi-word expressions really syntactic freezes? Second language acquisition from the perspective of usage-based linguistics. In Marja Nenonen & Sinikka Niemi (eds.), Collocations and idioms 1: Papers from the First Nordic Conference on Syntactic Freezes, Joensuu, May 19–20, 2007 (Studies in Languages, University of Joensuu, Vol. 41), 86–99. Joensuu: Joensuu University Press. Eskildsen, Søren W. & Teresa Cadierno. 2015. Advancing usage-based approaches to L2 studies. In Teresa Cadierno & Søren W. Eskildsen (eds.), Usage-based perspectives on second language learning, 1–18. Berlin: De Gruyter Mouton. Eskildsen, Søren W. & Johannes Wagner. 2015. Emodied L2 construction learning. Language Learning 65, 2, 419–448. Eskildsen, Søren W., Teresa Cadierno & Peiwen Li. 2015. On the development of motion constructions in four learners of L2 English. In Teresa Cadierno & Søren W. Eskildsen (eds.), Usage-based perspectives on second language learning, 207–232. Berlin: De Gruyter Mouton. Goldberg, Adele. 2003. Constructions: A new theoretical approach to language. TRENDS in Cognitive Sciences 7(5). 219–224. Goldberg, Adele & Devin M. Casenhiser. 2008. Construction learning and second language acquisition. In Peter Robinson & Nick C. Ellis (eds.), Handbook of cognitive linguistics and second language acquisition, 197–215. New York: Routledge.

314 

 Søren W. Eskildsen

Gries, Stefan Th. & Stephanie Wulff. 2005. Do foreign language learners also have constructions? Evidence from priming, sorting, and corpora. Annual Review of Cognitive Linguistics 3. 182–200. Gries, Stefan Th. & Stephanie Wulff. 2009. Psycholinguistic and corpus-linguistic evidence for L2 constructions. Annual Review of Cognitive Linguistics 7. 163–186. Hakuta, Kenji. 1974. Prefabricated patterns and the emergence of structure in second language acquisition. Language Learning 24(2). 287–297. Heritage, John. 1984. A change-of-state token and aspects of its sequential placement. In J. Maxwell Atkinson & John Heritage (eds.), Structures of social action, 299–345. Cambridge: Cambridge University Press. Hopper, Paul J. 1998. Emergent grammar. In Michael Tomasello (ed.), The new psychology of language, vol. 1, 155–175. Mahwah, NJ: Lawrence Erlbaum. Huth, Thorsten. this volume. Playing with turns, playing with action? A social-interactionist perspective, 47–72. Langacker, Ronald W. 1987. Foundations of cognitive grammar, vol. 1. Theoretical prerequisites. Stanford: Stanford University Press. Langacker, Ronald W. 2000. A dynamic usage-based model. In Michael Barlow & Suzanne Kemmer (eds.), Usage based models of language, 1–65. Stanford: CSLI. Lantolf, James P. 1997. The function of language play in the acquisition of L2 Spanish. In Ana Teresa Perez-Leroux & William R. Glass (eds.), Contemporary perspectives on the acquisition of Spanish. Vol. 2: Production, processing, and comprehension. Somerville, MA: Cascadilla Press. Lantolf, James P. 2006. Sociocultural theory and L2. State of the art. Studies in Second Language Acuqisition 28. 67–109. Li, Peiwen, Søren W. Eskildsen & Teresa Cadierno. Tracing an L2 learner’s motion constructions over time: A usage-based classroom investigation. The Modern Language Journal 98(2), 612–628 Lieven, Elena. 2010. Input and first language acquisition: Evaluating the role of frequency. Lingua 120. 2546–2556. Lieven, Elena, Heike Behrens, Jennifer Speares & Michael Tomasello. 2003. Early syntactic creativity: a usage-based approach. Journal of Child Language 30(2). 333–370. Lieven, Elena, Dorothé Salomo & Michael Tomasello 2009. Two-year-old children’s production of multiword utterances: A usage-based analysis. Cognitive Linguistics 20(3). 481–508. MacWhinney, Brian. 1975. Pragmatic patterns in child syntax. Stanford papers and reports on child language development 10. 153–165. MacWhinney, Brian. 1982. Basic syntactic processes. In Stan Kuczaj, II (ed.) Language development, volume 1: Syntax and semantics, 73–136. Hillsdale, NJ: Lawrence Erlbaum. MacWhinney, Brian. 2004. A multiple processes solution to the logical problem of language acquisition. Journal of Child Language 31. 883–914. Markee, Numa. 2000. Conversation analysis. Mahwah, NJ: Lawrence Erlbaum. Mellow, J. Dean. 2006. The emergence of second language syntax: A case study of the acquisition of relative clauses. Applied Linguistics 27(4). 645–670. McDonough, Kim & Yujin Kim. 2009. Syntactic priming, type frequency, and EFL learners’ production of wh-questions. The Modern Language Journal 93(3). 386–398. Myles, Florence, Janet Hooper & Rosamond Mitchell. 1998. Rote or rule? Exploring the role of formulaic language in classroom foreign language learning. Language Learning 48(3). 323–363.



The emergence of creativity in L2 English: A usage-based case-study 

 315

Myles. Florence, Rosamond Mitchell & Janet Hooper. 1999. Interrogative chunks in French L2. A basis for creative construction? Studies in Second Language Acquisition 21(1). 49–80. Nattinger, James R. & Jeanette S. DeCarrico. 1992. Lexical phrases and language teaching. Oxford: Oxford University Press. Ohta, Amy S. 2001. Second language learning processes in the classroom: Learning Japanese. Mahwah, NJ: Lawrence Erlbaum. Olsher, David. 2004. Talk and gesture: The embodied completion of sequential actions in spoken interaction. In Rod Gardner & Johannes Wagner (eds.), Second language conversations, 221–245. London: Continuum. Ortega, Lourdes. 2014. Ways forward for a bi/multilingual turn in SLA. In Stephen May (ed.), The multilingual turn: Implications for SLA, TESOL and bilingual education, 32–53. New York: Routledge. Pawley, Andrew & Frances Hodgetts Syder. 1983. Two puzzles for linguistic theory. In Jack C. Richards & Richard W. Schmidt (eds.): Language and communication, 191–226. Harlow: Longman. Pekarek Doehler, Simona & Evelyne Pochon-Berger. 2015. The development of L2 interactional competence: evidence from turn-taking organization, sequence organization, repair organization and preference organization. In Teresa Cadierno & Søren W. Eskildsen (eds.), Usage-based perspectives on second language learning, 233–268. Berlin: De Gruyter Mouton. Pinker, Steven. 1999. Words and rules. London: Weidenfeld and Nicholson. Pomerantz, Anne & Nancy Bell. 2007. Learning to play, playing to learn: FL learners as multicompetent language users. Applied Linguistics 28(4). 556–578. Reder, Steve. 2005. The “Lab School”. Focus on Basics 8(A). Available online: http://www.ncsall.net/fileadmin/resources/fob/2005/fob_8a.pdf Reder, Steve, Kathryn A. Harris & Kristen Setzler. 2003. A multimedia adult learner corpus. TESOL Quarterly 37(3). 546–557. Roehr-Brackin, Karen. 2014. Explicit knowledge and processes from a usage-based perspective: The developmental trajectory of an instructed L2 learner. Language Learning 64. 771–808. Saville-Troike, Muriel. 1988. Private speech: evidence for second language learning strategies during the “silent period. Journal of Child Language 15. 567–590. Schmidt, Richard W. 1983. Interaction, acculturation, and the acquisition of communicative competence: A case study of an adult. In Neesa Wolfson & Elliot Judd (eds.), Sociolinguistics and language acquisition, 137–174. Rowley, MA: Newbury House Swain, Merril & Sharon Lapkin. 1998. Interaction and second language learning: Two adolescent French immersion learners working together. The Modern Language Journal 82(3). 320–337. Tomasello, Michael. 2000. First steps toward a usage-based theory of language acquisition. Cognitive Linguistics 11(1–2). 61–82. Tomasello, Michael. 2003. Constructing a language. Cambridge: Cambridge University Press. Tummers, Jose, Kris Heylen & Dirk Geeraerts. 2005. Usage-based approaches in cognitive linguistics: A technical state of the art. Corpus Linguistics and Linguistic Theory 1(2), 225–261. van Lier, Leo. 2000. From input to affordance: Social interactive learning from an ecological perspective. In James P. Lantolf (ed.), Sociocultural theory and second language learning: Recent advances, 245–259. Oxford: Oxford University Press

316 

 Søren W. Eskildsen

Vogt, Paul & Elena Lieven. 2010. Verifying theories of language acquisition using computer models of language evolution. Adaptive Behavior 18. 21–35. Waara, Renee. 2004. Construal, convention, and constructions in L2 Speech. In: Michel Achard & Susanne Niemeier (eds.), Cognitive linguistics, second language acquisition and foreign language pedagogy, 51–75. Berlin: Mouton de Gruyter. Waring, Hansun Zhang. 2013. Doing being playful in the Second Language Classroom. Applied Linguistics 34(2), 191–210. Weinert, Regina. 1995. The role of formulaic language in second language acquisition: A review. Applied Linguistics 16(2). 180–205. Year, Jungeun & Kim Gordon. 2009. Korean speakers’ acquisition of the English ditransitive construction: The role of verb prototype, input distribution, and frequency. The Modern Language Journal 93(3). 399–417. Yuldashev, Aziz, Julieta Fernandez & Steven L. Thorne. 2013. Second language learners’ contiguous and discontiguous multi‐word unit use over time. The Modern Language Journal 97(S1). 31–45.

Jiyun Kim

12 Teaching language learners how to understand sarcasm in L2 English Abstract: This study reports on a project that aimed to develop students’ ability to detect and understand sarcasm in L2 English as used in spoken discourse. Derived from Vygotsky’s theory of consciousness and Gal’perin’s associated theory of educational development, the study helped learners by providing direct instruction on identifying and interpreting L2 sarcasm. Among the multiple data sets collected, this chapter particularly focuses on analysis of preand post-instruction individual interviews used to gauge learner development in understanding the concept of English sarcasm. By developing theoretical knowledge of sarcasm, students established a solid cognitive framework for more readily understanding the L2 concept. More importantly, learners gained a sense of empowerment by finally understanding subtle features of sarcasm that they had not previously recognized. The results highlight the importance of instructional quality and teacher-learner dialectics in which learners interact with an expert tutor who offers pedagogically designed psychological tools and semantic-pragmatic explanations to promote a functional understanding of subtle concepts such as sarcasm.

1 Introduction Sarcasm is a cultural phenomenon ubiquitous in contemporary societies. People use sarcasm to achieve various communicative goals. Some use it to show their negative or positive attitudes (see Kim 2014:193 for an outline), while others use it to skillfully hide their emotion and intent (Giora, 2004; Ekman, 1984). In using and comprehending sarcasm, speakers of English consider particular types of linguistic signals such as visual, prosodic, lexical, and contextual cues that exist within the conventions of the society. Particularly in the U.S., sarcastic characters are easily found in numerous scripted and unscripted television shows. Achieving conceptual understanding of sarcasm in another languaculture (Agar, 1994) is no easy matter. It requires examination of the motivation, intensity, means of expression, and goals of sarcasm in that particular culture. Mere exposure to the target language setting or having already achieved high proficiency in the language does not guarantee that learners will gain this ability. Bouton’s (1999) study on the second language (L2) acquisition of irony also DOI 10.1515/9781501503993-013

318 

 Jiyun Kim

disproves any co-relation between learners’ language proficiency and irony comprehension. Despite the need for instruction, teaching learners how to understand sarcasm in L2 English is an under-researched area in applied linguistics. This problem originates partly from the fact that the ambiguous nature of sarcasm makes the concept difficult to concretize into a tangible form and a teachable unit. Developing learners’ understanding of this complex concept requires an intervention carefully designed with theoretical and empirical rigor. This chapter reports on Kim’s study (2013) that introduces an innovative pedagogical approach for teaching the concept of sarcasm to L2 learners. In particular, this chapter provides a complementary account of learners’ concept development by tracing the trajectories of learner knowledge prior to and after a particular form of instruction referred to as Systemic-Theoretical Instruction (STI). To show evidence of development, this chapter documents three major stages of learner growth: learners’ pre-understanding prior to STI, learner-teacher dialectics during STI, and learners’ communicated thinking as evidenced through verbalization data collected after STI.

2 Background 2.1 Sarcasm in English Sarcasm is a complex social phenomenon that can hardly be thoroughly explained with a single theory or definition. Sarcasm as a speech act (Austin 1962; Searle 1969, 1979) is a linguistic behavior intentionally generated by goal-directed individuals acting to achieve specific communicative goals. These individuals frame (Goffman 1974) speech events in a way that steers their interlocutors to see their actual intent. The detection of sarcasm is triggered by a perceived mismatch between the literal and intended meanings of the potential sarcasm, and the type and number of cues incorporated by a speaker influence interlocutors’ comprehension (Yus 1998, 2000). The degree of emphasis placed in producing these cues also affects the comprehension process. For instance, if the speaker’s raised eyebrow is salient enough, that single cue could be sufficient for the interlocutor to detect sarcasm. However, the level of salience from each cue is subjective; that is, interlocutors are likely to resort to interpreting the cues most relevant to themselves (Sperber and Wilson 1981, 1992). The types of cues and strategies speakers choose to use to project sarcasm are culture-specific. Not all cultures perceive eye rolling or placing one’s hands on one’s hips as possible indicators of sarcasm. These cues may carry less sarcastic meaning or no meaning at all to interlocutors



Teaching language learners how to understand sarcasm in L2 English 

 319

from some cultures. Certain visual cues (e.g., a downward mouth pull for contempt) are likely to be more universal than others (Ekman 1993). Some of these nonverbal cues may leak into conversation (Ekman & Friesen 1969; Ekman et al., 1987) regardless of the speaker’s intention. Little consensus has been reached on the distinction between the terms sarcasm and irony. Some scholars differentiate the two, while others treat them as the same (see Attardo 2000: 795 for an outline). The two terms can no longer be realistically differentiated by Attardo et al.’s (2003) position that “a shift in meaning for the word irony seems to be taking place with sarcasm occupying what was previously the semantic space of irony” (p. 243). Partington’s (2007) definition of implicit irony is particularly useful since it is derived from the examination of real-life corpus data. The key factor in irony, according to his data analyses, is an implied reversal of the evaluative meaning of the utterance. The speaker intentionally and strategically poses a pair of contrastive narratives and generally switches “from favorable evaluation in the dictum to unfavorable evaluation in the implicatum” (p. 1565). Indeed, as for implicit irony, the key to understanding sarcasm is speaker intentionality. For a full appreciation of sarcasm, interlocutors need to comprehend why the speaker is exhibiting particular linguistic behaviors and to accomplish what functional goals. In fact, analyses of sarcasm on U.S. TV shows in the current study indicated that the intentional juxtaposition of two contrasting realities is a key strategy for creating sarcasm. The analyses revealed that the ironic contrast of realities can occur at multiple levels and can be achieved through both verbal and non-verbal channels. Even without producing any utterance, raising the corners of one’s lips with a tilted head and wide-open eyes can be sufficiently interpreted as an ironic facial expression. That is, interlocutors can detect a contrast between a given situation and what is likely to be an inappropriate facial expression(s) in that context. Creating a theatrical situation in which the speaker acts as if s/he is not aware of the actual reality behind ironic speech is another typical strategy speakers use. For this strategy, interlocutors perceive the mismatch between the actual and the artificially created theatrical realities, which guides them to read the underlying message. As the speaker’s intent, conveyed through words and expression, can range on a continuum from negative to positive, interlocutors’ interpretations also move along the same scale. The analyses of the following examples of sarcasm will illustrate these production and comprehension mechanisms in more detail. Negative sarcasm usually conveys bitter intent and attitudes with one or more of the following communicative goals: to ridicule, to insult, to criticize, to show anger, to vent frustration, etc. Speakers incorporate multiple strategies (simultaneously)

320 

 Jiyun Kim

to convey their sarcastic messages. They frequently use irony to create contrast between two different realities and hyperbole to maximize the gap. In Excerpt 1, Judge Judy creates a significant contrast between the obvious modern reality of using dollars as currency and a made-up reality of using rice and beans as currency. The obvious contrast between the hypothetical and actual situations is highlighted further via her simultaneous use of non-verbal cues (i.e., facial expressions, line 6). The judge intentionally ridicules the defendant in order to acquire the information she seeks. Excerpt 1: Negative sarcasm ( Judge Judy, 2010) 1 Judy: how much do you earn? 2 Defendant: a good amount. 3 Judy: I’m asking you a ↑question. a good amount is not an 4 answer. 5 Defendant: like (.) what do you want. dollar amount? figures? Judy: yes. it’s usually in dollars unless they pay you in rice and 6 7 beans. ((opens mouth, rolls eyes, and makes a blank face)) 8 Audience: @@ 9 Defendant: I do well. 10 Judy: I asked a question. the question is, how much do you 11 ↑earn. 12 Defendant: I don’t think I need to display how much I make. Speakers also attempt to demonstrate their wittiness through sarcasm while making their points. Excerpt 2 depicts a video clip from The View, ABC Daytime’s morning chat show that features a team of five female co-anchors debating and discussing topics in the news. In the episode transcribed in Excerpt 2, the CEO of AshleyMadison.com, Noel Biderman, was invited as a guest to discuss the controversial nature of his business, which was designed to facilitate infidelity among married people. Before the anchors welcomed him to the stage, they played an advertisement for AshleyMadison.com. Barbara Walters, one of the co-anchors, uses sarcasm primarily to criticize Biderman for running the business (and to show her disapproval of him to viewers) (line 18). Excerpt 2: Negative sarcasm (The View, 2011) Hasselbeck: so (.) imagine you’re a married man or woman sitting at 1 2 home watching TV when the following commercial comes on (.) and you may want to get the kids out of the room 3 4 right now before we show it to you. take a look. 5 Ad Narrator: isn’t it time for AshleyMadison.com? ((the ad features a man and a woman kissing in a bedroom)) 6



Teaching language learners how to understand sarcasm in L2 English 

 321

7 Audience: no:: ((booing)) 8 Hasselbeck: wow ((her eyebrows go up and down and she smiles)) (..) 9 okay, are you as outraged as our audience is, or curious 10 and racing to your computer? please welcome one of the most hated men on the internet, ((smiles)) founder and 11 12 CEO of AshleyMadison.com, Noel Biderman. welcome 13 Noel (.) okay, you’re obviously hearing a reaction from 14 our audience that doesn’t sound too positive. [lines omitted] 17 your website essentially caters to those in committed married relationships who are looking to have an affair. 18 now, you are married. you have two young children,= 19 20 Walters: =they must be so proud. your children. ((staring at Biderman with a blank face)) 21 22 Audience: @@@ 23 Biderman: ((looks down)) 24 Hasselbeck: other than the obvious reason for making a buck (.) why do 25 this? Through her utterance in line 20, Walters intentionally creates a contrast between the literal meaning of her utterance and what is likely to be perceived as commonsense reality: facilitating illicit affairs among married individuals is unethical. As demonstrated through these two examples, sarcasm in English is frequently associated with the speaker’s intention to be witty and humorous while also being critical. Positive sarcasm, on the other hand, entails positive intent with one or more of the following communicative goals: to be humorous, to show affection, to be friendly, to break the ice, to bond with others, and to promote group solidarity. One of the common strategies associated with positively intentioned sarcasm is the inclusion of criticism the speaker directs at herself or her situation in order to generate humor, as in the following example (Excerpt 3). In this example from the television drama Grimm, Nick’s wife expresses sarcasm about a situation that she finds quite dull. Excerpt 3: Positive sarcasm (Grimm: Season 1, Episode 12, 2012) ((Nick and his wife are on the phone.)) Nick: hey, what’s up? Wife: some ve::ry exciting laundry. ((smiling)) Nick: ah, no wonder you called.

322 

 Jiyun Kim

Similarly, in Excerpt 4, Bob shows his sarcastic attitude towards an individual’s unfortunate experience in order to convey sympathy. Excerpt 4: Positive sarcasm (Observational data, 2012) ((One person’s friend Bob has observed her experiencing a series of unfortunate events in one day: getting splashed by muddy water on the way to work in the morning, losing a wallet at a restaurant in the afternoon, and having her vehicle towed in the evening.)) Bob: This must be the luckiest day of your life. Bob attempts to create humor through a sarcastic remark to show sympathy for his interlocutor. The examples of positive sarcasm presented above illustrate cases where sarcasm is directed at particular circumstances in which interlocutors have found themselves. Nevertheless, according to my observational data, young generational speakers of American English seem to demonstrate an increasing tendency towards using their interlocutors as victims to create positive sarcasm, rather than creating positive sarcasm by criticizing themselves. However, more research is required to support this idea. Overall, speakers of American English convey messages through sarcasm by skillfully orchestrating visual, prosodic, lexical, and contextual cues. In order for L2 learners to grasp the underlying meaning of sarcasm, they need to gain the ability to deconstruct contexts into the many separate cues that together constitute sarcastic meaning. The researcher/instructor in this study focused solely on sarcasm in verbal contexts while excluding situational, Socratic, dramatic, and any other types of unintended irony. The researcher also ignored perspectives from third parties, such as audience members at TV shows, and focused exclusively on the two parties involved in the communication—speakers and interlocutors.

2.2 Pre-understanding from Korean L2 learners of English Some might wonder why anyone needs to teach L2 learners how to understand sarcasm. The rationale behind this necessity originates from learner empowerment. The following commentary by one of the participants in the current study effectively attests to a pedagogical imperative in this area for L2 learners: When someone uses sarcasm to me in Korean, I can at least show that I am offended by making an angry face or something…but in the English-speaking context, I simply become stupid. Even when someone is being sarcastic to me, I will not understand what is actually going on and they will think of me as a stupid person, which is obviously not true. (…) Just



Teaching language learners how to understand sarcasm in L2 English 

 323

because of cultural differences and my limited ability to understand the language, I become a stupid person, and I hate that reality. 

(Cho, 2013)

The primary goal of teaching sarcasm is to expand learners’ linguistic capacity for more nuanced forms of interaction in the L2 and, ultimately, to empower them to become more effective communicators. The first step in implementing effective instruction is discovering the current stage of students’ development—that is, what students already know about the target concept. After that, an instructor can connect this pre-understanding (see Miller 2011) to new knowledge through mediation. Once learners have fully acquired the concept, they will no longer need aid from materials and their instructor (Vygotsky 1987). The central property that defines learners’ developmental potential during this process is the quality of the psychological tools that are provided to them in the course of instruction. When the psychological tool(s) coherently and systematically present essential characteristics of the concept(s) under study, instruction results in profound developmental progress (Arievitch & Stetsenko 2000). For this reason, identifying the workings of learners’ L1 equivalent concept(s) is a crucial step in teaching the target L2 concept. Kim’s (2014) empirical study examined how Korean EFL learners understand sarcasm and found that they conceptualize sarcasm through their L1 equivalent concepts, ban-eo (irony) and bi-kkom (sarcasm). Specifically, the Korean participants perceived sarcasm exclusively as a taboo act, whereas monolingual English-speaking participants associated sarcasm more with humor and acknowledged its dual role—both positive and negative. Additionally, the Korean participants treated sarcasm as a speech act through which speakers can indirectly express their negative intent or cover their emotions, while no monolingual English-speaking participant reported the same possible speaker intent for any utterance. In order to help learners develop a conceptual understanding of sarcasm, it is necessary to inhibit irrelevant L1 knowledge and add new relevant knowledge through systematically designed instruction. This will ultimately enable learners to make a more effective interpretation of sarcasm in L2.

2.3 Why Systemic-Theoretical Instruction? Effective interpretation of L2 sarcasm requires pedagogical intervention designed to promote learners’ conceptual development. Previous studies (Bouton 1994, 1999; Davies 2003; Bell 2005, 2006) have indicated that abilities to interpret and produce

324 

 Jiyun Kim

implicatures like irony improve over time through accumulated exposure to the target language environment. Although this appears to be the general tendency, the question at stake is why anyone should wait and let learners haphazardly reach decisions through trial and error when effective instruction can instead shorten the entire process of learning the concept. Moreover, due to the complex nature of sarcasm, it is necessary for students to learn about multiple sarcasm-related cues that simultaneously occur and constitute sarcastic meaning. Learners will benefit from this bottom-up approach, which will eventually lead them to achieve greater accuracy in comprehension, as is the case for NSs who are not necessarily capable of identifying every cue (and thus may benefit from a top-down approach) but possess the necessary intuition for successful detection. The quality of instruction is guaranteed when the target concept is presented to students in a complete and coherent manner. This can be accomplished through pedagogically designed psychological tools that effectively illustrate the core characteristic(s) of the concept. Piotr Gal’perin (1989, 1992) argued for the importance of such systematic explanations of theoretical concepts to generate conceptual development in learners. He particularly developed the notion of Schema for the Orienting Basis of Action (henceforth SCOBA). SCOBA entails the presentation of the target knowledge in a more tangible form and in a more principled way, generally in a visual format, such as through figures, diagrams, or objects. Interaction with learners through these meditational tools should provoke genuine understanding of the concept, rather than enticing students to simply memorize the concept. An increasing number of studies report the effectiveness of using SCOBA in various L2 instructional settings. Oboukhova et al. (2002), for example, used a cartoon story presented via computer as a SCOBA to materialize the grammatical concept of aspect in French. The L1 Russian speakers in the experimental group who engaged in the SCOBA outperformed learners in the control group who received traditional instruction without exposure to the SCOBA. In Serrano-Lopez and Poehner’s (2008) study, the instructor taught Spanish locative prepositions (de, en, a) by presenting them with the conceptual principles of each preposition’s usage. During instruction, the researchers asked students to create figures through clay modeling that depicted the use of each preposition. Following instruction, students were tested on their use of locative prepositions. On a delayed post-test administered two weeks after the initial post-test, the clay-modeling group significantly outperformed the other two control groups, indicating that the SCOBA-engaged instruction had a longer-lasting impact due to its imagistic and tactile qualities. Findings from more recent studies corroborate the previous research into concept-based language teaching. Van Compernolle (2014) taught U.S. learners of French how to choose appropriate pragmatic features of L2 French second(tu/vous) and first- (on/nous) person pronouns and negation (ne … pas / …pas).



Teaching language learners how to understand sarcasm in L2 English 

 325

He developed concept cards that illustrated the fundamental principle at work for choosing these forms that best met the learners’ communicative needs. By engaging in this systematic orienting basis, students learned that the selection of the form depends on how speakers view the relationship between themselves and their interlocutors (e.g., social distance, power hierarchies) as well as on the degree of indexicality (Silverstein 2003). Zhang (2014) integrated use of Cuisenaire rods in instruction to examine the effects of STI in the development of L2 Chinese topicalization. While most of the educational projects carried out within STI for L2 development have focused on grammatical features, the current study reported in this chapter used a series of diagrams and charts to coherently explain the concept of sarcasm as well as to depict the cues for its detection in speech. The researcher/instructor in this study developed eight pedagogical diagrams to provide learners with an orienting basis through which she sought to coherently and systematically explain the concept of sarcasm during the STI interaction. The study addresses the following research question: In what ways, if any, does STI promote development in understanding the concept of sarcasm among L2 learners of English?

3 Methods 3.1 Participants Nine Korean advanced-level university learners of L2 English participated in the study. An email advertisement was sent to Korean PhD students and post-doc researchers in the departments of engineering, economics, computer science, and other science- and math-related majors at a large U.S. research university. A total of twenty-three people volunteered, and ten1 participants were selected through a screening process that entailed a pre-test and an individual interview. The pretest consisted of 10 test items and utilized the following materials: 10 video clips, a video scene script for each clip, and test sheets. For each test item, the twentythree volunteers were asked to watch a short video clip and fulfill three tasks: (1) a sarcasm identification task, (2) a speaker intent comprehension task, and (3) a potential sarcasm cue identification task. All three tasks were completed in the presence of the researcher.

1 Among the ten participants, one withdrew from the study in Week 4, due to a family emergency. The remaining nine participants all completed the sixteen-week study.

326 

 Jiyun Kim

Once the test was over, the researcher held individual interview sessions with the twenty-three volunteers. During the interviews, the researcher asked questions mainly to check participants’ understanding of the definition and use of sarcasm in both L1 and L2. If necessary, the researcher also asked participants to elaborate more on their answers for Task 3 test items. Volunteers were also asked to verbalize their inferencing processes (i.e., explain why they identified speech as sarcasm) and how they discerned the speaker’s intended meaning and attitude. The researcher selected ten participants from the twenty-three volunteers based on their test scores and interviews. This recruitment was based on the following three criteria, in this order. First, volunteers who correctly identified more than fifteen sarcastic utterances out of twenty in the pre-test were excluded. Second, the researcher prioritized volunteers who previously had not watched the majority of the video clips that were used in the three tests and instruction. Third, after the first two steps, volunteers who expressed interest (e.g., motivation and the desire to learn) in the instruction were prioritized more than others.

3.2 Systemic-Theoretical Instruction and SCOBAs The goals of the instruction in the present study were to promote learners’ internalization of the concept of English sarcasm as well as to assist them in using this conceptual knowledge to detect and appropriately interpret sarcasm when uttered by native speakers of English. As previously mentioned, a mere presentation of knowledge itself is not sufficient for in-depth understanding of the concept. Researchers instead should devote more attention to how the knowledge is presented to and internalized by learners (Lantolf, 2006). The eight diagrams were developed by the researcher to function as conceptual macrostructures that create a general orientation framework to enable learners to analyze concrete instances of sarcasm when it is used in spoken discourse. Due to limited space in this chapter, only SCOBAs 1 and 8 are presented in detail, and the others are briefly described. SCOBA 1 depicts various speaker intentions and communicative goals that speakers intend to accomplish via sarcasm. Speaker attitudes and intent exist along a continuum, at one end of which sarcasm conveys negative messages (i.e., to insult), while at the other end sarcasm expresses positive intentions (i.e., to bond). The tutor provided examples of sarcasm and situations in which these utterances are likely to be used to accomplish each communicative goal. While explaining the ubiquity of sarcasm in U.S. culture, the tutor further elaborated on possible types of emotions that may prompt speakers to convey various sarcastic intentions and attitudes. The tutor also provided real-life examples (i.e., observational

To criticize To insult To mock, ridicule and make fun of someone To complain To vent frustration/anger/displeasure (or other negative emotions) 6. To not show anger (or other negative emotions) directly 7. To cover speakers’ subversive motives 8. To save speaker’s face 9. To show one’s maintaining self-control 10. To lessen vulnerability 11. To admonish 12. To show frustration

1. 2. 3. 4. 5.

bitter

NEGATIVE SARCASM

Contempt Disgust Anger Disagreement Disapproval Dislike Frustration Jealousy Other unhappy & negative feelings

negative

1. To save interlocutor’s face 2. To persuade 3. To show off personal skills (i.e., to be funny)

DIVERSE COMMUNICATIVE GOALS

TO ACHIEVE

INTENT & ATTITUDES

TO SHOW

Using Sarcasm

PROMPT

EMOTIONS

1. 2. 3. 4.

To give compliments To show affection To promote group solidarity To improve relationships

light-hearted

POSITIVE SARCASM

Affection Bond Positive feelings

positive

 Teaching language learners how to understand sarcasm in L2 English 

 327

SCOBA 1: How different emotions may prompt a speaker to use different types of sarcasm to achieve various communicative goals

328 

 Jiyun Kim

verbal data) in which the various communicative goals are accomplished through uttering sarcasm in different contexts. SCOBA 2 introduces ‘verbal irony’ as an overarching concept that entails five subcategories: ironic sarcasm, hyperbole, understatements, double entendre, and rhetorical questions. SCOBA 3 displays pictures of multiple types of facial expressions that sarcasm users tend to exhibit either intentionally or unintentionally. These expressions include a sneer, deadpan look, fake smile, and face with a downward pull. After discussing each facial expression, the instructor provided video-clip examples in which sarcasm users incorporated these facial expressions. SCOBA 4 concretizes the notion of sarcastic facial expressions by providing examples of potential sarcastic eye and eyebrow movements of a face. The instructor provided video-clip examples of these expressions, which learners practiced analyzing and deconstructing by identifying how each of the detailed parts of the face moved. SCOBA 5 is similar to SCOBA 4 but focuses on the mouth area by discussing movements of the lips, lip corners, jaw and cheek muscles, and tongue. SCOBA 6 illustrates potential sarcastic gestures by providing pictures. After leading a discussion on each gesture in the diagram, the instructor provided video-clip examples in which speakers incorporated the potential sarcastic gestures discussed. SCOBA 7 provides examples of body movements that sarcasm users tend to adopt. They include a tilted head, raised chin, fast or slow nodding, and so on. SCOBA 8 is a flowchart that presents a series of steps to be followed as a guide for helping learners detect sarcasm when it is used in verbal ­discourse. The initial step asks learners to determine if there is a contrast between a speaker’s utterance and an interlocutor’s expectation of a non-sarcastic utterance, given the context of the communication. In other words, the interlocutor should sense some kind of mismatch between the speaker’s utterance and non-sarcastic utterance(s) that could generally be expected within the given situation. For instance, if the speaker utters, “What great weather!” when it has been thundering and raining for five consecutive days, the interlocutor should sense some kind of incompatibility between the way the speaker evaluated the circumstance and the utterance(s) the interlocutor might expect in



Teaching language learners how to understand sarcasm in L2 English 

 329

How to Understand Ironic Sarcasm Speaker’s Utterance

Your understanding of the reality

CONTRAST YES

Enough contrast? Not sure?

Check the degree and/or number of incompatibilities.

Speaker is using ironic sarcasm.

Speaker’s encyclopedic knowledge about the culture of the community Facial expression Gesture Body movements Prosodic cues Physical settings Biographical data (e.g., personality) Previous context (e.g., dialogue, situation) Current emotional state Social relationship b/w speaker and interlocutor Lexical cues (e.g., hyperbolic adverbs, adjectives)

Enough degree and/or number of incompatibilities? YES

NO Ambiguous

SCOBA 8: How to detect and understand sarcasm

non-sarcastic communication. The flow chart guides the interlocutor to possible next steps integrated with the visual, prosodic, and contextual cues needed for him/her to conclude the correct speaker intent of sarcasm or non-sarcasm. Learners’ engagement in the SCOBAs included the following three steps: presentation of a SCOBA, discussion of the SCOBA, and discussion of video clip example(s) that represent and exemplify the concept in the SCOBA. Discussions of the pedagogical diagrams always included video-clip examples that concretized the relevant concepts. Throughout the tutorials, learners were provided with more than 30 video clips that contained various instances of sarcasm with different combinations of potential sarcastic cues. The clips were taken primarily from U.S. TV shows (e.g., non-scripted talk shows, sitcoms, and cartoons) and YouTube video clips that featured political debates, TV news programs, and so on. Additionally, the tutor used written examples of sarcasm taken primarily from Amazon product reviews and Twitter and Facebook commentaries. The discussions also included analogies between learners’ L1 and L2 in terms of the meanings and use of relevant concepts and the multiple cues for processing sarcasm introduced in the SCOBAs. Instruction was conducted once a week for a twelve-week period. The tutor/ researcher met with the students individually once a week. Each tutorial session

330 

 Jiyun Kim

lasted for an hour and was video-recorded for analysis. The entire instruction, including the administration of three tests, was carried out by the author over a sixteen-week period.

4 Data collection All learners were given a pre-test before instruction began as well as a post-test one week after instruction ended. A second delayed post-test was administered four weeks after the initial post-test. All the assessments consisted of two parts: a written test and an individual interview. A written test was conducted first, followed by the individual interviews. The primary sources of ethnographic data consisted of five main subsets: (1) Individual interviews (i.e., pre-STI interview, post-STI interview); (2) In-class interactions (e.g., verbal explanations of homework); (3) Group activities (i.e., visual cue miming); (4) Focus-group discussions; (5) Student writing (i.e., interpretive-essay writings; movie script writing); and (6) Student-produced SCOBAs. Learners’ verbalization data is a crucial source for tracing the stages of development in internalizing the target conceptual knowledge (see also Swain’s 2006 study for the notion of languaging). This data reveals the nature of the interactions that occurred during instruction, which is crucial to understanding in what ways the STI interaction promoted (or did not promote) learners’ internalization of the concept and conceptual development. Individual interviews Each learner participated in two individual interviews throughout the tutorial. The initial interviews were held before STI. The primary purpose of the initial interview was to discover the background knowledge of learners’ understanding of the notion of sarcasm both in their L1 (Korean) and L2 (English). It was expected that the learners would not have much conscious awareness of the details of how sarcasm functions in either their L1 or L2. During the instruction, the tutor attempted to enable the learners to make their implicit knowledge of sarcasm in L1 conscious, and then to connect this knowledge to the new knowledge they were acquiring in the English L2 context. In this way, the developmental process started from what the learners already knew: their knowledge of L1 sarcasm. The post-instruction interview aimed to investigate students’ learning processes and outcomes in regards to the instruction they received. The students were asked to verbally explain (1) how they defined the concept of English sarcasm, (2) how they understood certain sarcastic utterances, which were prerecorded to ensure uniformity of presentation for each learner, and (3) their



Teaching language learners how to understand sarcasm in L2 English 

 331

learning experience and overall thoughts and feelings about the instruction they received. The communicated thinking required learners not to simply memorize the definition of the concept but rather to use the conceptual framework, the SCOBA, as a guide to explain the concept to the teacher and peers (during focus group discussions). Focus-group discussions Focus-group discussions were video-recorded with the expectation that learners would express their opinions in conversation with other participants. The nine participants, in groups of three, gathered in a conference room equipped with a wide-screen TV and watched a three-minute video clip taken from a non-scripted TV courtroom show. The instructor provided a script for the video clip and asked the participants to underline every instance of sarcasm they identified while viewing the clip. The participants were also asked to prepare rationales for their selections. Learners then discussed which utterances they identified as sarcasm and provided the underlying meanings and possible speaker intent, attitudes, and communicative goals for each sarcastic utterance. Participants actively expressed their agreements and disagreements with each other’s interpretations of the selected sarcastic utterances and the roles and meanings of the relevant cues. The current chapter particularly presents the learners’ communicated thinking gathered before (i.e., individual interviews), during (i.e., in-class interaction), and after STI (i.e., individual interviews), in which they explained to teachers or peers their understanding of the instructed concept(s) and of their performance while using the concept(s).

5 Analysis Development was measured based on three criteria: (1) concept development in the definition and use of sarcasm as evidenced in verbalization data; (2) sarcasm detection and comprehension ability as evidenced by performance on the three tests; and (3) the ability to use appropriate cues in detecting sarcasm as evidenced in the answers and the scores from the three tests. Pre-, post-, and delayed posttest scores were compared via the two-tailed Wilcoxon signed-ranks test to assess for improved sarcasm comprehension ability following STI. Among the multiple sets of data, this chapter focuses exclusively on students’ verbalization data collected from their individual interviews before and after STI. All communication between the tutor and learners occurred in Korean. Excerpts of data presented have been transcribed and then translated from Korean to English using the conventions shown in Appendix A.

332 

 Jiyun Kim

6 Results and discussion The analysis of learners’ verbalization data indicated learner development in their understanding of the concept of sarcasm. Before STI, learners demonstrated either incorrect or incomplete understanding of sarcasm, primarily due to their pre-understanding that centered on the L1 equivalents of English sarcasm: bi-kkom (sarcasm) and ban-eo (irony). After STI, however, students showed greater control over sarcasm by demonstrating a more mature understanding of the concept after experiencing mediation by both the SCOBA and the tutor. The following sections present focused snapshots of “what instruction looks like” and “what happened” across time to trace trajectories of student development at individual levels. 6.1 Pre-understanding Analysis of pre-STI interviews and STI interaction revealed particular sources of confusion in learners and areas that could be focused on to provide instruction. Before STI, Wang (Excerpt 5) defined sarcasm in a non-principled way by relying heavily on his intuition. Excerpt 5: Wang, defining sarcasm in the pre-STI interview Wang: something you would feel bad about after thinking about it? (..) at first you wouldn’t know but after two seconds you would realize. Tutor: can you think of any example? Wang: I don’t remember any (..) I believe I might have heard many when I was serving in the army. Tutor: okay Similarly, Yoh (Excerpt 6) did not demonstrate substantial theoretical knowledge about the concept and drew upon the definition of Korean sarcasm, bi-kkom, when asked to define English sarcasm. The term bi-kkom itself literally translates as ‘an act of twisting’ in Korean. Excerpt 6: Yoh, pre-STI interview Tutor: how would you define the term? Yoh: maybe something like twisting the situation or going against the interlocutor’s intention? Mia (Excerpt 7) provided incomplete knowledge about the concept and expressed great uncertainty while defining the term.



Teaching language learners how to understand sarcasm in L2 English 

 333

Excerpt 7: Mia, pre-STI interview Mia: um (.) metaphorical expressions? figurative expressions? Tutor: I see Mia: or maybe not? ((laughs)) Tutor: I see. that’s how you understand it. Mia: yes. All nine participants, during pre-STI interviews, provided incorrect or incomplete definitions that lacked understanding of the core characteristics and functionality of sarcasm: reversed-meaning, sarcasm-related cues, and the speaker’s communicative goals.

6.2 Mediation Focus on the central component of sarcasm—reversed-meaning Over the course of the instructional program, the teacher taught learners to disregard their L1 knowledge as irrelevant to English sarcasm and added new knowledge associated with the target concept. Excerpt 8 presents part of a political debate between a liberal and a conservative on a U.S. radio show. Learners were asked to identify visual cues that constituted sarcastic meaning and write an interpretive essay as homework. Excerpt 8: Example of sarcasm (Political debate, YouTube) 1 Redner: you called me a liar, I’m not a liar, I called you fat, you ARE Katz: [and you’re very aggressive, 2 Redner: FAT. ((shaking his open left hand up and down at Katz)) 3 4 Katz: [and you’re very aggressive, you’re very aggressive, and it’s 5  obvious that you wanna hit me, but you haven’t done it yet, so we’re all waiting, ((raising eyebrows, opening hands, 6  and tilting head)) um my advice to you is that you need to 7 work on that control issue, and let’s keep talking about 8 the issues= 9 10 Redner: =but am I a liar? During the instruction held in week 4 (Excerpt 9), one of the students, Mia, asked about the difference between sarcasm and name-calling (lines 2-3). She was unsure whether the utterance ‘you are fat’ (lines 1-4 of Excerpt 8) is sarcasm or not. The tutor reminded her of the core characteristic of sarcasm— irony or reversed-meaning—by showing her SCOBA 2, with which the tutor had

334 

 Jiyun Kim

discussed the relationship between sarcasm and irony during weeks 2 and 3. The tutor also mentioned that speakers usually incorporate irony as a strategy to create sarcastic utterances (lines 4-6, Excerpt 9). Mia signaled her acknowledgement and understanding by nodding (lines 7 and 11). The tutor translated the transcript of the video into Korean by using particular word endings and prosodic cues to highlight the sarcastic nuance embedded in lines 5 to 8 of Excerpt 8. However, Mia (Excerpt 9) went back to the original question she had asked and more explicitly indicated where her confusion resided—whether overtly calling the interlocutor fat with negative intent (e.g., to ridicule) should be considered sarcasm or not (lines 28-30). The tutor once again informed her that the reversed-meaning aspect is the key component of sarcasm and provided a more explicit instruction. She asked Mia to identify if irony occurs between the literal meaning of the speaker’s utterance, ‘you are fat,’ and the reality, or how the speaker and the interlocutor would perceive reality (lines 31-34). Excerpt 9: Mia, STI interaction in week 4 1 Tutor: did you identify any sarcasm by any chance? 2 Mia: sarcasm + but should I consider this part as sarcasm or simply 3 ridiculing? ((pointing to the script of the video)) 4 Tutor: ((checking the sentence Mia is pointing to, pulls out SCOBA 2, and 5 points to the figure on top)) remember one of the frequently used 6 strategies to make sarcastic utterances was using verbal irony? 7 Mia: oh yes yes ((nodding)) 8 Tutor: and so if it’s sarcasm it’s highly likely that there’s some actual 9 intended meaning that seems different from the literal meaning of the utterance 10 mhm:: ((nodding)) 11 Mia: [Lines omitted in which the tutor explains meanings of each utterance in the script. She uses Korean equivalent translations with particular prosodic cues that convey the similar nuanced meanings and attitudes that the L2 utterances do.] the issue that I keep being confused about is that (.) if you call 28 Mia: a skinny person fat, it could be considered as sarcasm but 29 since he called a fat person fat, since it is the fact, maybe it is 30 31 not sarcasm? I’m not so sure 32 Tutor: yes there are a lot more instances of sarcasm that contain the 33 reversed- meaning aspect but do you see any meaning 34 difference between the literal meaning of what he said and the 35 reality? or like you said with the fact?



Teaching language learners how to understand sarcasm in L2 English 

 335

[Lines omitted in which the tutor further clarifies ‘the reality’ and ‘the fact,’ which, after all, are what is understood as real and factual in the culture of the community to which the interlocutors belong.] Concretization of the areas where irony occurs Next, the tutor further elaborated on how the aspect of irony is manifested in different instances of sarcasm, as exemplified in Excerpts 4 and 10. Excerpt 10: Ironic sarcasm from Friends [Chandler is in an ATM vestibule. He starts to walk towards the door to leave, and the power goes off, the lights go out, then flicker back on.] what? ((grabs the door handle and shakes it, the doors are 1 Chandler: 2 locked)) (.) oh great. this is just- ((turns and sees that he is 3 locked in the ATM vestibule with a beautiful woman, the 4 famed Victoria’s Secret model Jill Goodacre)) (...) ((turns back and makes a gesture with his fists 5 6 tightened and raised upward.)) ++ °great° ((mouths to himself quietly with a smile)) 7 The tutor pointed out that a contrast occurs between the utterance in line 2 (Excerpt 10) and (what Chandler perceives to be) the reality, as well as between Bob’s utterance (Excerpt 4) and what the friend understands the interlocutor would consider a lucky day. The tutor also explained possible underlying meanings of sarcasm in each excerpt using SCOBA 1, where diverse speakers’ emotions and communicative goals are illustrated on a continuum that ranges from negative to positive. Chandler’s frustration (and possibly anger) might have triggered him to utter (a) in order to release his stress. Whether it was the speaker’s habitual speech act or not, it is understood that Chandler is being sarcastic about an unfortunate situation in which he is helpless. This interpretation derives primarily from the verbal irony he incorporated in ‘oh great’—saying the opposite of the reality, which is generally understood to be a less-than-ideal situation. The same principle of how to incorporate verbal irony in formulating sarcastic utterances applies to Bob’s utterance in Excerpt 4. Bob is describing what may be the worst day of the interlocutor’s life as ‘the luckiest day of your life’ to show his empathy for the interlocutor and to express sarcasm towards the multiple series of unfortunate situations that the individual had to experience that day. After the discussion on irony as a strategy to produce sarcasm, the tutor brought Mia’s attention to lines 4 to 6 of Excerpt 8, where Katz states ‘and it’s obvious that you wanna hit me, but you haven’t done it yet, so we’re all waiting’ with a particular facial expression, gesture, and body movement (i.e., raised

336 

 Jiyun Kim

eyebrows, open hands, and tilted head). This conversation is detailed in the following excerpt: Excerpt 11: Mia, STI interaction in week 4 (Continued from Excerpt 9) Tutor: then, what did you think about this utterance? ((pointing to the 1 2 utterance we’re all waiting in line 5 of Excerpt 4)) 3 Mia: well (.) not so much. not so sure. [Lines omitted where the tutor again provides the Korean equivalent translation.] 7 Mia: (.) does it mean we are waiting for you to verbally argue in a 8 more aggressive manner? 9 Tutor: hmm: 10 Mia: uhm (.) ((tilting her head with curiosity)) isn’t the situation 11 about two politicians arguing regarding their political opinions? 12 Tutor: yes there was some of those in the beginning I suppose. but it 13 got to the point where they call each other fat and a liar. talking 14 about appearances and temper issues. which you can’t even 15 imagine to happen on Korean mass media right? 16 Mia: yes ((laughing)) indeed. I can never understand what he’s trying 17 to say. ((looking down to read the script)) (.) I thought it was a 18 comic parody show or something. (.) then this part we’re all 19 waiting, (.) 20 Tutor: yes what do you think about that part? 21 Mia: ((utters the meaning of the utterance in Korean silently)) does it 22 mean the actual meaning of hitting? it doesn’t look like it 23 Tutor: is that so? 24 Mia: yes 25 Tutor: do you think the speaker and other people except for the interlocutor are actually waiting for the interlocutor to hit the 26 speaker? 27 mhm (.) is he aiming to irritate the interlocutor continuously so 28 Mia: 29 that he finally gets exploded? that could be one of the possible goals 30 Tutor: I think that’s his intent 31 Mia: 32 Tutor: okay. well it’s not like he is really waiting for that moment 33 where this guy hits him right? who would want to get punched 34 by anyone? 35 Mia: right. I don’t think so either. so this utterance could sound sarcastic to many people, 36 Tutor: 37 especially since the speaker is saying something different from



Teaching language learners how to understand sarcasm in L2 English 

 337

38 his real intent and the goals he’s trying to achieve is not to wait 39 for the guy to hit him but to make fun of the fact that he has a bad 40 temper and maybe to keep annoying him to see if he explodes 41 like you pointed out mhm ((nodding)) 42 Mia: 43 Tutor: ((pointing to the top continuum of SCOBA 1 using a pencil)) so his 44 utterance obviously was triggered by his negative feelings 45 towards this guy and where would you find the actual goal here? 46 47 Mia: ((looks at the SCOBA)) (.) ((points to the bottom continuum of SCOBA 1 using a pencil)) (.) 48 Tutor: 49 maybe to mock and make fun of? ((looks at Mia))= =mhm ((nodding)) 50 Mia: 51 Tutor: or like you just pointed out in order to intentionally provoke 52 him= ((writes/adds the intent on the SCOBA)) Mia: =yes 53 54 Tutor: so when you actually look at his facial expression and gesture, 55 ((pulls in her laptop to play the clip again)) ((leans forward to move her face closer to the screen of the laptop)) 56 Mia: 57 Tutor: take a look 58 Video: […] so we’re all waiting 59 Tutor: [that moment. ((stops the video)) 60 Mia: his eyebrows are rising ((laughs)) and [((replays the video)) 61 Tutor: 62 did you see? 63 Mia: so we’re all waiting ((mimics the prosody, raised eyebrows, raised lip corners, open arms and hands, and tilted head)) 64 the one that I told you about 65 Tutor: ((uses her finger to raise her lip corner)) raising one lip corner? 66 Mia: 67 Tutor: yes. so the lip corners get tightened, the head gets tilted like this,= =((tilts her head the same way)) yes yes 68 Mia: and so did you see all those? 69 Tutor: 70 Mia: yes ((laughing)) so the visual cue combinations are ((pulls out SCOBAs 3 to 7 and 71 Tutor: displays them in front of Mia)) (.) so starting with the face, what 72 about the facial expression? 73 74 Mia: ((points to one of the face pictures presented on SCOBA 3)) When asked what she thought about the utterance, Mia was unsure if it is sarcasm or not (line 3). Then, after hearing the translation again and contemplating it a

338 

 Jiyun Kim

bit, she asked if the verb ‘hit’ is used in a figurative way with the meaning ‘to argue aggressively’ (lines 7-8). The tutor encouraged her to think more (line 9) and then attempted to confirm her understanding of the context—whether it is a situation in which two politicians are arguing about political issues (line 10). The tutor then explained how the discussion escalated to the point where the two show intense animosity and attack each other by commenting on appearances and temper issues (lines 12-14). Mia’s response (lines 14-18) reveals how difficult it was for her to understand the nature of the interaction and the speakers’ utterances. Her confusion originated in her reasoning based on her L1 schema: (1) her expectations about the genre of ‘political debates aired on mass media’ and (2) incorrect understanding of sarcasm as involving metaphor and figurative language use. Mia’s interpretation of political debates centered on her corresponding L1 concept, related to the fact that provocative language use and violent antagonistic behaviors is much more restricted in Korean mass media than in the U.S. Focus on speaker intent and communicative goals Then the tutor (Excerpt 11) informed Mia of the cultural difference regarding the level of regulation in mass media between the two settings—promoting inhibition of her L1 knowledge and addition of L2 encyclopedic knowledge regarding a particular genre of discourse. Mia began to focus more on figuring out the meaning of the utterance ‘we’re all waiting’ (line 18). However, her interpretation stayed at the level of word usage and hinged on the possibility that the verb ‘hit’ (line 5 of Excerpt 8) might have been used figuratively (lines 20-23). The tutor now brought her focus to figuring out the possible goal of the utterance—why the speaker is uttering it—and explained the contrast between the literal meaning of the utterance ‘we’re all waiting’ and the actual reality (i.e., we—Katz and other people— are not actually waiting for Redner to hit Katz) (lines 24-25). Mia then came up with a possible speaker intent, which is to keep aggravating the interlocutor until he loses control (lines 26-29). The tutor again discussed the issue of verbal irony, but this time by highlighting the contrast between the literal meaning of the utterance and the encyclopedic knowledge that Katz assumes Redner has—Katz clearly is aware that Redner also knows no one wants to get punched by anyone, especially when they are being video-taped and aired on a mass media show. The tutor addressed this fact (that is, she noted that Katz does not want to be punched by Redner but chooses to say it) by asking a question in line 31. Mia agreed with this view (line 32), and the tutor explained why general interlocutors would find the utterance sarcastic—primarily due to the reversed-meaning component that signals the speaker’s negative intent (i.e., to mock the interlocutor in lines 33-36) along with combinations of other cues (i.e., visual, prosodic, and contextual) that are conventionally used to convey sarcastic attitudes.



Teaching language learners how to understand sarcasm in L2 English 

 339

Interaction with other participants followed the same principle in which the tutor focused on the essential features of sarcasm by (1) first clarifying the concept through SCOBA and then (2) deconstructing examples of sarcastic utterances into visual, prosodic, and other contextual cues. Due to limitations of space, this chapter presents excerpts from the STI interaction with Mia only to illustrate how exactly mediation occurred during the instruction. 6.3 Concept development Contrary to the incomplete and non-theoretical knowledge they provided prior to the start of STI, the learners demonstrated mature understanding of the concept during the final stage of the program. This concept development originated from the ability of learners to conceptualize sarcasm through both theoretical and experiential lenses. Learners were able to appropriately define sarcasm after acquiring the essential components—irony, relevant cues, and speaker intent. During the post-STI interview (Excerpt 12), unlike the attitude he exhibited during the pre-STI interview, learner Wang defined the concept with greater confidence. He particularly indicated the need to consider various factors to understand sarcasm in its many facets: the multiple cues to consider for detection and comprehension of sarcasm, various communicative goals and speaker intent, and different levels of harshness that exist. In line 2, Wang explained how— before STI interaction—he used to simply focus on words or expressions that contained the opposite meaning of the actual situation or speaker’s intent in order to detect sarcasm. He demonstrated his functional understanding of sarcasm by acknowledging the existence of diverse speaker intents and communicative goals for using sarcasm. Additionally, he showed awareness of different ranges of rudeness that interlocutors need to consider in order to identify speakers’ intentions (lines 3 to 6). He also reported that the concept of sarcasm had become broader for him due to his increased awareness of both diverse speaker intentions and various kinds of sarcasm-related cues (lines 28 to 30). Excerpt 12: Wang, post-STI interview so before 1 Tutor: 2 Wang: [it is because before I simply thought of meanings that are opposite, and meanings to attack someone but now I have to 3 4 think about the level of severity if it’s to attack someone and 5 also even though it’s not to attack someone there seem to be cases in which you use it just to release your stress and so forth 6 7 Tutor: I see

340 

 Jiyun Kim

13 Wang: but I guess there is one common thing which is to move beyond the [lines omitted in which Wang provides examples of possible other speaker intent] 14 surface meaning 15 Tutor: ahh: so you think there is always some kind of underlying meaning for sure (in sarcastic utterances). 16 Wang: ((nods)) yes. so the common thing is that it always moves beyond 17 the surface meaning and there is always some real intent behind it= 18 =and if you want to understand what that intent is you have to consider the situation, you have to see the speaker’s face and 19 20 listen to his/her voice, and so on [lines omitted in which the tutor reiterates what Wang had said] 26 Wang: 27 28 29

so it’s an act of conveying underlying intended meaning through skills of using certain expressions or using some kind of methods? I think it’s a more expanded meaning than what I used to think in the past ((nods))

Additionally, through lines 19 to 21, Wang explained how it is necessary to consider various cues (i.e., contextual, prosodic, and visual) in order for an interlocutor to understand the intended meanings of sarcastic utterances. It is noteworthy that Wang provided this knowledge even before the tutor asked him to discuss the types of cues he would consider for sarcasm comprehension. His readiness to comment on sarcasm-relevant cues while defining the term (i.e., latching in line 19) shows his development since the pre-STI interview. In the earlier interview, he defined sarcasm as ‘utterances you would feel bad about after thinking about them’ without providing any particular types of cues before being asked. Mia (Excerpt 13), during the post-STI interview, was able to include some central tenets of English sarcasm: its ubiquity and its association with humor. Excerpt 13: Mia, post-STI interview Tutor: how would you define sarcasm= Mia: =well first of all it looks like it’s their (Americans’) characteristic to use sarcasm in their everyday life, Tutor: ((nods)) Mia: and secondly they seem to want to pursue the quality of wittiness by using it mhm what made you think about that intent, being witty? Tutor: Mia: I sort of realized that while doing homework and during class. and you know through those examples we saw? I see Tutor:



Teaching language learners how to understand sarcasm in L2 English 

 341

Another learner, Lee (Excerpt 14), indicated that he now focuses on what speakers intend to accomplish and stated that his attention had shifted to figuring out the goals of the utterances rather than the literal meaning of expressions. Excerpt 14: Lee, post-STI interview rather than simply focusing on generally understood meanings of Lee: expressions (.) I now focus more on goals?= =well during the first week I thought of things through the notion of irony or ironic expressions? but now I think about goals. the (communicative) goals that sit at the bottom (of utterances). I see things based on that now Importantly, learners’ increased awareness and ability to discern subtle nuances of messages enabled them to have a positive attitude towards L2 learning. Wang (Excerpt 15), for example, reported that his attitude towards English learning had changed after receiving STI. His increased sensitivity (lines 2-8) and ability to analyze communication (lines 12-20) enabled him to more readily understand the actual meanings that speakers convey. This ultimately made him perceive English learning as a fun activity (line 22). Excerpt 15: Wang, post-STI interview 1 Tutor: okay, then do you see any changes before and after,= 2 Wang: =sensitivity 3 Tutor: oh yes? 4 Wang: yes definitely ((nods)) I became a lot more sensitive to situations 5 Tutor: you mean the conversations that you participate in? or [not only that but also the situations that I happen to hear and 6 Wang: watch when other people talk to each other. 7 I see you mean in English? 8 Tutor: 9 Wang: yes 10 Tutor: mhm (.) in what ways do you think you became more sensitive? maybe more analytic? it’s fun to analyze people’s talk like that 11 Wang: 12 Tutor: hmm 13 Wang: and so I get to think like oh: they say these things to mean those things 14 15 Tutor: mhm: 16 Wang: if it was like before I wouldn’t have cared much= 17 =I would let them all pass. they would just pass by. but now I 18 hear those things that I didn’t used to hear and that makes me pay more attention to what people say and go wow they use these 19 20 expressions to say this?

342 

 Jiyun Kim

21 Tutor: 22 Wang:

I see and so English became more fun to me now. it’s fun.

Overall, learners’ post-STI interviews showed that the STI promoted development in their overall conceptual understanding of sarcasm. Before instruction, learners’ understanding was primarily grounded in their L1 understanding of sarcasm. However, throughout the course of the STI, many learners re-conceptualized sarcasm by considering its actual use, examples, and the functional goals it achieves.

7 Conclusion This study examined the teachability of the concept of sarcasm in English as a second language. As such, it focused on the extent to which a specific pedagogical approach grounded in Vygotskian principles of developmental education is effective in assisting learners in grasping the concept of sarcasm in English, including how it is signaled through a variety of linguistic and extra-linguistic means, and the extent to which learners are able to use this conceptual knowledge to detect and appropriately interpret sarcasm in English. Analysis of student verbalization data showed evidence of concept development. Learners were able to connect a theoretical essence (i.e., a SCOBA) and concrete phenomenon (i.e., actual manifestations of sarcasm in videos), which resulted in holistic and genuine understanding of the concept of sarcasm. The interactions with learners included a series of mediations to help them clarify ambiguities and better understand the concept of sarcasm. For example, Mia initially struggled with distinguishing sarcasm from two other ideas: name-calling and metaphor. The tutor disabused Mia from relying heavily on her prior knowledge of sarcasm, which centered on metaphor and figurative language. Instead, the tutor encouraged Mia to shift her orientation towards the central property of sarcasm—irony. Throughout this process of concept clarification, the tutor highlighted specific examples of reversed-meaning in multiple video clips depicting sarcasm. Despite the incompleteness of their frameworks at the outset of the program, participants found their own ways to appropriate the concept by showing unique developmental trajectories. Wang placed a greater emphasis on the need to include various kinds of sarcasm-related cues in defining sarcasm, which contrasts with the more limited definition he provided during the pre-STI interview. He not only demonstrated greater confidence in defining sarcasm but also developed a positive attitude towards English learning. Mia added new components—ubiquity and humor, both of which are highly associated with English sarcasm—while she re-defined the concept after the STI. By adding



Teaching language learners how to understand sarcasm in L2 English 

 343

L2 sarcasm-related components and inhibiting unrelated components, the STI provided learners with an analytical lens that strengthened the quality of their understanding of English sarcasm. Additionally, this framework enabled some learners to re-conceptualize their overall objectives of English learning and to develop a new approach to L2 communication. As learners like Lee stated, the focus in their language learning shifted from finding literal translations of expressions to figuring out speakers’ intents and the functional goals of their utterances. The findings confirmed a dialectical relationship between scientific knowledge and spontaneous knowledge in which each influences the other (Vygotsky 1978). Learners developed a mature conceptual understanding of sarcasm, which transformed their analytical ability and thus their control over sarcastic contexts. This improved performance recursively reconstructed and completed their theoretical knowledge.

8 Implications The current study produced generally positive results. However, limitations remain in terms of its scope of applicability. The STI interactions in this study occurred in one-on-one tutoring sessions. Practitioners may find it difficult to apply the same intensive approach to a number of students within a larger classroom context. Of course, this does not indicate that it is impossible to apply the same types of meditational support in language classrooms. Although not presented in this chapter due to limited space, learner development in the study occurred mostly through communicative thinking not only with the tutor but also with the learners’ peers. Nevertheless, a greater level of sensitivity may be required from the teacher in order to effectively mediate a group of learners while remaining aware of group ZPDs (Vygotsky 1998). An ESL classroom comprised of students with mixed L1s may be even more challenging for practitioners who aim to implement the curriculum used in this study. As demonstrated above, it is important for the mediator to acknowledge students’ previous understandings of the concept of sarcasm. This initial knowledge is likely to be based on the students’ L1 understanding of the concept, which will be different according to their cultural backgrounds. Teachers might obtain an understanding of learners’ L1 equivalent concepts of sarcasm from various types of homework assignments (e.g., have learners bring video examples of L1 sarcasm and write an interpretive essay about them). However, the amount of the teacher’s knowledge may vary depending on the degree of similarities and differences between the particular L1 and L2 concepts of sarcasm. A teacher’s lack of knowledge in students’ particular L1s may not effectively enable him or her promote their conceptual

344 

 Jiyun Kim

development. This is because a teacher’s degree of knowledge about learners’ pre-understanding and their psychological processes matters—it determines the quality of the dialectics between the teacher and learners. The medium of instruction in this study was the participants’ L1. The tutor was able to interact with the students during communicative thinking in the L1, and the students were also able to use their L1 in group communicative thinking. This allowed the students to express their thoughts and feelings freely and in great detail, which heavily influenced the quality of the teacher-student dialectics. However, in an ESL classroom with linguistic diversity, the instruction would likely be held in English, which may be a great challenge for both the teacher and students when the circumstances demand rigor in analysis and sophistication in verbalization. It remains to be seen how instruction in a concept such as sarcasm unfolds in this type of environment.

Transcription Conventions (.) Short pause (about one beat); additional periods denote longer pause [ Overlapping speech (text) Unclear speech . Falling, final intonation , Slightly rising, continuing intonation ? Rising intonation (not necessarily a question) : Sound lengthening; additional colons denote longer lengthening = Latching (no pause between words) ↑ Higher pitch relative to surrounding talk ↓ Lower pitch relative to surrounding talk – Abrupt cutoff Faster speech relative to surrounding talk >text< Slower speech relative to surrounding talk h Audible out-breath; additional hs denote longer out-breath .h Audible in-breath; additional hs denote longer in-breath @ Laughter (about one beat); additional @ denote longer laughter



Teaching language learners how to understand sarcasm in L2 English 

→ ((text)) Text

Indicates line of interest in the discussion Transcription notes Text in italics indicates a foreign or referred to word

 345

References Agar, Michael. 1994. Language shock: Understanding the culture of conversation. New York: Morrow. Arievitch, Igor M. & Anna Stetsenko. 2000. The quality of cultural tools and cognitive development: Galperin’s perspective and its implications. Human Development 40. 69–92. Attardo, Salvatore. 2000. Irony as relevant inappropriateness. Journal of Pragmatics 32. 793–826. Austin, John Langshaw. 1962. How to do things with words. Oxford: Oxford University Press. Bell, Nancy. 2005. Exploring L2 language play as an aid to SLL: A case study of humour in NS-NNS interaction. Applied Linguistics 26(2). 192–218. Bell, Nancy. 2006. Interactional adjustments in humorous intercultural communication. Intercultural Pragmatics 3(1). 1–28. Bouton, Lawrence F. 1994. Conversational implicature in a second language: Learned slowly when not deliberately taught. Journal of Pragmatics 22(2). 157–167. Bouton, Lawrence F. 1999. Developing nonnative speaker skills in interpreting conversational implicatures in English: Explicit teaching can ease the process. In Eli Hinkel (ed.), Culture in second language teaching and learning, 47–70. Cambridge: Cambridge University Press. Briggs, Jean L. 1970. Never in anger: Portrait of an Eskimo family. Cambridge, MA: Harvard University Press. Claridge, Claudia. 2010. Hyperbole in English: A corpus-based study of exaggeration. New York: Cambridge University Press. Davies, Catherine Evans. 2003. How English-learners joke with native speakers: An interactional sociolinguistic perspective on humor as collaborative discourse across cultures. Journal of Pragmatics 35. 1361–1385. Ekman, Paul. 1984. Expression and the nature of emotion. In Klause R. Scherer & Paul Ekman (eds.). Approaches to emotion, 319–344. Hillsdale, NJ: Erlbaum. Ekman, Paul. 1993. Facial expression and emotion. American Psychologist 48. 384–392. Ekman, Paul, Wallace V. Friesen & Maureen O’Sullivan. 1987. Universals and Cultural Differences in the Judgments of Facial Expressions of Emotion. Journal of Personality and Social Psychology 53(4). 712–717. Ekman, Paul & Wallacw V. Friesen. 1969. Nonverbal Leakage and Clues to Deception. Psychiatry 32(1). 88–106. Gal’perin, Piotr. Ia. 1989. Organization of mental activity and the effectiveness of learning. Soviet Psychology 27(2). 45–64. Gal’perin, Piotr. Ia. 1992. Stage-by-stage formation as a method of psychological investigation. Journal of Russian and East European Psychology 30(4). 60–80. Giora, Rachel. 2004. On the graded salience hypothesis. Intercultural Pragmatics 1. 93–103. Goffman, Erving. 1974. Frame analysis: An essay on the organization of experience. Boston, MA: Northeastern University Press.

346 

 Jiyun Kim

Kim, Jiyun. 2013. Developing conceptual understanding of sarcasm in a second language through concept-based instruction. University Park, PA: The Pennsylvania State University dissertation. Kim, Jiyun. 2014. How Korean EFL learners understand sarcasm in L2 English, Journal of Pragmatics 60. 193–206. Haenen, Jacques. 1996. Piotr Gal’perin: Psychologist in Vygotsky’s footsteps. New York: Nova Science Publishers. Haenen, Jacques. 2001. Outlining the teaching-learning process: Piotr Gal’perin’s contribution. Learning and Instruction 11. 157–170. Lantolf, James P. 2006. Sociocultural theory and L2. Studies in Second Language Acquisition 28. 67–109. Lantolf, James P. & Matthew E. Poehner. 2014. Sociocultural theory and the pedagogical imperative. Vygotskian praxis and the theory/practice divide. New York: Routledge. Miller, Ronald. 2011. Vygotsky in Perspective. New York: Cambridge University Press. Oboukhova, Ludmila. F., Alexander V. Porshnev, Elena R. Porshneva & Sogia A. Gaponova. 2002. ‘Konstruirovanie komp’uternoj obuchayushej programmy na osnove teorii P.I Gal’perina’. [Constructing a concept-based instructional program on the basis of P. Y. Gal’perin‘s theory]. Voprosy Psikhlogii 5. 103–114. Partington, Alan. 2007. Irony and reversal of evaluation. Journal of Pragmatics 39. 1547–1569. Poehner, Matthew E. 2009. Group Dynamic Assessment: Mediation for the L2 Classroom. TESOL Quarterly 43(3). 471–491. Ratner, Carl. 2000. A cultural-psychological analysis of emotions. Culture and Psychology 6. 5–39. Searle, John R. 1969. Speech acts. Cambridge: Cambridge University Press. Searle, John R. 1979. Expression and meaning. Cambridge: Cambridge University Press. Serrano-Lopez, Maria & Matthew E. Poehner. 2008. Materializing linguistic concepts through 3–D clay modeling: A tool-and-result approach to mediating L2 Spanish development. In James P. Lantolf & Matthew E. Poehner (eds.), Sociocultural theory and the teaching of second languages, 321–350. London: Equinox. Sperber, Dan & Deirdre Wilson. 1981. Irony and the use-mention distinction. In Peter Cole (ed.), Radical pragmatics, 295–318. New York: Academic Press. Sperber, Dan & Deirdre Wilson. 1992. On verbal irony. Lingua 87. 53–76. Silverstein, Michael. 2003. Indexical order and the dialectics of sociolinguistic life. Language and Communication 23. 193–229. Swain, Merrill. 2006. Languaging, agency and collaboration in advanced language proficiency. In Heidi Byrnes (ed.), Advanced language learning: The contribution of Halliday and Vygotsky, 95–108. Washington, D.C.: Georgetown University Press. Ushakova, Tatiana N. 1994. Inner Speech and Second Language Acquisition: An ExperimentalTheoretical Approach. In James P. Lantolf and Gabriela Appel (eds.), Vygotskian Approaches to Second Language Research. 135–156. Norwood, NJ: Ablex. van Compernolle, Remi Adam. 2014. Sociocultural theory and L2 instructional pragmatics. Bristol: Multilingual Matters. Vygotsky, Lev S. 1987. The collected works of L. S. Vygotsky: Vol 1. New York: Plenum. Yus, Francisco. 1998. Irony: context accessibility and processing effort. Pragmalinguistica 5–6. 391–411. Yus Ramos, Francisco. 2000. On reaching the intended ironic interpretation. International Journal of Communication 10(1–2). 27–78. Zhang, Xian. 2014. The teachability hypothesis and concept-based instruction : Topicalization in Chinese as a second language. University Park, PA: The Pennsylvania State University dissertation.

Natalie Lefkowitz and John S. Hedgcock

13 Anti-language: Linguistic innovation, identity construction, and group affiliation among emerging speech communities Abstract: The invention and use of anti-language can enable language users to resist norms, provoke mainstream disapproval, strengthen in-group solidarity, cultivate covert prestige and peer approval, and exclude (“other”) non-initiated individuals. This chapter examines anti-language use in language contact, social media, and L2 development settings where perceived nonstandard patterns implicitly or explicitly compete with standard norms. To explore patterns of linguistic defiance, we analyze qualitative and quantitative data from our prior research on: (1) the French language game Verlan; (2) novel registers evolving among francophone social media users; (3) deliberate underperformance behaviors among classroom L2 learners; and (4) Spanish heritage-language learners’ appropriation of high- and low-prestige varieties. Analyses reveal conflicts between users’ recog­ nition of “standard” linguistic norms and practices that contravene normative standards by crossing code boundaries and merging registerial features. Binary contrasts between high-prestige and vernacular styles, public and private communication, and orality and literacy reinforce paradoxical tensions, which we examine and compare. A synthesis and reframing of these divergent research strands suggests novel directions for systematic exploration of anti-language in relation to linguistic innovation, identity construction, and insider status among emerging speech communities.

1 Introduction Social and multilingual turns in studies of language development have precipitated decisive shifts in our thinking about language, language use, and language learning (Block 2003; May 2013; Rampton 2013). Research on language contact and change, identity construction, and translingual practice has thus influenced conceptualizations of language learning and use. In particular, parallels between individual language development processes and macro-level language change and innovation have highlighted learners’ efforts to construct social identities, engage in audience design, and achieve legitimacy as members of multiple speech communities and discourses (Blackledge and Creese 2010, 2013; DOI 10.1515/9781501503993-014

348 

 Natalie Lefkowitz and John S. Hedgcock

­ lommaert 2010, 2013; Ellis and Larsen-Freeman 2009; García and Wei 2013; B Larsen-­Freeman and Cameron 2008; McWhorter 2011; Norton 2013). One means by which emerging speech communities realize these socio-affective goals involves the use of anti-language, a productive linguistic tool deployed strategically to establish and maintain counter-realities, express opposition to mainstream practices and values, and demonstrate social resistance. Like comparably innovative forms of language play, anti-languages often emerge as social dialects or registers within marginalized groups of language users whose standing relative to society is precarious or liminal (Montgomery 2008). In this chapter, we examine evidence of anti-language use in settings where standard and nonstandard linguistic norms compete – namely, in language contact, L2 learning, and social media spaces in which authority governing acceptable language use can be overtly and covertly contested. Drawing on our prior research, we focus chiefly on communication practices that reflect sociolinguistic defiance and deviation from perceived mainstream norms, impulses that underlie and connect creative behaviors observed in a range of communicative settings. The data that we will examine derive from our studies of: (1) the French language game Verlan; (2) novel registers evolving among francophone social media users; (3) deliberate underperformance in the oral production behaviors of classroom learners of French and Spanish; and (4) Spanish heritage-language learners’ appropriation of standard and vernacular varieties (Hedgcock and Lefkowitz 2000, 2011, 2016; Lefkowitz 1989, 1991, 2003, 2011; Lefkowitz and Hedgcock 2002, 2006; O’Connor and Lefkowitz 2014). By reanalyzing selected data sources, we will expose conflicts between participants’ recognition of high-prestige linguistic norms and language practices (overt and covert) that contravene normative standards. These innovative practices often involve the crossing of code boundaries and the blending of registerial resources to contest prevailing norms, create counter-identities, and produce “alternative realities” (Halliday 1976).

2 Defining anti-language Anti-language, a construct formalized by Halliday (1976), refers to a linguistic variety generated by an “anti-society” that wishes to assert its uniqueness relative to a superordinate discourse community and its “everyday language.” An antilanguage can construct “social structure of a particular kind, in which certain elements are strongly foregrounded.” Such foregrounding imbues an anti-language with “a special character in which metaphorical modes of expression” emerge at the phonological, lexicogrammatical, and semantic levels (570). Montgomery



Anti-language: Innovation, identity construction, affiliation 

 349

(2008) characterized anti-languages as “extreme versions of social dialects” that evolve among marginal and precarious subcultures (113). These subcultures (or anti-societies) may even be perceived as antagonistic to a society. As creative semiotic systems, anti-languages set subcultures apart from (or in opposition to) mainstream discourse communities. They also strengthen group solidarity while excluding the Other – making anti-languages and anti-societies “impenetrable to outsiders” (Montgomery 2008: 115). For example, Smitherman (2000) portrayed Black English as an anti-language that signifies a “counter-ideology” and rebellion against oppression. Nonetheless, as Montgomery (2008) stipulated, an antilanguage should not be construed “as an absolute category to which particular varieties must conform on an all-or-nothing basis.” Rather, an anti-language may center on “an idea to which given instances approximate more or less closely” (118). The anti-language categories that we examine exemplify prototypical antilanguage functions and features to varying degrees and in localized ways. Initial investigations of anti-languages and anti-societies focused chiefly on subcultures positioned “outside the law,” such as prisoners, criminal gangs, and members of religious sects and cults (Montgomery 2008). Examples include studies of: various argots (Bullock 1996; Nielsen and Scarpitti 1995); Gypserka, Polish prisoner talk (Podgórecki 1973); pelting speech, the Elizabethan English of street vagabonds (Halliday 1976); the language of the Calcutta underworld (Mallik 1972); Polari, the secret language of gay men (Baker 2002); and CB radio slang (Montgomery 1986). Anti-languages arising in other liminal speech communities extend to contemporary rap and hip-hop music, expressive forms produced and claimed by marginalized ethnic minorities and social groups (Alim, Ibrahim and Pennycook 2009; Blommaert 2010, 2013; Cutler 2014; Montgomery 2008; Pennycook 2010; Terkourafi 2010). Anti-language markers can be classified in terms of the full range of linguistic structures. As an anti-language develops, its users may engage in generative processes that transgress conventional boundaries, thereby producing novel oral and written registers (Crystal 2011). The list below presents the productive processes that we will examine in four contexts of language use: –– Phonological and morphological innovation (e.g., mispronunciation, including the self-conscious violation of normative phonological patterns; playful application of productive processes such as metathesis; invention of novel verb inflections); –– Relexicalization (e.g., partial relexicalization, the substitution of novel word forms for existing word forms, a process of indexical reassignment and  semantic expansion; overlexicalization, the introduction of alternative and completely novel lexical items to express meanings within an anti-­society);

350 

 Natalie Lefkowitz and John S. Hedgcock

–– Lexical borrowing from outside the matrix (superordinate) language or standard variety; –– Formal simplification (e.g., reduction of word- and phrase-length forms to abbreviations, acronyms, and initialisms; contractions; minimal punctuation; truncation; use of emoticons and emojis as substitutes for prosodic clues); –– Taboo language (e.g., the use of obscenities, vulgarity, and argot – criminal vocabulary designed to be unintelligible to outsiders); –– Registerial blurring and cross-over (e.g., the merging of oral and written features in messages destined for a single medium, such as digital spaces; transiting across registers or straddling boundaries of appropriateness; violating appropriateness norms for register choice).

3 Anti-language in emerging speech communities Members of emerging speech communities engage in anti-language behaviors both to resist external pressure and to enhance internal cohesion. Electing an oppositional code can cultivate solidarity and affirm separateness from the dominant speech community in terms of language, neighborhood, ethnicity, culture, age, social class, occupation, and sexual orientation. Francophones who elect to use Verlan, for instance, may wish to demonstrate resistance to the sociopolitical values, ideologies, and economic structure of mainstream French society while expressing a distinct identity that reformulates and plays on “standard” French – a variety prized as a signifier of national identity. In the digital milieu of French social media communication, anti-language forms may reflect users’ efforts to adapt their linguistic performance to the unique dimensions of online communication and to accommodate the expectations of their audiences (Giles and Ogay 2007). These uses may likewise signify overt resistance to mainstream, “standard” linguistic practices: Francophone social media users know that their online communication practices depart from – and often defy – conventions that govern polite speech and written communication. These dynamic practices establish solidarity and maintain phatic communion with online interlocutors. Similarly, in the foreign language (FL) context, instances of deliberate underperformance among English-speaking learners of Spanish and French reflect efforts to resist the authority of the pedagogical standard in the classroom. The highly anglicized Spanish and French self-consciously produced by subgroups of students in our research signifies allegiance to subversive classroom cliques



Anti-language: Innovation, identity construction, affiliation 

 351

whose aim is to mock the target language and the teacher. Underperformance produces a kind of anti-language that is both creative and supportive of solidarity and covert prestige — a status resulting from positive attitudes toward disfavored vernacular speech norms (Holmes 2008; Labov 1972). The novel Spanish varieties produced by heritage users of Spanish in the U.S. similarly emerge organically and unconsciously, though some speakers evince an explicit desire to flout highprestige varieties and to undermine their users. The “Spanglish” attested among heritage speakers is nonetheless perceived as a variety that is something less than “standard” Spanish – and thus a threat to the latter. These low-status, hybrid varieties nonetheless identify and unify their users as members of a marginalized speech community.

3.1 Verlan as anti-language The first example that we will examine from our data is Verlan. Of the anti-­ language representations that we explore, Verlan exhibits the highest frequency of properties common to anti-languages and represents one of the “more extreme, hard-edged cases” (Montgomery 2008: 117); the remaining three categories follow in descending order. Originating as a French language game requiring the inversion of syllables, Verlan has evolved into a youth sociolect – the term itself comprising the inverse of l’envers (“backwards”) (Doran 2004, 2007; Lefkowitz 1989, 1991; Méla 1988, 1991, 1997). Verlan is a covertly prestigious code, a “linguistic bricolage” spoken by multicultural, multi­ethnic, multilingual banlieue adolescents (banlieusards) who inhabit the marginal­ ized, low-income housing projects (cités or habitations à loyer modéré [HLM]) in immigrant-populated suburbs of Paris and the outskirts of large metropolitan centers in France (Doran 2004). These dwellings and their occupants are frequently characterized by stigmatized images and predominantly negative depictions related to “ethnic alterity, criminality and violence” (Cannon 1997), as well as unemployment, delinquency, poverty, and intercultural tensions (Doran 2007; Hassa 2010). A tool for asserting identity, Verlan expresses coolness (avoir la tchatche, “having the talk”), sense of place and belonging, linguistic and cultural hybridity, solidarity, insider status, outsider exclusion (othering), and youthful camaraderie. The use of Verlan is marked by resistance to mainstream authority, values, and high-prestige (often prescriptive) linguistic norms. These tendencies are evident in Verlan’s propensity for crude vocabulary, ritual insults (les vannes), secrecy, and cross-linguistic borrowing. According to

352 

 Natalie Lefkowitz and John S. Hedgcock

Hassa (2010), the use of foreign terms “provides a certain linguistic camouflage” that makes difficult topics easier to discuss (58). In addition, these lexical ­borrowings symbolically reflect the linguistic diversity of their users, who have North African (Arab and Berber), West African, Roma (“Gypsy”), Creole, Asian, Portuguese, and Caribbean origins, among others. Sloutsky and Black (2008) referred to Verlan as “un véritable creuset linguistique,” a linguistic melting pot that incorporates expressions from hip-hop, rap, reggae, popular songs, old French slang, African-American vernacular, as well as minority community languages (Doran 2004). Describing rap music that contains Verlan, Hassa (2010) observed that the use of a “linguistic patchwork” functions as a “provocative mirror of the multicultural, multi-layered identity of banlieue youth” (62). For Verlan speakers, as well as the other emerging speech communities under discussion, reappropriating language and creating an alternative code or anti-language confers covert prestige by creating a positive self-image, a counter-­ identity, and a source of pride and self-validation in being a mainstream outsider. Thus, what might have been viewed as a linguistic stigma is effectively convert­ed into an emblem that overtly diverges from mainstream language and culture (Bourdieu 1991; Doran 2004). The need for a mechanism that confirms sociocultural identity is understandable for banlieusards, who live in an interstitial culture that blends cultural and linguistic practices from their parents’ birthplace and from the host country (Calvet 1994). Verlan constructs a “discursive ‘third space’ in which to enact . . . aspects of identity” that are distinct from mainstream French culture and that resist the dominant social order and its linguistic practices (Doran 2007: 505). Pressure to adhere to a high-prestige linguistic standard is generated by prevailing sociopolitical, cultural, and educational currents. Attitudes that privilege conformity with “standard” French dominate in France, known for its prescriptivist penchant for the mastery of standard French (le bon usage), forcefully upheld by the Académie Francaise, official legislation, conventional print and broadcast media, and the educational system. Speaking backwards – against “standard” norms – becomes a metaphor of opposition, of talking back. Méla (1988) characterized this phenomenon as a “backwards universe” where the social order undergoes a symbolic reversal. Using Verlan becomes a small-scale symbol of revolt where opposite values prevail and where sounding bad is good. The first instance of Verlan defying high-prestige linguistic convention is the use of metathesis, a phonological innovation involving the deliberate manipulation of syllables resulting in the violation of high-frequency linguistic patterns as seen in Tab. 1.



Anti-language: Innovation, identity construction, affiliation 

 353

Tab. 1: Metathesis Open Syllable Type (CVCV) Standard French branché câblé moto pourris putain salut

Verlan → → → → → →

chébran blécâ tomo ripoux tainpu lusa

Phonetic transcription [ʃebʁã] [bleka] [tomo] [ʁipu] [tɛ̃py] [lysa]

English gloss “in,” “cool” “in,” “cool” “motorcycle” “corrupt” “whore” “greetings”

Closed Syllable Type (CVCVC) Standard French balaise basket bonjour gonzesse musique

Verlan → → → → →

laiseba sketba jourbon zessegon ziquemu

Phonetic transcription [lɛzba] [skɛtba] [ʒuʁbɔ̃] [zɛsgɔ̃] [zikmy]

English gloss “hefty” “sneakers” “hello” “woman” “music”

(Based on Lefkowitz, 1989, 1991, 2003)

The second representation of defiance involves partial relexicalization and over­ lexicalization, or reverlanization, a tool for semantic innovation. This process renders lexical items less transparent to the uninitiated, and in so doing, allows word forms to convey their own new meanings (Lefkowitz 1989). Sloutsky and Black (2008) described this process of lexical renewal as “linguistic hijacking or rerouting.” Tab. 2: Reverlanization (Partial relexicalization and overlexicalization)1 arabe “Arab” [aʁab] → [aʁabə] → [bəaʁa] → [bəaʁ] = [bœʁ] → [bœʁə] → [ʁəbə] → [ʁəb] = [ʁœb] femme “woman” [fam] → [famə] → [məfa] → [məf] = [mœf]→ [mœfə] → [fəmə] = [fœmø] flic “cop” [flik] → [flikə] → [kəfli] → [kəfl] → [kəf] → [kœfə] → [kəfə] → [fəkə] → [fək] = [fœk] (Based on Lefkowitz 1989, 1991, 2003) 1 Original transcriptions, based on Méla (1988, 1991, 1997) and Lefkowitz (1989, 1991, 2003), have been edited to conform to IPA convention.

354 

 Natalie Lefkowitz and John S. Hedgcock

In the first example, the Verlan word beur, derived from arabe, shifts its reference meaning from Arab to Arab immigrant children residing in France. Arabe refers to ethnic and cultural origin, whereas beur and rebeu index an Arab-French identity (Méla 1997). As beur has penetrated standard French and lost its rebellious status, alternatives come into use, such as the reverlanized rebeu and the truncat­ed reb [ʁœb]. The Verlan term meuf refers to an adolescent woman, whereas the reverlanized form feum is sexually charged, portraying the referent as an object of desire. The final example, the French slang word flic (“cop”), becomes more irreverent and less accessible than the verlanized kefli [kəfli] and the simplified keuf [kœf], when it converts to the reverlanized feuk [fœk], a word homophonous with the English expletive “fuck.” This semantic expansion, also called jongleries langagières (“linguistic juggling”) and contre-légitimité linguistique (“lin­ guistic counter-legitimacy”), promotes solidarity by assigning positive meanings to otherwise negative associations held by dominant French culture toward racial, ethnic, and banlieusard terms connected with the cités (Goudailler 2002; Sloutsky and Black 2008). A third instance of deviation from mainstream French linguistic norms is the verlanization of lexical items referring to taboo behaviors. These terms reflect users’ marginalized, multiethnic, multicultural banlieue surroundings and often incorporate argot items that index illicit or illegal activities such as drug use, interethnic sex, violence, theft, and gang affiliation. In addition to obscenities and ritual insults, taboo language also includes foreign lexical borrowings, as the data samples in Tab. 3 show. Tab. 3: Taboo language: Criminal vocabulary Standard French braquer taper tirer chopé herbe shit shoot joint poudre

→ → → → → → → → →

Verlan

English gloss

[kebʁa] [peta] [ʁeti] [peʃo] [bə] [təʃ] [təʃu] [wɛ᷈ʒ] [dʁəpu]

“to steal” “to rob” “to rip off” “arrested” “marijuana” “hash” “fix” “joint” “heroin”

(Based on Lefkowitz 1991: 117)

Not surprisingly, in the face of the well-attested French desire for linguistic purity and the pervasive standard language myth (Lippi-Green 2012), these defiant uses



Anti-language: Innovation, identity construction, affiliation 

 355

of Verlan attract strong mainstream disapproval, as attested in the opinions of educated speakers of standard French presented in Tab. 4.2 Tab. 4: Perceptions of Verlan as expressed by speakers of standard French – I never use Verlan. This language is killing French! – I don’t know it and I don’t want to know it because it is not useful and not very beautiful! – I never use it because it is unnatural, it is impolite, and it deforms the language. (Le Verlan n’est pas du français. C’est n’importe quoi et je ne vois vraiment pas à quoi ça sert.) (“Verlan isn’t French. It’s random, and I don’t really see its purpose”). – Je pense sincèrement que le langage inversé est une dégradation intempestive du français ou même des autres langages. A défaut d’une langue propre à leur génération, certains jeunes utilisent ce moyen d’expression pour s’isoler d’un monde où ils ne se trouvent pas. Jamais je ne parle le Verlan, car Montesquieu, Voltaire sont des maîtres pour moi. Le français est une langue trop belle pour la dégrader de cette façon. (“I sincerely believe that inverted language is an inopportune deterioration of French and even of other languages. Lacking a language of their own generation, some young people use this means of expression to set themselves apart from a world where they’re out of place. I would never speak Verlan; Montesquieu and Voltaire are masters and my role models. French is too beautiful a language to be degraded in this way.”) (Lefkowitz 1991)

Mainstream disapproval will be shown to recur among all of the cases of antilanguage that we consider here and below.

3.2 Anti-languages in French social networking: Twitter and Facebook Our second example of anti-language, though less robust than Verlan, is the use of Twitter and Facebook in French social media. As in the case of Verlan, French social networking includes nonconventional linguistic acts that are used for playful purposes, as well as for shaping and asserting identity, construct­ ing sociality, cultivating covert prestige, developing peer-group affiliation, and excluding outsiders. Known for its diverse abbreviated and compressed forms, or textisms, social media communication is governed by the constraints of the medium itself. Textisms (sometimes pejoratively labeled textese, slanguage, and digital virus) include initialisms, acronyms, shorthand, highly expressive punctuation, ­emoticons, and emojis (Crystal 2011). These nonstandard – and often ­transgressive – semiotic innovations will be illustrated with examples of formal 2 Participants were speakers of French who attended a highly prestigious Parisian lycée.

356 

 Natalie Lefkowitz and John S. Hedgcock

simplification, lexical borrowing, taboo language, semantic expansion, and more distinctively, registerial blurring and crossover. Sherzer (2002) claimed that speech play entails “the manipulation of elements and components of language in relation to one another, in relation to the social and cultural contexts of language use, and against the backdrop of other verbal possibilities in which it is not foregrounded” (1). It should therefore come as no surprise that such linguistic departures from prescriptive norms result in mainstream disapproval. The first example of deviant linguistic behavior is that of formal simplification through ne deletion in the following two examples, the first from Twitter and the second from Facebook: –– Je suis tjrs la quand il ne le faut pas. Je suis Jamais la quand j’aurais du y être. (“I am always there when I shouldn’t be. I am never there when I should have been there”) –– Je veux pas te perdre bebe ! ♥ Ne me laisse pas seul un jour car je serais perdu sans toi  ! (“I don’t want to lose you baby! ♥ Don’t leave me alone one day because I would be lost without you!”). (O’Connor 2014: 21) In written genres, convention holds that the pre-verbal ne must be retained, even though it is expressed redundantly by the post-verbal particle pas. However, van Compernolle and Williams (2009) found that ne is only retained in about 20% of cases of informal spoken French in France and that Canadian and Swiss francophones regularly omit ne. O’Connor (2014) similarly revealed that, in a Twitter corpus of 2,000 tweets, negation occurred 652 times, with 62 tokens exhibiting complete negation (circumfixation with ne + V + pas) and the remaining 590 exhibiting incomplete negation (post-verbal pas alone). In other words, only 9.5% (62/652) of the negated strings retained ne. In a Facebook corpus of 500 status updates, negation occurred 461 times, with 172 tokens demonstrating complete negation and the remaining 289 demonstrating incomplete negation: Only 37.3% (172/461) of the tokens retained ne. Differences in ne retention rates between the two platforms could be attributed to a couple of factors. First, brevity could play an important role. Tweets are restricted to 140 characters, unlike the unconstrained length of status updates and their accompanying limitless comments on Facebook. Because ne is semantically redundant (pas fulfills the negation role in the absence of ne), its deletion does not alter meaning. The constraints of the medium require Tweets to be brief, thereby supporting ne deletion. Second, user age may influence negation behaviors. In France, Twitter appeals to a younger audience, whereas the age of the Facebook community is widely distributed. This age-differentiated distribution parallels Ashby’s (1981) observations of negation patterns in spoken French: “The rate of ne deletion is considerably higher among speakers of the younger age-stratum than among those of the older” (682).



Anti-language: Innovation, identity construction, affiliation 

 357

The second type of linguistic innovation that defies standard usage entails registerial blending or blurring, frequently observed in oral transcriptions such as those shown in Fig. 1: Standard orthography

Speech form (Phonetic transcription)

“I am”

New digital orthography

[Ʒə sɥi]

[ ʃʃɥi] i]

(O’Connor and Lefkowitz 2014) Fig. 1: Registerial blurring:

In vernacular speech, [ǝ] can be dropped in rapid speech in utterances such as [Ʒə sɥi] (, “I am”) if the deletion does not result in a triple-consonant cluster. Deletion of [ǝ] from [Ʒə sɥi] precipitates regressive assimilation of the voiceless alveolar fricative [s] on the voiced post-alveolar fricative [Ʒ], transforming [Ʒ] into the voiceless post-alveolar fricative [∫]. Thus, [Ʒə sɥi] becomes [∫ɥi]. In his investigation of social media users’ adherence to orthographic rules and phonetic transcription practices, O’Connor (2014) reported that his Facebook corpus contained 697 tokens of je: Of these, 106 (15.2%) involved elision of je with the following consonant (). The proportion of elided tokens in the Twitter corpus was much higher: Of 924 tokens, 222 (24.0%) showed the same phonetic elision pattern. Also noteworthy in O’Connor’s findings is that the written rendering of occurred in 25 tokens (1.5% of the combined Twitter and Facebook corpora). Despite this modest showing, that occurred at all was surprising and suggests that certain vernacular speech patterns have permeated social media language. These trends demonstrate crossover from written to oral to digital registers, revealing how informal language – generally reserved for oral contexts – ­undergoes transformation in online communication. Although these examples represent an emerging and novel social media form that is neither spoken nor written, they also suggest an anti-language-like departure from the strictly prescribed norms associated with the well-attested mainstream French preoccupation with linguistic purity and le bon usage. The constantly shifting hybrid registers of digital communication incorporate and merge elements from both written and spoken modalities (Crystal 2011; Sergeant and Tagg 2014; Sindoni 2013; Thurlow 2013). The following sample from O’Connor’s (2014) Facebook data illustrates a second oral phenomenon: (“I love you, my love”)  (25)

358 

 Natalie Lefkowitz and John S. Hedgcock

Although one could infer that this utterance reflects an orthographic error, we would argue that it actually exemplifies the influence of a widespread pattern in spoken French, liaison, the process by which a mute final consonant is pronounced and joined to the adjacent vowel-initial syllable. In this example, (“my”) is pronounced [mɔ̃], but because it is followed by a word that begins with a vowel, (“love”), the final [n] in mon is fully realized and linked to the next syllable, resulting in [mɔ̃.na.’muʁ]. Liaison is strictly a phonological phenomenon, but the transcription of [mɔ̃.na.muʁ] as represents a merging of spoken and written features by combining French orthography and phonetic (rather than phonemic) spelling (O’Connor 2014). Further evidence of registerial blurring can be found in the examples of negation presented previously, where written cyberlanguage mirrors oral language. In vernacular speech, adjacent strings in an utterance can show ne deletion or ne retention, as in Je suis tjrs la quand il ne le faut pas. Je suis Jamais la quand j’aurais du y être (“I am always there when I shouldn’t be. I am never there when I should have been there”) (O’Connor 2014: 21). Had these online exchanges been governed solely by conventions for writing, we would expect all cases of negation to be uniform and complete (i.e., to display full circumfixation: ne + V + pas). The next deviation from prescriptive norms involves the use of lexical borrowings, taboo language, simplifications, and semantic expansion, samples of which appear in Tab. 5. Tab. 5: Lexical borrowing, taboo language, simplification, and semantic expansion Sample

Source

Kiss on your ass! Bitch please! Bitch #Bitch Fuck WTF Nigger Shit LOL

Twitter Twitter Twitter Twitter Facebook Facebook Facebook Facebook Facebook and Twitter

(Based on O’Connor 2014, 34–35; O’Connor and Lefkowitz 2014)

In our examination of Verlan, we observed that vulgarity and foreign lexical borrowings – particularly from English – may confer coolness, covert prestige, insider status, outsider exclusion (othering), and youthful camaraderie (Lazar 2012). Moreover, using a language other than one’s own establishes a linguistic disguise, a camouflaged sense of protection, and even privacy, hidden behind



Anti-language: Innovation, identity construction, affiliation 

 359

a faceless and foreign linguistic shield. As English has become the most widely used Internet language (Crystal 2011), it leaves an indelible mark on the social media language of French-speaking users: English-based technological terminology is pervasive, and English lexical items “are quickly introduced into Internet jargon and then adapted to suit the needs of the rapid and efficient communication required in cyberspace” (Untied 2009: 9). Translingual practices such as the adoption of English lexis, initialisms, abbreviations, and acronyms express opposition to mainstream French, often provoking the consternation and disapproval of educated French speakers, including self-proclaimed proponents of strict adherence to (largely fictitious) standards of linguistic purity. For example, groups such as the Comité de lutte contre le langage sms et les fautes volontaires sur Internet (“Committee for the Fight against SMS Language and Deliberate Mistakes on the Internet”), which boasts over 20,000 members, have dedicated themselves to counteracting the influence of cyberlanguage. As the textisms shown in Tab. 5 illustrate, initialisms abound in digital communication (Crystal 2008, 2011). This iconic lexical feature of online discourse constitutes a shorthand consisting of abbreviations produced by combining the initial letters of the words in a phrase or lexical bundle. Initialisms differ from acronyms, in that they are pronounced as individual letters rather than as words. A common example, LOL, has expanded its range of reference and social meanings to include meanings in French that are absent from English. In addition to implying that the user is laughing out loud, LOL serves to underscore an utter­ ance, to show irony or discomfort, as a filler when the user has nothing to say, or even as a form of punctuation (Bérard 2010). The following example demon­ strates one or more of these uses: User 2: Je rigole! (“I’m joking”) User 2: Non en fait je rigole pas LOL (“No, in fact, I’m not joking. LOL”) 

(O’Connor, 2014: 26)

In this extract, User 2 had commented on a third party’s status and mentioned a bodily injury that she had incurred. User 2’s use of LOL (“laugh out loud”) in the second turn is especially interesting. To an English speaker, LOL might appear to contradict the fact that she is not joking; however, in this exchange, LOL emphasizes that she is not joking at all. O’Connor (2014) hypothesized that this departure from the original English reference meaning shows contrast with comparable French-origin initialisms that convey similar meanings: MDR = mort de rire (“dying from laughter”), PTDR = pété de rire (“broken from laughter” - vulgar), and XPDR = exploser de rire (“exploding from laughter” - vulgar). O’Connor (2014) noted that the latter did not appear in his Twitter or Facebook corpora, demon­ strating that the English-based LOL may have displaced its French equivalents.

360 

 Natalie Lefkowitz and John S. Hedgcock

It is also worth noting that, as in our Verlan data, a number of these borrowed and relexicalized items express vulgar meanings, a common feature of anti-languages (Halliday 1976; Montgomery 2008). By disturbing social and linguistic conventions, these language practices typically trigger disapproval. According to Jones and Schieffelin (2012), violations of linguistic norms are associated with “the deterioration of face-toface sociality; cyber-bullying; compulsive behavior; and teens’ use of coded messages to coordinate illicit activities such as sex and drug use” (201). Proponents of what Cameron (1995) described as “verbal hygiene” fear that developing online language conventions threaten stable, standard norms. As Crystal (2011) observed, “the prophets of doom have been out in force, attributing every contemporary linguistic worry to the new technology, and predicting the disappearance of languages and a decline in spoken and written standards” (3). Alarmist media representations have contributed to a “genre of worry” and a spirit of “scorn and moral panic” regarding commonly held attitudes toward language use on the Internet (Crystal 2006; Tannen 2013; Thurlow 2006).

3.3 Underperformance in the FL classroom as anti-language The third setting in which we have observed evidence of anti-language behavior is the foreign language (FL) classroom, where the pedagogical norm is frequently synonymous with a high-prestige FL standard, typically a per­ceived national or superstrate dialect such as British English, Mandarin Chinese, Parisian French, or Castilian Spanish (Del Valle 2014; Gass, Bardovi-Harlig,  Magnan and Walz 2002; Magnan and Walz 2002; Valdman 1976, 1989). Like any linguistic ecosystem, the FL classroom provides fertile ground for the emergence of new codes and interregisters (van Lier 2004). In FL instruction, superordinate codes converge toward the pedagogical norm, embodied in the official FL curriculum and in textbook content – and usually represent­ed by the teacher’s own speech behaviors (Valdés and Geoffrion-Vinci 1998). In ­contrast, ­subordinate codes or interregisters may overtly or covertly conflict with the pedagogical norm, aligning their users with social factions who gain cultural capital by occasionally defying the publicly esteemed standard. The social purposes of these non-compliant codes parallel prototypical antilanguage functions, which include resistance to mainstream practices, the reconstruction of identities, the pursuit of peer approval, social solidarity, the assertion of counter-realities, the provocation of mainstream disapproval, and exclusion.



Anti-language: Innovation, identity construction, affiliation 

 361

Our classroom studies initially examined how social pressure and prestige might affect the oral performance of young-adult learners of French and Spanish in secondary and post-secondary FL classrooms (Hedgcock and Lefkowitz 2000; Lefkowitz and Hedgcock 2002, 2006). We were compelled to investigate this relationship by our observation that skilled and even highly proficient FL students sometimes exhibited surprisingly poor oral production skills: We speculated that their non-target-like speech might reflect their efforts to pretend that their oral skills – their pronunciation accuracy, in particular – were not as well-developed as we knew them to be. We thus investigated how FL students’ pronunciation was influenced by the extent to which they connected prestige with the approval or disapproval of their peers and instructors. In a sequence of iterative descriptive studies, we examined learners’ decisions to accept prestigious (often prescriptive) FL speech norms as a means of gaining overt prestige (i.e., by aspiring to achieve target-like pronunciation) or to reject the pedagogical standard (if only occasionally) and thus gain covert prestige by subverting the FL target through calculated underperformance. In particular, we investigated: (1) students’ perceptions of their own and their peers’ FL speech skills; (2) their appraisals of the value of “native-like” pronunciation; (3) their perceptions of social conditions favorable or unfavorable to developing target-like speech skills; and (4) the accuracy of participants’ perceptions of target-like FL pronunciation. Predictably, data from French and Spanish learner cohorts strongly sug­ gested a professed ambition to master “standard” and “native-like” pronuncia­ tion in their FLs. Notable, however, were striking incongruities between par­ ticipants’ perceptions and observed production patterns, as well as their keen awareness of their own and their peers’ overt and furtive efforts to speak French and Spanish poorly – often with the clear aim of parodying the pedagogical norm and implicitly deriding its users. For example, questionnaire data supplied by 268 FL participants indicated broad agreement that they highly valued target-like speech: 88% reported that they liked and admired the sound of “native” speech, 80% reported that it was very important for them to develop “excellent” FL pronunciation skills, and 84% asserted a desire to satisfy the instructor’s expectations for pronunciation accuracy (Hedgcock and Lefkowitz 2000; Lefkowitz and Hedgcock 2002, 2006). Not unexpectedly, only 30% of participants admitted that they occasionally or frequently tried to speak the FL in a non-target-like way; meanwhile, 67% of respondents reported noticing their peers’ purposeful underperformance, with 56% expressing disapproval of this behavior. These “under-the-radar” social practices undertaken by a sizeable minority of students give rise to a localized dialect that typifies anti-language. Adopting these anti-language forms and behaviors (i.e., deliberately producing

362 

 Natalie Lefkowitz and John S. Hedgcock

non-target-like FL speech) constructs a speech community or “counter-reality” (Halliday 1976) within the classroom ecology. Viewed from an anti-language perspective, “[a new] speech community emerges as an arena of competing affil­iations and antagonistic differences” (Montgomery 2008: 119). Data from audio-recorded FL class sessions, semi-structured interviews with FL learners, and questionnaire responses exemplify some of the formal characteristics of anti-language common in Verlan, francophone social media communication, and HL Spanish and Spanglish (see Section 3.4). In the following interview extracts, for instance, Spanish FL students report observing and sometimes participating in purposeful efforts to produce non-target-like speech: –– You know, a lot of times, I- … even though I’m saying “Hola,” I’m going, “God, why am I sayin’ “Hola” [ˑhola]? There’s no “h.” You know? So … (Chris: ­242–243)3 –– I’m not sure if I have a good sense of … fine pronunciation, but I can tell if [fellow students] sound more like a gringo speaking. It’s like they know, but they do it anyway? Like [jow voj æl ˑpɑɹkei] (Yo voy al parque)? (Sharon:  24–25) These representative extracts portray a multiglossic classroom milieu where L2 learners recognize and aspire to reproduce a pedagogical standard, yet where the pedagogical standard serves as a counterpoint to a covert discourse signified by volitional efforts on the part of certain learners to deviate from – and even undermine – that target. Intriguingly, FL students who displayed underperformance behaviors were often those who had achieved a high level of proficiency in Spanish or French. We speculated that mispronunciation behaviors might relate as much to the social dynamics of the FL classroom as to FL skill levels. On first inspection, one might wonder why high-achieving FL learners would hold back when given the opportunity to practice and display their developing L2 speech skills. After all, the ability to produce highly target-like speech can unquestionably confer cultural capital, or overt prestige, in the FL classroom. Nonetheless, a striking number of participants reported observing purposeful mispronunciation, even when among accomplished learners: –– Even guys who speak good Spanish, you know? They’ll, like, not pronounce “r”s right, like [ˑɹɑɹow] instead of, like, [ɾaɾo]? Seems like they do it on purpose, like for fun. Yeah, there’s ah- just real American … pronunciations. (Terry: 7) –– … it’s [deliberate mispronunciation is] definitely support for sounding nonnative. [Laughs] … Um … [in high school] … we had a lot of tape-work, copy 3  Participant names are pseudonyms.



Anti-language: Innovation, identity construction, affiliation 

 363

after the tape and so people would just kind of, you know, make fun of it in a way, sort of, not really try to pronunciate … how you should. (Sharon) –– … all the cool people and even the people that got good grades said things … the way that they wanted to say it. That sounded like a cool high schooler … (Stacy) These extracts paradoxically suggest that underperforming (or pretending to underperform) confers a different kind of cultural capital – covert prestige – which might be equally valued by subversive minority factions in the classroom ecology (Holmes 2008; Trudgill 1972, 2000). By producing highly anglicized FL speech, underperformers may “court … global stigma … in the interests of building a joint non-institutional persona” and affiliating with a lower-status (but highly popular) faction (Eckert 2000: 227). Like the Moroccan adolescent immigrants in Jaspers’s (2005a, 2005b) study of “styling,” play-acting, and “doing ridiculous” in the “standard” Dutch of a Belgian high school, our FL learners sometimes found that trying to speak Spanish or French well was a distinctly uncool thing to do – a signifier of learning the language of authority, yielding to institutional power, and being “nerdy” or “boring.” Learners who can manage both dialects can achieve both overt and covert prestige by earning legitimacy within both highand low-prestige social groups. They adjust their linguistic behavior (engage in communication accommodation, or audience design) by converging upwardly on a high-prestige dialect or pedagogical norm or downwardly on a lower-status or stigmatized dialect (Giles and Ogay 2007; Valdés and Geoffrion-Vinci 1998). Such individuals can selectively “straddle” and negotiate non-intersecting, adjacent, or competing discourses (Jaspers 2005a, 2005b; Lubrano 2004), moving back and forth across linguistic divides to perform the social-semiotic functions required by a given speech event (Halliday 1985; Rampton 1995, 2013; Silva-Corvalán 1994). In line with anti-language use examined in the contexts of Verlan, francophone social media interaction, and HL Spanish (Spanglish), purposeful mispronunciation in the FL classroom reflects the skillful manipulation of semiotic resources to fulfill social functions, including the demarcation of competing ­discourses (Gee 2014). In addition to pursuing covert prestige and consequent alliances with “cool” peers, FL underperformance can express defiance, if not outright mockery, of the perceived mainstream target as these extracts illustrate: –– [Our] Spanish teacher … she’d get really upset … after a repeated, mistake was made? So it got to the point almost … where … we each had our certain mistake … that we would do wrong and we’d do it on purpose just to piss her off. (Chris) –– … there was a girl who, uh … was not … of the popular set and she, but her French, you know, she pronounced it very well … and kids would … like,

364 

 Natalie Lefkowitz and John S. Hedgcock

they’d make fun of her and … they’d say “Oh, why do we need to read? Let’s just have her read and listen …” [there was] kind of pressure to sound bad. (Josh) –– … in high school I noticed [deliberate mispronunciation] … you know? It’s just ‘cause they were goofing around, like, “Oh, this is French class,” you know? They’re- they weren’t taking it too seriously. (Heather) Underperformers who display their linguistic knowledge and skill to challenge and deride accomplished peers and authority figures such as the teacher wield power in a milieu where “sounding bad” curries favor with minority factions in the classroom (cf. Jaspers 2005a, 2005b; Rampton 1995). Moreover, although most FL learners professed and exhibited oral skills aligned with the pedagogical norm (a high-prestige standard), those who described and admitted to underperformance behaviors hinted strongly at an inclination to resist mainstream expectations and even to make fun of them.

3.4 Heritage Spanish as anti-language Verlan, francophone social media registers, and purposeful underperformance in FL classrooms exhibit numerous formal features that typify previously investigated anti-languages; these features similarly fulfill several social and ideological functions. Our research on the language and literacy socialization of heritage language (HL) speakers of Spanish in the U.S. (Hedgcock and Lefkowitz 2011, 2016; Lefkowitz 2011; Lefkowitz and Hedgcock 2007, 2008, 2011, 2012; Reichelt, ­Lefkowitz, Rinnert and Schultz 2012) suggests that contexts of HL development also promote the emergence of anti-languages, which manifest themselves in subtler ways. The HL designation refers to “a language other than English used by immigrants and possibly their children that symbolically and linguistically represents their country of origin” (Lee and Suárez 2009: 138). Though definitions vary, a heritage language learner (HLL) in the U.S. is a bicultural individual who first acquired the HL at home and in the community — and who later becomes proficient in English in childhood (mainly at school). With limited interaction in the HL outside the home, adolescent and adult HLLs often display high-level oral-aural skills but limited HL literacy. A salient socio-affective dimension of HL development is that many HLLs undertake formal study of their HLs in early adulthood “to connect with communities of speakers” and reclaim or affirm their cultural identities (Carreira and Kagan 2011: 40). The divide between vernacular oral-aural competence and literacy skills in the HL (notably, academic literacy) tends to become pronounced in adolescence and early adulthood (Achugar 2003; Bowles, Toth and Adams 2014; ­Chevalier 2004;



Anti-language: Innovation, identity construction, affiliation 

 365

Cohen and Gómez 2008; Hornberger and Wang 2008; Montrul 2008). Frequently viewed as lower-prestige, the stylistic range of HLLs is often narrower than that of more educated NSs (Carreira and Kagan 2011; Chevalier 2004; Hedgcock and Lefkowitz 2016; Hoopingarner 2004; Valdés 2001; Valdés, Fishman, Chávez and Pérez 2006; Wiley, Peyton, Christian, Moore and Liu 2014). It is undeniable that users of lower-prestige varieties may enjoy limited access to cultural capital and discourses where high-prestige varieties predominate (e.g., educational institutions, professions, workplaces, and so on). Perhaps because of its lower social status, a low-prestige variety can become “a language of intimacy” among its users, especially in heteroglossic settings where high- and low-prestige varie­ ties co-exist and compete (Wardhaugh and Fuller 2014). HL varieties of Spanish can similarly be viewed as interregisters, developing varieties that approximate a superordinate target register (e.g., a pedagogical “standard”) entailing “precise ways of speaking for particular domains” (Valdés and Geoffrion-Vinci 1998: 494). HLLs whose Spanish dialect qualifies as an interregister “must work consciously to acquire ways of speaking that characterize the groups to which they aspire to belong” – namely, speakers of high-prestige dialects (Valdés and Geoffrion-Vinci 1998: 476). Frequently characterized as English-dominant, HLLs often engage in translingual behaviors that may suggest gaps in their knowledge of Spanish. Importantly, these behaviors represent systematic, creative uses of linguistic resources from both Spanish and English. Regrettably, the port-manteau term Spanglish often refers pejoratively to novel vernaculars that have emerged among populations of Spanish-­English bilinguals and HL speakers in the U.S. (Lipski 2006; Valdés et al. 2006; Valdés, González, García and Márquez, 2008; Zentella 1997, 2008). For good and for ill, Spanglish represents a hybrid variety that signifies the hybrid identities of HLLs, who often characterize their subject positions as tentative, vulnerable, and marginal. Although they participate in mainstream and peripheral discourse communities, HLLs tend to be assigned hyphenated identities (Wright 2004), which can develop when languages shift from older to younger generations among immi­ grant populations – and almost inevitably toward the code of the superstrate culture (Fishman 1977; Giampapa 2004; Wong-Fillmore 1994). No more homogeneous than “standard” varieties of Spanish, Spanglish exhibits several common features that index its users’ translingual, transcultural socialization, including cross-linguistic morphophonological influences (primarily anglicization), lexical borrowing, the regularization of false cognates, and higher-frequency occurrence of nonstandard verbal inflections. The data samples in Tab. 6 exemplify attested forms that are traceable to the translingual practices of HLLs in our study and other Spanish-English ­bilinguals – and notable for how they depart from high-prestige varieties of Spanish popularly viewed as “standard” (Del Valle 2014).

366 

 Natalie Lefkowitz and John S. Hedgcock

Tab. 6: Spanish  Spanglish Innovations “Standard” Spanish form

“Spanglish” form English Gloss

Category(-ies)

almuerzo estacionar competencia seguro enlazar lo mejor asistir

lonche parquear competición aseguranza linkear el top atender

lunch to park competition insurance (policy) to (hyper)link the best to attend

nadie hablaste

nadien hablastes

nobody, no one you (fam.) spoke

Lexical borrowing Lexical borrowing Anglicization Anglicization Anglicization Anglicization Anglicization; False cognate Colloquialization Colloquialization

These samples and innumerable others like them illustrate several processes thought to signify anti-language development, including phonological and morphological innovation, and lexical borrowing. Somewhat paradoxically, use of such forms is often perceived as symptomatic of constrained linguistic creativity that predictably generates disapproval from mainstream speakers and writers of “standard” varieties of Spanish, including language educators, Spanish-English bilinguals who are not HLLs, and monolingual Spanish speakers. Guadalupe, a participant who self-identified as an educated native speaker of Latin American Spanish, expressed a common attitude toward Spanglish and comparable hybrid varieties, claiming that Spanglish is “inferior.” She reported the following in an interview: “I have heard those [Spanglish forms] and that made me feel very strange when I hear these words. I can’t pronounce when that doesn’t exist. I don’t feel fine that everybody is using it as normal vocabulary.” She confirmed that “there is a bad Spanish,” a category that includes Spanglish. In comparing her regional dialect to the dialects of her Mexicanborn classmates, Guadalupe asserted that her variety was “totally different,” observing that HLLs “do a combination on English and Spanish.” Although she noted that code-switching and code-mixing were commonplace and even contagious, Guadalupe’s portrayal of her contemporaries’ varieties of Spanish suggests a negative evaluation of their competence and thus of their legitimacy as users of the language. We recognize that the power of “standard language” has diminished and that notions of “linguistic purity” largely amount to conven­ ient (and misleading) myths (Coupland 2010; Lippi-Green 2012; Niedzielski and Preston 2003; Preston 1999; Rampton 2015). Nonetheless, misguided beliefs about – and attitudes toward – standard language and vernaculars remain pervasive among members of the general public and even among FL educators (as our own research has shown). Though ill-founded, the e­ ntrenched standard



Anti-language: Innovation, identity construction, affiliation 

 367

l­ anguage myth continues to influence educational practice and learners’ linguis­ tic self-esteem. Like the non-heritage Spanish speakers in our study, our HL participants frequently evinced similarly unfavorable images of themselves as users of Spanish in accounts of their interactions with instructors and with non-HL Spanish speak­ ers, often expressing feelings of inferiority. The following interview extracts exemplify the strong tendency among HLLs to devalue their language proficiency and practices: –– When speaking Spanish, “I pause, I hesitate, I like to sound more fluent than the way I do . . . I am never satisfied where I am at and I can usually identify the areas that I need to improve in” (Aurelia); –– “I sometimes use wrong words because of the combination of English and Spanish . . . I translate the words [like] Vamos a tomar un break . . . It’s like not standard” (Oswaldo); –– Parents and grandparents “get on our case if we speak English around them. In certain situations, speaking English makes it seem like you are turning away from your own . . . culture to others and if you don’t speak as much Spanish, that means you are not as Mexican as you should be” (Verónica); –– Spanglish is “not correct, so you should not use it … We can say more stuff. It gives a better picture [but it’s] … not professional” (Manuel); –– In México, “people can tell that I’m not from there. They can tell my accent. That’s an English accent. I like to speak Spanish… but I feel like I am an Amer­ ican.” Spanglish amounts to “Spanish in the wrong way” (César). Unlike Verlan and the subversive codes created among deliberately underperform­ ing FL learners (discussed in Section 3.3), varieties of HL Spanish (including Spanglish) do not result from HLLs’ volitional effort to defy perceived norms or linguistic standards. Indeed, these varieties confer little, if any, cultural capital on their users vis-à-vis adherents of “standard” Spanish. Non-users of HL ­varieties (speakers of standard varieties, bilingual speakers who are not HLLs, language teachers) often perceive HL varieties in ways that parallel how the linguistic ­mainstream views anti-languages – as substandard vernaculars (Hedgcock and Lefkowitz 2016; Lefkowitz and Hedgcock 2007, 2008, 2011, 2012). Our survey and interview data reveal strong biases against both HL varieties and their users, inevitably contributing to HLLs’ feelings of incompetence, as well as alienation and marginalization. For example, Antonia indicated that she had been told by her in­­ structors that she did not speak Spanish well, observing that they expect HLLs “to know everything.” María’s recollections of classroom interactions likewise highlighted uncomplimentary appraisals of her speech: The professor, she said, “was always making negative comments and that made me feel bad … My vocabulary

368 

 Natalie Lefkowitz and John S. Hedgcock

was not the one the professor was expecting. The professor was constantly correcting my vocabulary and said you should say this instead of that.” Hardly unusual, similar reports of earning the disapprobation, if not scorn, of authority figures who are socially positioned to judge the speech and writing of HLLs reveal tension between the pedagogical norm and HLLs’ lower-status home and community varieties of Spanish. As with Verlan, the anglicized language of FL students, and the evolving language of social media communication, the hybridized language practices of HLLs are often perceived as anti-languages, alienating them from matrix discourse communities. A target of denigration, Spanglish has nonetheless become a signifier of solidarity among some of its users. Although our study did not explicitly investigate the status of Spanglish, researchers such as Fairclough (2003) and Zentella (2008) have argued that its use has become an emblem of bicultural identity and solidarity in Latino communities. As Otheguy and Stern (2010) noted, “many Hispanics in the USA use the term Spanglish with pride, considering it a badge of identity and self-esteem.” Indeed, “artists, professors, journalists, and other opinion makers … proudly proclaim that they speak Spanglish, according this term a level of covert prestige” (96). Though not overtly manifest in our own data, these trends suggest that the translingual creativity that makes Spanglish a vital and dynamic code has also established its linguistic counter-legitimacy.

4 Discussion and conclusion By revisiting and reframing data from our previous studies through the antilanguage lens, we have indeed affirmed Montgomery’s (2008) assertion that anti-language does not represent an absolute category that requires uncompromising adherence to all of its observed features. Rather, anti-languages evolve along a loosely constructed continuum of characteristics that allows for flexibil­ ity according to the unique circumstances in which anti-languages develop. In other words, any given anti-language is likely to manifest some (but likely not all) of the tendencies associated with the phenomenon. Our research illuminates salient features common to these emerging speech communities and their antilanguage practices related to (1) Verlan, (2) novel registers evolving among francophone social media users; (3) deliberate underperformance behaviors among classroom FL learners; and (4) Spanish heritage-language learners’ appropriation of high- and low-prestige varieties. In descending order, Verlan exhibits the greatest number of anti-language properties, followed by French social media; underperformance in the FL classroom and Heritage Spanish tie for final position, as illustrated in Fig. 2.

Linguistic Characteristics

P P P P P P P P 15

TOTAL

P P P

P P P*

9

P P

P P 14

P P P P

P

P

P

Underperformance in FL Instruction

P P P P P

P

P

P

P

P P

French Social Media

P P

Identity Maintenance & Affirmation Opposition, Resistance, & Defiance Covert Prestige Counter- (Alternative) Reality Mainstream Disapproval Verbal Competition & Display Secrecy & Concealment Solidarity, Social Relations, & Peer Approval Exclusion of Non-Initiates (Othering)

Relexicalization Simplification Phonological & Morphological Innovation Lexical Borrowing Taboo Language Registerial Blurring

Verlan

9

P

P P P P

P

P*

P

P

Heritage Spanish (Spanglish)

Highest Frequency    Lowest Frequency

Anti-language: Innovation, identity construction, affiliation 

Fig. 2: Comparison of Anti-Language Characteristics and Functions

* Owing to space constraints, our analysis does not include explicit evidence demonstrating these phenomena. Our data nonetheless include observations of Verlan and Heritage Spanish (Spanglish) users deploying informal conventions in formal settings, transgressing a ­ ppropriateness boundaries for register choice.

Sociolinguistic Functions



 369

370 

 Natalie Lefkowitz and John S. Hedgcock

Because of its association with marginal, criminal countercultures and its consequent value as a signifier of in-group membership, identity marking, and linguistic alternative realities that oppose mainstream linguistic practices, Verlan exem­ plifies the most overtly defined anti-language status; indeed, Verlan serves as a prototype. Accordingly, it occupies the highest position on the continuum. For banlieusards, the use of Verlan contributes to a “collective identity” that “transcends their diverse countries of origin, religion, and sociolinguistic factors as a response to the economic and social conditions they all share” (Hassa 2010: 61). Furthermore, the social stature of the perceived high-prestige standard and its inextricable connection to French civilization and culture lends itself to equally strong opposition by those who feel marginalized. As Bullock (1996) observed, the ability to speak Verlan well requires a certain linguistic virtuosity: “[O]ne object of this linguistic behavior must surely be to sound conspicuously differ­ ent from the standard language” (189). Doran (2007) similarly underscored the striking lexical, semantic, and phonological innovations in Verlan that depart dramatically from standard (prescribed) French, and whose use can be seen as “a refusal of the measured, careful pronunciation of normative French, the language of a cultural elite by whom minority youths feel negatively judged” (501). Figure 2 demonstrates that, of the 15 linguistic and social categories associated with anti-languages, Verlan displays 15 out of 15, French social media 14 out of 15, and both FL underperformance, and HL Spanish 9 out of 15. Also worth noting is the variable frequency of anti-language features across codes, as well as the features that all four groups share: identity maintenance and (re)affirmation; opposition, resistance, and defiance of mainstream norms; covert prestige; mainstream disapproval; solidarity, social relationships, peer approval, and insider status; exclusion of outsiders and non-initiates; phonological and morphological innovation; and lexical borrowing. According to Montgomery (2008), “the notion of an anti-language … can be used to illuminate certain kinds of social dialect” and “to clarify the notion of speech community” (119). The four emerging speech communities examined here enhance our understanding of the ways in which young people in language contact, social media, and L2 development settings engage in creative linguistic innovation in order to (re)construct novel identities, express individuality, ­cultivate solidarity and covert prestige among peers, and resist dominant linguistic and cultural standards. “The perspective of the anti-language,” argued Halliday (1976), “is one in which we can clearly see the meaning of variability in language: in brief, the function of alternative language is to create alternative reality. A  social dialect is the embodiment of a mildly but distinctly different world view—one which is therefore potentially threatening, if it does not coincide with one’s own” (581).



Anti-language: Innovation, identity construction, affiliation 

 371

References Achugar, Mariana. 2003. Academic registers in Spanish in the U.S.: A study of oral texts produced by bilingual speakers in a university graduate program. In Ana Roca & M. Cecilia Colombi (eds.), Mi lengua: Spanish as a heritage language in the United States, 213–230. Washington: Georgetown University Press. Alim, H. Samy, Awad Ibrahim & Alastair Pennycook. (eds.). 2009. Global linguistic flows: Hip hop cultures, youth identities, and the politics of language. New York: Routledge. Ashby, William J. 1981. The loss of the negative particle ne in French: A syntactic change in progress. Language 57(3). 674–687. Baker, Paul. 2002. Polari: The lost language of gay men. New York: Routledge. Bérard, Coline. 2010. Lol,