UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Peer language socialization in an internationalized study abroad context : norms for talking about language Surtees, Victoria 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2019_february_surtees_victoria.pdf [ 3.08MB ]
JSON: 24-1.0375711.json
JSON-LD: 24-1.0375711-ld.json
RDF/XML (Pretty): 24-1.0375711-rdf.xml
RDF/JSON: 24-1.0375711-rdf.json
Turtle: 24-1.0375711-turtle.txt
N-Triples: 24-1.0375711-rdf-ntriples.txt
Original Record: 24-1.0375711-source.json
Full Text

Full Text

 PEER LANGUAGE SOCIALIZATION IN AN INTERNATIONALIZED STUDY ABROAD CONTEXT: NORMS FOR TALKING ABOUT LANGUAGE by Victoria Surtees B.A. University of British Columbia, 2006  D.E.S.S., University of Montreal, 2007 M.A., Concordia University, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Teaching English as a Second Language)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2018  © Victoria Surtees, 2018   ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  Peer language socialization in an internationalized study abroad context: Norms for talking  about language  submitted by Victoria Surtees in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Teaching English as a Second Language  Examining Committee: Dr. Patricia Duff Supervisor  Dr. Steven Talmy Supervisory Committee Member  Dr. Sandra Zappa-Hollman Supervisory Committee Member Dr. Monique Bournot-Trites University Examiner Dr. Claudia Ruitenberg University Examiner    iii  ABSTRACT Study abroad (SA) students often expect to have opportunities to interact with peers in the target language during their sojourns. However, while SA students report valuing peer relationships, to date, few studies have explored the role of peer interaction in SA (McGregor, 2016) and even less research has attended to peers’ perceptions of their role in SA students’ language learning (Kinginger, 2017). To address this gap, this qualitative multiple case study investigates peer language socialization at an internationalized English-medium university in Canada. It focuses on how language (e.g., grammar, use, lexis, and pronunciation) and language learning were oriented to in peers’ conversations, and the norms around how such topics were to be managed in informal talk. The focal students were three Japanese undergraduate SA students, each of whom recruited several English-speaking peers with whom they recorded weekly conversations. The data, which included interviews with peers and SA students as well as the recorded conversations, were analyzed using micro-analytic approaches, including membership categorization analysis (Housley & Fitzgerald, 2015; Sacks, 1992) and discursive approaches to stancetaking (Du Bois, 2007; Stevanovic & Peräkylä, 2012). Findings show that SA students had difficulty forming peer relationships, despite their engagement in extracurricular activities. The interview data also reveal that SA students valued peers who were multilingual and had experience with international students, and that peers valued SA students who asked for language help and displayed willingness to improve their English. While peers’ reports in interviews depicted discussions of language as relatively simple interactions, analyses of the peer interaction data demonstrated that SA students and peers required significant linguistic resources iv  and prior knowledge to successfully engage in talk-about-language (Levine, 2009) and that not all SA students’ requests for language help were successful. These findings point to how “doing novice” and “doing expert” may be learned practices and highlight the need to conceive of peers as historical multifaceted individuals who may or may not be willing or able to appropriately “do expert” in interaction with SA students. As such, this study makes a significant contribution to applied linguistics in the areas of SA, peer interaction, and language socialization.       v  LAY SUMMARY This multiple case study explores how three Japanese study abroad students, with intermediate English skills, discussed language and language learning in conversations with English-speaking peers at a Canadian university. Data include interviews with the study abroad students and their peers as well as 11 hours of recorded interactions. Findings show that peers were reluctant to correct students and expected study abroad students to ask for language help. Study abroad students highlighted that peers with international experience were more supportive conversation partners than native speakers with less international experience. They also reported that they had to learn to ask for language help. Analyses of recorded conversations show that asking for language help was a complex task and that the advice peers provided was not always accurate. The results point to the need to reconsider displays of “language expertise” as a learned behaviour rather than an inherent property of native speakers.  vi  PREFACE This study has undergone an ethical review process which was approved on December 9th, 2015 by the University of British Columbia Behavioural Research Ethics Board. The Human Ethics Certificate #H15-02937 for “Peer language socialization in study abroad” is valid until November 3rd, 2018. This dissertation is original, independent work by the author, V. Surtees.  A version of Chapter 7 has been accepted for publication. Surtees, V. (2018). “As a friend, that’s the one thing I always am very conscious not to do”: Categorization practices in interviews with peers in the host community. Study Abroad Research in Second Language Acquisition and International Education.     vii  TABLE OF CONTENTS ABSTRACT ................................................................................................................................... iii LAY SUMMARY ............................................................................................................................v PREFACE ...................................................................................................................................... vi TABLE OF CONTENTS .............................................................................................................. vii LIST OF TABLES ........................................................................................................................... i LIST OF FIGURES ........................................................................................................................ ii LIST OF EXCERPTS .................................................................................................................... iii LIST OF ACRONYMS ................................................................................................................. iv ACKNOWLEDGEMENTS .............................................................................................................v DEDICATION ............................................................................................................................... vi Chapter 1: Peer interaction in study abroad .....................................................................................1                 Chapter 2: Study abroad at an internationalized university ...........................................................18                       Chapter 3: Theory and analytic focus ............................................................................................40         viii               Epistemic stance ..................................................................................... 65 Deontic stance ........................................................................................ 66     Chapter 4: Methods ........................................................................................................................68                                    Excerpt selection .................................................................................... 93 Focus on categorization .......................................................................... 94 Excerpt selection .................................................................................... 96 Discourse analytic approaches to stance .............................................. 100   ix    Chapter 5: Stories of peer socialization .......................................................................................103                                         Chapter 6: Expert and novice peer identity categories ................................................................135                               Chapter 7: Rights and responsibilities in talk-about-language ....................................................169         x                  Chapter 8: Accomplishing talk-about-language in peer conversation .........................................195  Word searches ...................................................................................... 201 Other-initiated repair ............................................................................ 205 Doing learning ...................................................................................... 210 Failed bids for completion.................................................................... 214 Preemptive repair ................................................................................. 218 Interrupting to ask ................................................................................ 223 Distributed expertise ............................................................................ 228 Collaborative telling ............................................................................. 232         Chapter 9: Conclusion and implications ......................................................................................240                             xi    REFERENCES ............................................................................................................................261                    LIST OF TABLES Table 3.1 SA Studies with Recorded Interaction .......................................................................... 47 Table 4.1 Summary of JC Participants ......................................................................................... 77 Table 4.2 Summary of JC Peer Participants ................................................................................. 79 Table 4.3 Summary of Interview Data .......................................................................................... 83 Table 4.4 Transcription Conventions ............................................................................................ 92 Table 4.5 Examples of Talk-About-Language Sub-Categories .................................................... 98 Table 4.6 Frequencies for Talk-About-Language in Ami's Remote Observation Data ................ 99 Table 4.7 Frequencies for Talk-About-Language in Samantha's Remote Observation Data ....... 99 Table 4.8 Frequencies for Talk-About-Language in Lisa's Remote Observation Data ................ 99 Table 6.1 Novice Peer Categories and Predicates ...................................................................... 153 Table 6.2 Expert Peer Categories and Predicates ....................................................................... 163  ii  LIST OF FIGURES Figure 1.1 Talk-about-language ...................................................................................................... 8 Figure 3.1 Multifaceted nature of language learning and teaching .............................................. 56 Figure 4.1 Remote observation timeline ....................................................................................... 86 Figure 4.2 Focal case map ............................................................................................................ 89 Figure 4.3 “She’s not ashamed” memo......................................................................................... 95 Figure 4.4 Analytic memo for repair sequence “midterm” ......................................................... 101 Figure 5.1 Language ideologies in JC peer stories. .................................................................... 133 Figure 6.1 Relationship between language ideologies and categories. ....................................... 167 Figure 7.1 Asking and helping as category-bound practices. ..................................................... 191 Figure 8.1 Relationship between ideologies, categories, and practices ...................................... 237 iii  LIST OF EXCERPTS Excerpt 4.1 “She’s not ashamed”.................................................................................................. 94 Excerpt 4.2 “Midterm”................................................................................................................ 101 Excerpt 6.1 “I didn’t realize she was an exchange student” ....................................................... 141 Excerpt 6.2 “Mostly my friends are from America” Part 1 ........................................................ 143 Excerpt 6.3 “Mostly of my friends are from America” Part 2 .................................................... 145 Excerpt 6.4 “Lisa kind of takes initiative” .................................................................................. 148 Excerpt 6.5 “She’s not ashamed”................................................................................................ 150 Excerpt 6.6 “They accept me to speak” ...................................................................................... 155 Excerpt 6.7 “Maybe cause I’m used to it” .................................................................................. 158 Excerpt 6.8 “A lot of my friends are international” .................................................................... 161 Excerpt 7.1 “It’s her language I guess” ...................................................................................... 173 Excerpt 7.2 “No whenever she asks” .......................................................................................... 177 Excerpt 7.3 “I’m always here if she wants to practice” .............................................................. 179 Excerpt 7.4 “I'm not like an English expert” .............................................................................. 183 Excerpt 8.1 “How do you say” ................................................................................................... 202 Excerpt 8.2 “There is what? cores?” ........................................................................................... 205 Excerpt 8.3 “Is it correct?”.......................................................................................................... 211 Excerpt 8.4 “You’ve never seen this?” ....................................................................................... 215 Excerpt 8.5 “Sloths”.................................................................................................................... 219 Excerpt 8.6 “Do you know a wig?” ............................................................................................ 220 Excerpt 8.7 “Whipped cream?” .................................................................................................. 224 Excerpt 8.8 “Don't you say one piece?” ..................................................................................... 228 Excerpt 8.9 “Taking a movie is it correct?” ................................................................................ 233  iv  LIST OF ACRONYMS  JC  Japan-Canada L1  First language L2  Second/additional language LS  Language socialization PWU  Pacific Western University  SA   Study abroad TL  Target language      v  ACKNOWLEDGEMENTS A dissertation is not dissimilar to a marathon: it takes time, discipline, financial resources, and most especially the support of others. I was fortunate to have wonderful committee members who have supported my research agenda and pushed me to grow. To Dr. Talmy, who introduced me to the wild world of discourse analysis, I thank you for opening my eyes and ears to interaction in ways that have radically changed my views on language. To Dr. Zappa-Hollman, who has shown me what it means to be a strong leader in international programs, I thank you for sharing your experiences and your knowledge. I am especially indebted to my supervisor, Dr. Patricia Duff. Her steady belief in the value of this project and her kind and wonderfully thorough feedback has helped keep me on the path to completion. I would be remiss if I did not also thank my participants, who shared their stories, their time, and their conversations with me. It took courage and openness to participate in this project and I wish them all the best in their future endeavours. I must also express my gratitude for the generous financial support I received from my home department of Language and Literacy Education and from the Social Sciences and Humanities Research Council. Without their funding, this study would not have been possible.  In my journey, I have been fortunate to receive words of wisdom from many teachers and colleagues. In LLED and the faculty of Education more broadly, I have been surrounded by colleagues who have listened to my ideas, read drafts, and provided access to key resources for this study. To Dr. Reg D’Silva – this research could never have been completed without your generous support and openness. To my peers – we shared stories of difficulty, stories of triumph, tips and tricks for writing, drafts, and so much more. Although dissertation writing is in many ways a solitary process, the mentorship and comradery of colleagues like Natalia, Laura, Ryan, and Liam made me feel less alone and helped me explore my ideas in a safe space. At conferences, I was fortunate enough to meet Dr. Janice McGregor, Dr. Emma Trentman, and Dr. Wenhao Diao, all of whom have ushered me into the study abroad community. I also wish to extend my thanks to Dr. Celeste Kinginger, whose research inspired much of what is contained in this dissertation, and who kindly agreed to serve as external examiner. Before these incredible colleagues, I also had the benefit of strong support and encouragement at Concordia University, where Dr. Elizabeth Gatbonton (may she rest in peace) first suggested that I pursue a PhD. Truly though, I would not be here without the incredible love of my family and husband. My parents, Christine and Alan Surtees, and brother, Douglas Surtees, have stood by me when I thought my project would never end. They read drafts, kept tabs in the dark times, and celebrated my victories. Finally, to Russell Gardiner, my incredible husband, your patience and unwavering belief that I would succeed has been my driving force. You have seen me at my worst and at my best and you have shared in it all by my side. Throughout this entire process, you have shown me more kindness than I ever could have hoped for. Oh, and your patience. Did I mention your patience? There will never be enough words to thank you.  vi  DEDICATION For Dr. Elizabeth Gatbonton.  1  Chapter 1: Peer interaction in study abroad 1.1 Introduction Study abroad (SA) refers to programs where students participate in a temporary formal education experience (often in a different language) in another country or region for a fixed duration (Kinginger, 2009). It is no secret that post-secondary institutions worldwide are increasingly making SA a key component of their strategies to produce multilingual and global citizens (Teichler, 2015). According to the OECD (2017), around 3.3 million students studied abroad in 2015, with that number set to double by 2020. Given the huge number of students enrolled in both short-term and degree-length international programs, it is safe to say that SA is a global economic and social phenomenon of increasing significance in the field of higher education.  In general, students who choose to enrol in SA programs do so to discover new cultures, to increase their employment prospects, and – of particular interest to this study – to improve their foreign language skills. However, while large-scale reports have identified significant benefits of SA in these three areas (University of Oxford International Strategy Office, 2015), research on language learning outcomes for sojourns lasting one year or less has tended to observe widely varying results at the individual level (Kinginger, 2015a). For this reason, an increasing number of SA researchers are turning to qualitative methods to better understand how SA students’ expectations, experiences, and interactions with host-community members shape students’ target language (TL) development (Shively, 2018).1 This dissertation aims to contribute                                                  1 I have chosen to use the term target language (TL) development rather than the more widespread term second language (L2) development to allow for the fact that most participants in this study had at least some knowledge of 2  to a better understanding of how SA students’ (English) TL encounters shape their language learning by focusing specifically on the role of peer relationships and interaction. It does so by exploring the interactions and relationships of three Japanese undergraduate SA students and their English-speaking peers at a large western Canadian university, which I will call Pacific Western University (PWU). The three focal participants were enrolled in an eight-month program, the Japan-Canada (JC) program, and had opportunities to meet a wide variety of English-speaking peers through university residences, language exchanges, and extra-curricular activities. In what follows, I explain the ways in which this study contributes to a better understanding of the role of peer interaction in TL learning in the SA context.    1.2 Peer interaction in SA For many SA students, a key objective of the SA experience is to make friends in the TL community (Goldoni, 2013; Mitchell, 2015; Pyper & Slagter, 2015). However, as McGregor (2016) pointed out, despite SA students’ desires to engage with peers, “we still know very little about the ways in which sojourners co-construct and negotiate talk with local peers in conversation” (p. 178). Indeed, while many studies have examined the role of homestay experiences in SA (e.g., Diao, Freed, & Smith, 2011; DuFon, 2006; Greer, Brandt, & Ogawa, 2013; Iino, 2006; Ishida, 2011; Jackson, 2008; Kinginger, Lee, Wu, & Tan, 2014; Knight & Schmidt-Rinehart, 2010; Pekarek Doehler, 2018; Shively, 2016; Van Booven, 2017; Wilkinson, 1998), according to McGregor (2016), fewer studies have focused on the role of interactions with TL-speaking peers in the host community.                                                   more than two languages, including the SA participants. L2 is still used when referring to the additional languages of participants that were not the focus of this study and in citations from other authors.  3  SA studies that mention peer relationships often do so with the aim of quantifying or modelling the peer networks that SA students form (Dewey et al., 2014; Dewey, Ring, Gardner, & Belnap, 2013; Isabelli-Garcia, 2006; Mitchell, 2015; Zappa-Hollman & Duff, 2015). Alternatively, other qualitative and narrative studies mention peer interaction as one aspect of SA students’ experiences more generally. These studies explore the ways in which a wide variety of encounters (e.g., peer relationships, homestays, classes, extracurricular activities) impact SA students’ identities (Benson, Barkhuizen, Bodycott, & Brown, 2012; Jackson, 2008; McGregor, 2012; Stewart, 2010; Whitworth, 2006), or beliefs, attitudes, goals, or motivations (Allen, 2010; L. Miller & Ginsberg, 1995; Shively, 2016; Yang & Kim, 2011). This body of research typically highlights students’ difficulties in accessing opportunities for TL interaction with peers as well as the benefits of doing so. These studies show that those who succeed in forming relationships with peers via the TL often express a deep satisfaction with their sojourns. They also point to the ways in which educational institutions can create opportunities for TL-mediated relationships through residence arrangements, language pledges, language exchange or mentorship programs, internships, and extracurricular activities (Mitchell, 2015; Trentman, 2013a).  However, given these studies’ focus on a broad range of experiences, the specific contributions of interaction with peers often remain underexplored. In addition, because these studies typically rely heavily on self-report data from SA students, such as questionnaire, interview, and journal/diary data, the perspectives of host community peers are largely absent in this body of research. As such, while the perspectives of SA students are discussed in depth, the expectations, beliefs, and motivations that shape peers’ engagement with SA students often go unrecognized. The studies cited above also present little evidence about what students and their 4  peers might talk about or the ways in which they collaboratively engage in language learning activities.  To date, what we do know about the features of TL peer interaction comes from a small number of studies that have examined specific instances of interaction between SA students and their native-speaker peers (Diao, 2013, 2016; Dings, 2012, 2014; Fernández, 2016; Ishida, 2010; Levine, 2009; Masuda, 2011; McGregor, 2016; Shively, 2013). Some of these studies have analyzed peer talk as a way to observe SA students’ TL development across the sojourn. For example, Dings (2014) observed that US students developed a more varied repertoire for expressing alignment in Spanish conversations with native-speaking peers during a year-long sojourn in Spain. Others have shown that SA students sometimes discuss norms around TL use with peers, including pragmatic norms related to personal pronoun use (McGregor, 2016), differences between languages and language varieties (Fernández, 2016), and issues of pronunciation, lexis and morphosyntax (Levine, 2009).  While micro-analytic studies of peer interaction provide rich data on the ways in which peers and SA students engage with each other, peers’ perspectives are still not usually presented in these studies. The language practices observed are thus seldom discussed in the context of local expectations around how language learning can or should be managed between peers. Kinginger (2012, 2017) has repeatedly pointed out that SA research needs to more systematically include the perspectives of host community members, such as peers, host teachers, families, and program administrators, in order to better understand factors that mediate access to interaction and potential sources of misunderstanding. She argues that if language learning is theorized as a social and collaborative process (Duff & Talmy, 2011), then to understand SA students’ 5  experiences, it is at least as important to investigate the norms and values of hosts as it is to understand the perspectives of SA students themselves.  To date, only two studies (Diao, 2013; Kinginger & Wu, 2018) that I know of have adopted an integrated approach that examined both recordings and ethnographic data related to peer interaction. Diao (2013) investigated US students’ development of gendered particles in Mandarin in interactions with Chinese roommates in China. In her work, Diao examined how four US students developed patterns of use that more closely resembled the patterns observed in the speech of their Chinese roommates. She also pointed to instances in which participants explicitly discussed norms around the use of gendered particles. While the main focus of her research was language use, Diao also engaged in a substantial discussion of Chinese peers’ interest in English and international travel and the ways in which those interests mediated the Chinese peers’ engagement with the US students. In addition, her discussion of the results attended to the ways in which discourses about “foreigners” in China shaped the ways in which peers, as well as other locals, treated and interacted with the US students.  Diao’s (2013) study is relatively unique in that it not only investigates the specific affordances of peer interaction for language learning (i.e., the development of gendered particles in Mandarin) but it also addresses the underlying tensions and locally circulating norms that shaped and constrained the ways in which peers interacted. Following Diao’s example, in this dissertation I add to this small but growing body of findings by investigating the interactions and perspectives of three Japanese undergraduate SA students and their TL-speaking peers at an English-medium Canadian university.  6  1.3 Peer language socialization In this study, like Diao (2013), I adopt a language socialization (LS) framework to interpret peers’ practices in this context (Duff & May, 2017; Ochs & Schieffelin, 2012). LS, broadly defined, is the process by which novice language users move toward expert status through participation in local communicative practices (Ochs, 1996). Studies that adopt an LS approach emphasize the role of meaningful interaction and recognize that learning is a dynamic and contingent process shaped and constrained by social, political, and personal factors. An increasing number of studies are adopting or advocating for LS perspectives to examine TL learning in SA, particularly in homestay interactions (Cardellio, 2016; Kinginger, 2017; Kinginger et al., 2014; Shively, 2018; Wang, 2010).  LS takes as its starting point the theoretical orientation that language learning through interaction is a negotiated and situated process in which speakers must collaboratively manage knowledge, affect, and positionality to accomplish learning (Duff & Talmy, 2011). In this sense, LS researchers understand the process of (language) learning as multi-directional and co-constructed. For this reason, an LS framework works well for understanding the ways in which peers move fluidly in interaction between “doing” language expert/novice and “doing” other roles, such as friends, roommates, or teammates.2  Despite the utility of LS for examining moment-to-moment shifts in participants’ roles, most LS research has investigated interactions between speakers who more clearly “fit” into expert and novice categories, such as parents and children or teachers and students (for a review                                                  2 Throughout this dissertation, I use italics to signal my use of terms that refer to social categories. In this way, I seek to acknowledge that these types of labels refer to participants’ identities in interaction or to locally relevant social categories that can be used as resources in interaction.  7  see Ochs & Schieffelin, 2012). Research that applies an LS framework to peer interaction is less common. To date, work that has used LS to investigate recordings of informal peer talk has typically involved children (for a review, see Kyratzis & Goodwin, 2017). LS work on interactions between adult peers, like the participants in this study, has more often been investigated from an institutional perspective in terms of workplace socialization (Jacoby & Gonzales, 1991; Li, 2000) or academic discourse socialization (for a review, see Kobayashi, Zappa-Hollman, & Duff, 2017). Diao’s (2013) study on conversations between SA students from the US and their Chinese roommates is a notable exception (see discussion in previous section). Thus, this study contributes to knowledge on peer language socialization by applying an LS framework to informal interactions between young adult peers.  1.4 Talk-about-language in peer interaction While it would of course be interesting to examine all aspects of peer interaction in the SA context, it was necessary, practically speaking, to focus on a few select practices that were particularly salient in peer talk in the local context. Thus, in this dissertation, in addition to providing more general descriptions of the interactional contexts of students’ reported peer socialization experiences, I have chosen to focus on a specific event called talk-about-language, which Levine (2009) has defined as talk “engaged in by language learners about language in natural settings” (p. 19).  Talk-about-language includes interactions that deal overtly with aspects of language form, including pronunciation, morphosyntax, lexis, or social/pragmatic use. As an example, Levine (2009) provides the following excerpt of an interaction between a German native speaker 8  and Bill, an SA student from the US. In this excerpt, the talk-about-language moment comes when both speakers focus momentarily on the lexical item “share” (Figure 1.1).  1  NS1   und zieht jemand anders für (name) ein?                  and is someone else moving in for (name)? 2   Bill  nee weil er uh (..) he shares? what’s share                  no because he uh (..) 3  NS1   teilen                  share 4  Bill  er teilt er teilt sein zimmer mit herr (name)                  he shares he shares his room with (name) 5  NS1   und er kann die miete alleine bezahlen?                  and he can pay the rent alone?  Figure 1.1 Talk-about-language. Excerpt reproduced from Levine (2009, p. 24, original italics, bolding, and capitalization). Talk-about-language events can be quite brief and, as in the excerpt above, often take the form of small side sequences during conversation; here, Bill is searching for the German word (teilen/teilt) for “share,” which his German partner helpfully supplies. However, talk-about-language also includes interactions in which there is an extended discussion of linguistic form or use. For instance, McGregor (2016) has used the term to refer to lengthy metalinguistic or metapragmatic discussions about pronoun use in German. While Levine’s (2009) examples come exclusively from native speaker-L2 speaker interactions, he acknowledges that talk-about-language also occurs between learners as well as multilingual speakers.  Based on the incidental nature of talk-about-language events, Levine distinguished talk-about-language from the more traditional term “focus on form,” which is usually associated with institutional or educational contexts. In addition, while results from focus-on-form studies usually examine the acquisition of specific forms or meanings, according to Levine it is not 9  necessary to think of talk-about-language as primarily a means for learning specific language items. Rather, he suggests that talk-about-language should be conceptualized as a social practice that may afford learners opportunities for legitimate participation in local communities. In other words, the value of participating in talk-about-language may have less to do with the forms that are negotiated within a specific event and more to do with the role that the event itself plays in allowing learners to be recognized as active participants in the conversation.  In describing talk-about-language as social practice, Levine (2009) encourages researchers to reflect on how participants negotiate norms for engaging in this practice. On the surface, engaging in talk-about-language seems a simple task: when speakers encounter a language problem, they should address it briefly and move on, as Bill does in the excerpt from Levine (2009). However, as Levine remarks, in reality, talk-about-language is much more complex and involves the principled interweaving of pedagogical interactions with more mundane communication. He explains that: There are particular patterns or “rules” of order that determine not only whether […] a Talk-about-Language event can happen (e.g., consider whether a Talk-about-Language would be likely to occur between two strangers in a train), but also, crucially, the ways it can occur and what can count as acceptable outcomes of the event. (p. 35) In this way, Levine acknowledges that, interactionally speaking, initiating talk-about-language can be a risky business. For language learners, asking a question might mean interrupting a peer’s exciting story. For a peer, providing correction may feel like teaching or criticizing instead of engaging in a friendly chat.  There is also some research evidence to suggest that speakers orient to and understand the complexities of “doing learning” in otherwise non-learning focused conversations. Dings (2012) 10  found in her analyses of peer interaction that speakers oriented to the complex rules around engaging in talk-about-language in their initial conversations. Her data showed that during their first meeting, her two participants, a US learner of Spanish and a local Spanish peer, explicitly discussed expert and novice rights and responsibilities. For example, the Spanish speaker explained his empathy with the challenges of language learning and aligned himself with the SA student as a fellow learner by describing his experiences learning English. He also encouraged the US student to make mistakes and assured her that she would not be judged for them. Similarly, Theodórsdóttir (Eskildsen & Theodórsdóttir, 2017; Theodórsdóttir, 2018) found that a Canadian SA student studying Icelandic often pre-negotiated rights to engage in talk-about-language or “practicing” with service staff before conversing with them on other mundane topics. These findings point to the potential importance of knowing and negotiating the rules for engaging in talk-about-language in informal interactions outside the classroom. Therefore, in this research I pay particular attention to the situated production of norms and expectations surrounding peers’ engagement in talk-about-language and discuss these in the context of prevailing assumptions about English and about language learning in SA contexts.  1.5 Peers at an internationalized Canadian university Before presenting the research questions and describing the organization of this dissertation, it is worth taking a moment to address my use of the term “peer” in this research. Readers may well be wondering why I have not yet specified what I mean by “peer” – TL-learning peers, same-aged peers, locals, Canadians, native speakers of English, Japanese native speakers? The answer is that not one of these labels applied to all peers in this study. Although I discuss participant recruitment and profiles of participants in more detail in Chapter 4, here I 11  explain some of the dilemmas I had about how best to describe participants in terms of “peer” socialization.  As mentioned in the introduction, the institution at which this research took place, Pacific Western University (PWU, pseudonym), is a large English-medium university located in western Canada. As with many large urban universities in English-speaking countries, the student body at PWU was and is highly linguistically and ethnically diverse. In addition, a high proportion of students, both domestic and international, are from Asian backgrounds. Thus, when I asked the three focal SA students in this study to recruit any peers with whom they had regular conversations in the TL, English, they naturally recruited people from a wide range of backgrounds and with diverse linguistic histories. While almost all the peers were more proficient in English than the SA students, almost all were multilingual. For example, one SA student recruited an international student from Britain who had just come back from an SA sojourn in Morocco, where she had been studying Arabic. Another SA student recruited a bilingual English-Chinese Canadian student who had immigrated to western Canada as a child. Clearly, then, terms like native speaker were insufficient for capturing the complexity of these transnational peers’ histories and identities. These peers’ various institutional statuses (as international students, domestic students, and exchange students) also called into question my initial intention to refer to the Japanese focal participants as SA students, since some of the peers were also conventionally SA students. I considered labelling SA students as learners; however, doing so would run the risk of implying that peer interaction was always about “language learning,” which of course was not the case.  Since the term peer implies a relatively symmetrical non-institutional relationship without the implication of any particularly deep ties or intimacy, it ultimately seemed most 12  appropriate to refer to both categories of participants in this way. In addition, by referring to all participants as “peers,” I opened up analytic space to consider the ways in which the participants treated each other at specific moments in interaction (e.g., as representatives of a national culture, as language learners, as friends). In other words, this choice allows me to adopt an emic perspective to interpreting the various ways in which the participants understood their relationships and to better account for the ways in which peers shifted between talk-about-language and other activities in their interaction. In the remainder of this dissertation, the two categories of participants are simply distinguished by their program affiliations: the three focal SA students in this study are referred to as JC peers (after the program in which they were enrolled) and the peers they recruited are referred to as PWU peers (based on their enrolment as students at PWU). 1.6 Reflexivity  Throughout this dissertation, I have included hyper-reflexive comments (Byrd Clark & Dervin, 2014) such as those found in section 1.5 above. According to Byrd Clark and Dervin (2014), hyper-reflexivity refers to the ongoing and complex process whereby researchers engage, throughout all phases of the research project, with issues of representation, methodological dilemmas, and the researcher’s positioning. As Byrd Clark and Dervin explain, these sorts of hyper-reflexive practices demand “a willingness to go and sit with the uncomfortableness and messiness of one’s own ideological attachments, ways of representing and investing, and a willingness, at the same time, to flexibly engage and negotiate meanings with one another” (p. 25). These authors also highlight the role of ongoing collaboration and negotiation with both participants and colleagues for promoting deeper engagement with research data. Throughout 13  this dissertation, I have included comments that address my (ongoing) realizations about the ways in my research design and personal assumptions shaped my interactions with the participants and my interpretations of their experiences. I also point out unexpected incidents that arose during data generation, ongoing adaptations to the study procedures, dilemmas concerning representation, and shifts in my thinking about how to interpret the findings.  For example, as discussed above, I struggled throughout the writing phase to choose labels for different types of participants. The more I reflected on the subject, the more I realized that my choice of participant labels was a pragmatic referential issue as well as a theoretical and epistemological one. My emic approach to data analysis dictated that I choose a label that was locally relevant to participants. Thus, I sought out the label most used by participants to talk about their relationships. The data showed that peers most frequently referred to each other simply as “friends”; however, the label friend failed to reflect the field of research I was trying to address (i.e., peer socialization) and it also became apparent that many of the PWU peers were conventionally-speaking not friends (or at least not close friends) to JC peers. Instead, it was pointed out to me by a committee member that the explicit mention of the friend category seemed to be doing important interactional work in the interviews. This realization led me to shift the focus of my analyses to better understand how participants used the term friend in talk when discussing their relationships and involvement in JC peers’ learning. Thus, my ultimate decision to describe participants as JC and PWU peers emerged out of extensive reflection and consultation.  In most cases, by calling into question the ways in which aspects of the study, such as recruitment categories and interview questions, influenced the kinds of data that were generated, I was able to provide a more nuanced and ultimately more revealing account of peer interaction 14  in this context. As researchers are increasingly recognizing, by attending to assumptions that are embedded in the research design, analysts are better equipped to move beyond foregone conclusions to examine the subtle ways in which participants engage with those assumptions (Potter & Hepburn, 2012). For instance, while at first the diversity of participants’ linguistic expertise and histories posed a challenge for writing up this study, acknowledging it enables me to make an important contribution to research in SA that I did not foresee at the outset of the study. Specifically, this research addresses the ways in which the multilingual and international backgrounds of these peers were important factors that shaped the JC peers’ experiences. Thus, by assuming a systematically reflexive approach to data analysis (as well as to data generation, as I discuss in Chapter 4), this study contributes to critically engaging with taken-for-granted assumptions about language learning in SA (these assumptions are discussed in Chapter 2).  1.7 Research questions and organization of the dissertation The research presented in this dissertation is a multiple case study involving three Japanese undergraduate SA students (hereafter known as JC peers) and the English-speaking peers with whom they formed relationships during the second semester of their sojourns (hereafter known as PWU peers). The research was guided by the following questions:  1. How do JC peers describe the contribution of peer interaction and/or relationships to their English learning? How do PWU peers describe their contributions to JC peers’ English learning?  2. What role (if any) does talk-about-language play in interaction, how is it initiated, and how is it discursively managed?  15  The first set of questions aimed to uncover the ways in which JC and PWU peers discursively constructed expert and novice roles in interviews. In this study, I conceptualized expert and novice roles as identity categories that were made locally relevant in specific instances of talk. By focusing on participants’ descriptions of their roles, I sought to examine the ways in which they (and I) co-constructed norms and expectations around the rights and responsibilities of novice peers in this context and characteristics linked to expert peers who were described as especially helpful or supportive. The second research question aimed to investigate the discursive characteristics of talk-about-language in informal interactions between JC students and PWU peers. I discuss the resources that participants used to engage in this kind of talk, its frequency, and its outcomes. In examining these two related aspects, I shed light on the benefits of viewing talk-about-language as a collaborative and negotiated social practice in this context.  In the chapters that follow, I introduce readers to key assumptions about language learning in the field of SA, describe the theory and methods that informed this study, and present the findings of the research. In Chapter 2, I critically discuss assumptions in SA research with a particular emphasis on the importance of the native speaker (TL) interlocutor in both students’ and researchers’ understandings of the value of the SA context. I also present the unique features of the multicultural and multilingual context in which this study was undertaken and point to the specificities of studying English (rather than other languages) abroad. In Chapter 3, I discuss language socialization (LS) in more depth and review how it has been applied to examine discursive social practices in SA contexts. I also introduce three constructs that informed the analyses and interpretation of the findings in this study: language ideologies, identity categories, and stancetaking. Chapter 4 presents the design of the study, including data generation and recruitment strategies, transcription conventions, case selection, coding, and analysis. Chapters 16  5-8 present findings related to the research questions about participants’ descriptions of and engagement in talk-about-language practices. Chapter 5 provides a detailed description of the three JC peers’ experiences (“stories”) of peer socialization during their eight-month sojourns and situates their reported experiences in their local and ideological contexts. Chapter 6 explores the ways in which participants and I co-constructed local categories of expert and novice and the ways in which language ideologies mediated these category constructions in interviews. Chapter 7 examines the responsibilities that PWU peers assigned to themselves and to JC peers when describing their roles in talk-about-language. Chapter 8 analyzes instances of talk-about-language in recorded peer conversations. Through a micro-analysis of peers’ talk, it examines the ways in which JC and PWU peers initiated and negotiated talk-about-language and the extent to which the language problems that were discussed were successfully resolved. It also highlights the value of understanding talk-about-language as a form of social practice. Finally, in Chapter 9, I conclude by drawing connections between these chapters, addressing limitations, and pointing to future directions for research. 1.8 Inspiration for the work As is often the case with doctoral work, my research was initially inspired by my own SA experience. In 2004, I left my Canadian university to complete my third year as an undergraduate exchange student in France. I departed on my journey determined to speak as much French as possible and to make many French friends. Much like the participants in this study, I lived in a university residence with local French students as well as international and exchange students. Despite my best efforts, I found it hard to make friends with French speakers. They could be funny and assertive when I, with my limited French language skills, could not. I found myself 17  spending a lot of time with students from the United Kingdom and Canada with whom I spoke mostly English. Thus, as is the case for many SA students, I was initially very discouraged. However, despite the initial setbacks, after a few months I was fortunate to meet a few French-speakers in my classes and my residence who were interested in being my friends. To my delight at the time, they were all what might conventionally be called native speakers and I valued them for their “Frenchness” and for the access they provided to other sympathetic French speakers. They took the time to mentor me and were patient with my language development.  When I returned home after my sojourn, many people told me that I had succeeded in improving my French because I had worked hard and avoided speaking too much English. While it is true that I was a dedicated language learner, I could not have succeeded if no one had taken the time to speak with me in the TL. Even then, I felt my learning was a joint effort and was due in large part to the patience and kindness of my French-speaking friends. In this dissertation, I intend to unpack the ways in which the kind of support I experienced in conversations with my French friends was actually achieved (or not achieved) in talk between JC and PWU peers in a very different context. As such, this dissertation will provide an empirically grounded examination of some of the affordances of peer talk in SA contexts and some of the constraints on maximizing those potential affordances. 18  Chapter 2: Study abroad at an internationalized university 2.1 Introduction  In Chapter 1, I located this study within a specific context: an eight-month SA program in Canada hosted at an internationalized university. Therefore, an important first step in this dissertation is to explain more precisely what I mean by the terms “SA” and “internationalized university.” It might seem logical to approach this question from a programmatic perspective. However, given this study’s focus on local understandings of how SA students “should” learn English during their sojourns and how peers “should” and “do” help them, for the purposes of this research, here I will define SA in terms of its taken-for-granted benefits for language learning.3 Thus, this chapter provides a critical overview of common assumptions and expectations associated with language learning in SA as well as features frequently associated with multilingual and multicultural English-medium campuses. It also discusses the ways in which these assumptions have shaped trends in SA research.  By unpacking the common understandings of SA found in the research literature as well as SA advertising and policy, I will be better able in later chapters to interpret the ways in which JC and PWU peers drew on, reproduced, resisted, or transformed these widespread discourses about the value of SA. Taking stock of SA assumptions also opens up space to problematize naturalized conceptions about the features of SA as a context for language learning (Surtees, 2016) and to identify key gaps in the literature. The first half of the chapter addresses three common SA assumptions while the second half provides a critical discussion of SA in English-                                                 3 A program overview and definition are still provided in Chapter 4 as part of the description of the context for the study. 19  speaking contexts where internationalization efforts have led to increasingly diverse multilingual student populations.  2.2 SA conceptualizations The term “SA” encompasses a wide variety of program designs and components, involving everything from intensive language learning, service learning, and humanitarian work, to subject-area courses, homestays, or cultural excursions (or a combination of all of these). Some scholars use “SA” to refer to any sojourn for educational purposes outside of the country of origin while others use the label to refer to sojourns of between one week and one year which do not lead to a degree in the host country (Kinginger, 2009). Engle and Engle (2003) identified as many as seven categories of SA programs, ranging from study tours for beginner language learners to one-year programs for advanced learners. The SA label also covers individual student exchanges (like my own SA sojourn in France described in Chapter 1) and cohort-based programs in which students attend classes with co-nationals (like the program reported in this research). Most SA researchers likely would agree that SA refers to programs that foster intercultural awareness and global citizenship – and sometimes foreign language skills – through international mobility. However, given the wide array of formats that exist, defining the common features of SA is a more difficult task than one might expect. Despite the challenges inherent in identifying a clear programmatic definition, SA has indeed become a coherent area of research. There are now two academic journals fully dedicated to the subject: Frontiers, established by Barbara Freed in 1995, which is a general journal about linguistic and non-linguistic aspects of learning outside one’s own country, and Study Abroad Research in Second Language Acquisition and International Education, established in 2016, which focuses more specifically on linguistic 20  dimensions of SA. The existence of these journals suggests that scholars view SA as an important domain for various kinds of interdisciplinary research. For Engle and Engle (2003), what separates SA from study at home is mainly students’ experiences with “focused and reflective interaction with the host culture” (p. 4).4 According to these authors, the aim of any SA sojourn “should be to present participants with a challenge—the emotional and intellectual challenge of direct, authentic cultural encounters and guided reflection upon those encounters” (p. 7). As a result, they conclude that most SA programs should be designed to give students opportunities to acquire linguistic and intercultural competence through interaction in the host culture. These kinds of statements, which are reproduced through much of the literature on SA, represent three key widespread assumptions that I will examine in this chapter: the beneficial role of interactions within an immersive environment, linguistic, and cultural, in the host community, and finally, mobility as a privileged means for accumulating social capital. In the next sections, I focus specifically on the ways in which these assumptions appear in literature and advertising related to sojourns that include language learning as a major program component.  2.2.1 Immersion  The first assumption about SA involves the role of immersion (Freed, 1998). Learning objectives in SA have traditionally revolved around language and culture learning that occurs as a result of being immersed in daily activities in another country or region. This assumption is often stated directly within SA and SA-related policy. For instance, in the US, Allen and Dupuy                                                  4 Engle and Engle (2003) distinguish between “knowledge-based” SA, which covers programs in the sciences and engineering that are meant to foster specific knowledge, and “culturalist” SA, which focuses on the objectives mentioned above. Like Engle and Engle, in this dissertation I refer to the latter type of SA.  21  (2012) have reviewed the way in which the “Communities standard” of the US Standards for foreign language learning links mobility and immersion. They point especially to a line of the policy that recommends that US students learn foreign languages by “traveling to communities and countries where the language is used extensively to further develop their language skills and understanding of culture” (p. 468). By connecting “traveling,” “extensive use,” and “language learning,” this policy naturalizes the connection between immersion in an environment where the language is frequently used and development of language proficiency.  SA marketing and networking websites also boldly proclaim similar links between travel and immersion. Such websites are designed to guide users through the process of selecting, planning, and engaging successfully in an educational experience abroad. For example, the website StudyandGoAbroad.com, a high-traffic consumer portal run by Education Dynamics targeted at US students, is designed both as an information hub for SA students and as a third party recruitment tool for SA program providers.5 In a post about the 25 benefits of SA, the website describes language acquisition as one of the top SA benefits, explaining that “immersion in another country is the quickest way to master the local language.” (http://www.studyandgoabroad.com/study-abroad/program-types/study-articles/benefits-studying-abroad/). These claims in advertising, as well as policy (such as the “communities standard” policy, discussed above) appear to filter down to SA students in important ways. A number of studies (e.g., Jackson, 2008; Meier & Daniels, 2013; Mitchell, 2015; Yang & Kim, 2011) have                                                  5  Studyabroad.com targets over 1.7 million US students and has been operating since 1995 (Source: https://www.educationdynamics.com/inquiry-generation/consumer-portals).    22  noted that SA students often expect to be surrounded by the TL and to have many opportunities to use it. However, while policy, advertising, and SA students themselves continue to conceptualize SA as an immersive experience, SA researchers have found that most students do not experience “immersion” in the conventional sense (i.e., frequent opportunities for sustained interaction in the TL). As early as the 1990s, Freed (1998) pointed to the difficulties that students face in making connections with TL speakers in the host community. For example, in one study (Freed, Segalowitz, & Dewey, 2004), Freed found that an intensive learning program in the home country offered more opportunities for TL use than the SA context. Other researchers, such as Coleman (2013, 2015) and Mas Alcolea (2017), have also explained that SA students tend to interact more with each other and with other international students than with local native speakers. Kinginger (2013) also points out that with the rapid rise of technology and social networking, students can now stay in touch with family members and friends, often calling them multiple times daily, reducing incentives and opportunities to make relationships in the host community. Finally, special attention is now being paid to the experiences of Anglophone SA students (Mitchell, Tracy-Ventura, & McManus, 2017), who are often spoken to in English in countries around the world, regardless of their attempts to use the TL.  Despite this widespread acknowledgement that simply enrolling in SA does not guarantee immersion (or learning), there is still a tendency in SA research to construe sojourns as potentially immersive experiences. In this sense, what Doerr (2013b) refers to as “the discourse of immersion” appears to serve as a powerful resource for making sense of the value of SA.6                                                  6 Doerr (2013b) does not provide a definition of discourse but from her work, it seems clear that she is referring to what Gee (2015) refers to as big D discourses, that is, social and historical identities and attributes that are widely recognizable in talk and text and which are used to make sense of small ‘d’ discourses (i.e., specific instances of talk and text).   23  Kubota (2016) also recently identified the notion of immersion as a cornerstone of the SA literature. She remarks that SA research continues to assume that proficiency gains are best realized by maximizing TL exposure (often through homestay arrangements) and by minimizing first language use, a notion that is also reinforced by theory and research in second language acquisition from past decades that emphasize the role of input, interaction, and output for second language learning (see Duff & Surtees, 2018 for discussion). This emphasis on immersion constructs the ideal learner subject in SA as “learning-by-doing” (Doerr, 2015, p. 361), or in other words, as primarily experiential learners who have a high degree of autonomy and agency. One consequence of viewing SA as potentially immersive is that SA research, particularly research that investigates language learning, tends to focus on the roles of broad forms of cultural and linguistic learning fostered through participation in mundane social activities, such as making friends, shopping for groceries, ordering at cafés, or taking public transit. Doerr (2012) explains that this focus on out-of-class experiences is unusual in research on educational contexts, and that research in most degree-seeking programs, for example, tends to focus on disciplinary outcomes and classroom practice. As a result of this focus on exploring and experiencing the mundane, she explains that SA often has “no clearly set goals, no supervision, and no evaluation” (p. 263). Doerr acknowledges that most SA programs involve some form of coursework and assignments. However, she writes that although SA coursework often teaches knowledge about the host society, language, or culture, “such knowledge about the host society is valued less than mundane experience” (p. 262) by both programs and students.  This focus on the value of exploring and experiencing the broader cultural environment with few guidelines means that SA research also tends to focus on the cultural, personal, and linguistic learning that results from students’ social experiences rather than from their classroom 24  encounters. Since 2000, rather than focusing on gains in writing ability, for example, studies have more often examined how out-of-class experiences such as homestays impact identity development (Anya, 2017; Benson et al., 2012; Jackson, 2008; Kinginger, 2004, 2015b; McGregor, 2012; Mitchell, Tracy-Ventura, & McManus, 2015; Mitchell et al., 2017; Pellegrino, 2005) and social/pragmatic learning (Bardovi-Harlig & Dörnyei, 1998; Barron, 2007; Diao, 2016; Kinginger & Belz, 2005; Shardakova, 2005; Shively, 2011; Taguchi, 2011).  Doerr (2015) has also observed that much SA research has focused on shedding light on factors that mitigate immersive experiences and prevent contact with host communities (e.g., cohorts with the same L1, lack of motivation, discriminatory experiences based on race or gender, homesickness, and low language proficiency) or has focused on developing program design elements to enhance it. These elements often include assignments such as homestay interview activities (Diao et al., 2011; Knight & Schmidt-Rinehart, 2010) and ethnography projects (Jackson, 2006; Lee, 2012; Roberts, Byram, Barro, Jordan, & Street, 2001).  This study is one such project that focuses on students’ out-of-class experiences. Like the research described above, it is predicated on the assumption that in an SA program in an English-speaking country, students will naturally be interested in and able to access opportunities for contact in English. It also assumes that such interactions will likely be valuable language learning opportunities for speakers of other languages (Japanese, in this case) and seeks to shed light on institutional initiatives that promote or constrain students’ access to opportunities for interaction. Like many of the studies cited in the paragraph above, it focuses on informal spoken interaction rather than disciplinary competences or classroom talk. Therefore, it is safe to say that through these design choices, this project is to some extent complicit in reproducing assumptions about the immersive potential of SA.  25  While this study does assume the potential value of SA for providing access to opportunities to use English, and thus for English language development, among other forms of learning, the findings of this project also resist SA assumptions that intersect with discourses of immersion, particularly with regard to the nature of the host community in which SA students are “immersed.” Assumptions about the host community and the types of language input and practice they provide are discussed in next section.  2.2.2 The native speaker  Doerr (2012) explains that what separates SA from learning at home is the assumption that SA naturally confronts students with cultural difference, making the otherness of the host community a defining pedagogical feature of the SA experience.7 She argues that SA materials used by students construct “unfamiliarity as a place-specific attraction” (p. 264) that cannot be replicated on home campuses. However, in most SA contexts, not all types of difference are treated as equally beneficial or desirable for language (or culture) learning. Rather, in most cases, it is native speakers (or idealized versions of native speakers) who are depicted as the embodiments and privileged conduits of unfamiliarity in the SA context and thus as legitimate members of the host community (Kubota, 2016b). Therefore, immersion in SA is not a neutral concept but refers to experiences within a specific “native” cultural/linguistic community.8  In today’s globalized world, most scholars would agree that this idealized conception of the host community as a culturally unified space inhabited by primarily monolingual native speakers is illusory and antiquated. It is increasingly rare to find communities where the majority                                                  7 This is not necessarily only a characteristic of SA but of foreign language education more generally. 8 I recognize that many of these labels are problematic and do not account for the complex situations of individual speakers. 26  of speakers have no knowledge of another language, and even more rare to find such an environment in the universities or large cities where most SA programs are hosted. As Coleman (2015) notes, “each of us, in an age of global migrations and internationalised campuses, possesses a linguistic repertoire embracing more or less complete but overlapping systems” (p. 45). SA researchers are now acknowledging that the spaces that SA students travel to are racially, ethnically, and linguistically diverse (Anya, 2017; Glaser, 2017; Kaypak & Ortaçtepe, 2014; Shiri, 2005); however, while the diversity of the overall context is acknowledged, interactions with local native speakers remain the central objective of most SA sojourns and continue to be an important goal for most SA students (Doerr, 2015).9 For instance, most interaction research in SA still focuses exclusively on interactions between SA students and native speakers (e.g., Diao, 2016; Dings, 2012; McGregor, 2016; Shively, 2018). An implication of the centrality of the native speaker in SA as an agent of TL socialization is that SA research does not tend to recognize host community members as potentially multilingual (and non-“native”). This lack of recognition is in contrast with much recent research in the field of applied linguistics more generally, where advocates for the multilingual turn (e.g., May, 2014) are highlighting the benefits of conceiving of speakers with varied repertoires as multilingual (or plurilingual) language users rather than as deficient native speakers. In my view, host members’ multilingual repertoires are likely overlooked for several reasons. First, the idealized native speaker is most often assumed to be monolingual, a                                                  9 A good example of this can be found in an SA blog, where Nate, a young man from the US, identifies his greatest personal language learning victory as “having a full conversation with a native Beijinger”. (http://thestudyabroadblog.com/steps-to-improve-your-language-fluency/). However, there is some evidence in Erasmus contexts that students value multilingual communities over native speaker communities (Kalocsai, 2009). 27  phenomenon referred to as the monolingual bias (May, 2014; Rampton, 1990).10 The connection between the native speaker (as a category type) and monolingualism can be traced to both the Chomskyan notion of the idealized native speaker as well as what has been referred to as the Herderian ideology of “one language, one nation” (Ricento, 2006; Woolard & Schieffelin, 1994), in which the people of a particular geo-political place – usually a nation-state – are assumed to speak one specific language and only that language as their mother tongue. Secondly, native speakers are thought to be more “authentic” users of the TL; portraying them as multilingual may make them appear less authentic. Indeed, in their volume on authenticity in SA, Van Compernolle and McGregor (2016) explain that the lay notion of authenticity, which they term authenticity of correspondence, is intimately connected “with the lexicogrammatical conventions and/or sociolinguistic and pragmatic practices of native speakers” (p. 1). Since theoretically, the role of hosts in SA is principally to provide “authentic” TL models and experiences of “difference,” host knowledge that does not contribute to that goal, including knowledge of other languages such as the SA students’ L1, often goes unmentioned.11  In addition, the continued importance of the native speaker concept in SA has often meant that interactions with other speakers of the TL (e.g., non-native, international, bilingual, heritage, and near-native speakers) are not mentioned, are avoided, or in the best cases, are conceptualized as stepping stones toward interaction with native speakers. For example, Coleman’s (2015) model of the concentric circles of socialization in SA identifies three groups                                                  10 The monolingual bias also refers to the notion that native-speaker status is what TL learners should strive to attain despite the fact that they can never be monolingual (May, 2014).  11 An exception to this would be work on less frequently studied languages in SA, such as Arabic and Chinese, in which host members’ knowledge of different dialects is noted as well as their knowledge of or desires to learn English (Diao, 2016; Trentman, 2013b). 28  with whom SA students interact: co-nationals, internationals, and locals. While Coleman acknowledges that interacting with all types of speakers can afford different sorts of learning opportunities, he still views the movement from internationals to locals as a “progression of friendships” (p. 44), with the native-speaker friendships being the highest level of progression. Students also report prioritizing interactions with native speakers and avoiding interactions with others and particularly co-national peers (Magnan & Back, 2007). Wilkinson and Hall (2002) for example found that, much like I did in my own sojourn (described in Chapter 1), SA students in France tried to avoid contact with co-nationals as much possible in order to take advantage of the language learning possibilities afforded by native speaker interactions. As a result of the perceived hierarchy between native speakers and other TL speakers in SA, there has been very little research on how multilingual speakers contribute to language learning in these contexts. However, the little amount of research available clearly demonstrates the value of being more inclusive of various others in SA. For example, Kaloscai (2009) found that while Erasmus students with first languages other than English viewed English native speakers “as uncaring and inefficient communicators” (p. 40), they valued other L2 learning peers for the willingness to accommodate to different ways of using English. Within their lingua franca communities, these students often engaged in creative hybrid language practices and collaboratively solved language problems. In a study on the value of interaction between L2 learning peers in SA, Hassall (2015) found that Australian SA students in Indonesia supported each other’s language learning through a variety of practices, including explicit discussions of pragmatic norms. Hassall found that students frequently noticed pragmatic moves in their Australian peers’ TL speech that they had been unable to notice in the fast speech of native speakers of Indonesian. Fernández (2016) also explored interaction between multilingual 29  students in SA. In her study, she examined interactions between Kaelyn, an SA student from the US in Argentina, and Analía, a local Spanish speaker. She explained that the two had been paired as conversation partners so that Kaelyn could practice Spanish during her time abroad. She also noted that Analía had volunteered for the program because she already had experience teaching English and Spanish as additional languages and planned to go the US on exchange the following year. In their interactions together, the two women often discussed different uses of personal pronouns and in a later interview, Kaelyn explained that Analía had used tú (the informal version most commonly used across all Spanish-speaking countries) more often than vos (the informal, local version of the pronoun used in some South American countries, such as Argentina, Uruguay, Paraguay) with her because she had become used to doing so with her students. By understanding and acknowledging Analía’s language history, Fernández was better able to make sense of Analía’s language choices and the strategies she used to support Kaelyn. These findings highlight the urgent need for more research on how a range of interlocutors contribute to language learning in SA contexts. This dissertation addresses this need by identifying and discussing the various language backgrounds of PWU peers and by complicating the notion of native speaker privilege and expertise in SA.   2.2.3 Capital accumulation  SA programs are one of many internationalization initiatives that are intended to “develop a globally competitive national labour force” (Association of Universities and Colleges of Canada, 2014, p. 4). Thus, a final key assumption about SA involves its value as an index of “worldliness” and communication skills, which are viewed as significant assets for future employability. SA websites often highlight the value of SA for developing the elusive soft skills valued in today’s labour market, including flexibility, innovativeness, motivation, endurance and 30  problem-solving skills. The following quote taken from www.studyabroad.com (the website discussed in section 2.2.1), demonstrates this tendency: As cliché as it sounds, by studying abroad you will become a much worldlier person, and become more marketable to grad schools and employers because of the intangible life skills you’re going to pick up along the way. Think of it as gaining a sort of cultural capital. The fact that you’ve studied abroad will answer many people’s questions about what type of person you are, even before they have to ask. (Student Guide, studyabroad.com)    This connection between “intangible life skills” and SA experience is also supported to some extent by large-scale studies. For example, findings from a study of the impact of Erasmus mobility programs on students’ employability (European Union, 2014) found that “Erasmus students have better employability skills after a stay abroad than 70% of all students” (p. 14). Given SA’s focus on the skilling of individuals for optimal participation in the global labour market, Kubota (2016b) has located SA as being embedded within a neoliberal social imaginary. A social imaginary, according to Rizvi and Lingard’s (2010) definition, is a largely implicit “way of thinking shared in society by ordinary people” [which is] “embedded in ideas and practices, carrying within it deeper normative notions and images, constitutive of a society” (Kubota, 2016b, p. 34). For Kubota (2016b), “a neoliberal social imaginary constructs an image of the neoliberal subject as equipped with communication skills, a global mindset, and intercultural competence and thus as competitive in global labour marketplaces” (p. 349). In her article, she argues that SA students are neoliberal subjects, who, through their excursions, acquire capital in the form of language skills as well as the ability to thrive in a new and foreign 31  environment. As will become evident in this study, JC peers also clearly cited neoliberal rationales for their sojourns, including the desire to become more employable in Japan.  One implication of this neoliberal vision has been that SA research tends to focus on the skills that SA students accumulate during their sojourns (i.e., outcomes) rather than SA students’ contributions to host communities (with the exception of the economic impact of fee-paying students) (Parry, 2015). Viewed from another angle, because SA is viewed as an educational product, the main focus of SA research has essentially been on customer satisfaction – that is, whether or not SA students (the customers) achieve the promised results. From this perspective, host communities are often conceptualized as service providers, and as such, their satisfaction and the outcomes of their experiences are seldom explored. For instance, while there are a large number of studies on students’ homestay experiences (see discussion in Chapter 1), only a few studies have investigated the perspectives of host families and identified what they seek to gain from the hosting experience (Doerr, 2013a; Knight & Schmidt-Rinehart, 2002; Parry, 2015; Shively, 2016). These studies have demonstrated mismatches in student and host family expectations that point to students’ neoliberal understandings of the SA experience. For example, several of these studies observed that while host families expected students to behave as a member of the family and sometimes expected social relations to extend past the sojourn experience, students sometimes treated the homestay primarily as paid accommodation and did not invest emotionally in the ways that homestay families had hoped for (Doerr, 2013a; Knight & Schmidt-Rinehart, 2002; Parry, 2015). In this study, by focusing on the perspectives of both PWU and JC peers, I seek to recognize the impact that SA students can have on local interlocutors and to provide a more equitable vision of the value of peer interaction in this context.  32  2.2.4 Summary Based on the assumed value of international sojourns as described in the sections above, SA has long enjoyed the unquestioning support of the general public and governments worldwide. Indeed, its benefits in many ways have been naturalized as common sense (Twombly, Salisbury, Tumanut, & Klute, 2012). Through a discussion of three key SA assumptions, I have identified how taken-for-granted knowledge about SA has influenced the ways in which SA has been conceptualized as a context for learning, as well as how it has been researched. Examining these assumptions has also allowed me to uncover several key gaps in this SA literature, namely 1) a dearth of research on classroom learning in SA, 2) a lack of acknowledgement of hosts’ multilingual repertoires, and 3) a scarcity of studies on host perspectives. This study primarily addresses the latter two gaps through its attention to peer perspectives and linguistic histories; however, throughout this study, I attend to the ways in which all three assumptions are produced and reproduced via my project design choices as well as through participants’ descriptions of their experiences and goals.  In the following section, I describe the specificity of contexts in which English is the language of instruction and the importance of distinguishing students’ experiences in these contexts from the experiences of students learning other TLs.  2.3 SA at internationalized universities Glaser (2017) recently distinguished two different SA contexts for English learning. Firstly, she identified programs which offer courses in English, but which are located in countries or regions where the official language is not English. These contexts, many of which are in Europe, usually involve deliberately internationalized curriculums and are often referred to 33  as English as a lingua franca (ELF) or English as a Medium of Instruction (EMI) contexts. Secondly, there are contexts like the one in this study in which the official language of the broader society is English, but where a high proportion of speakers are dominant in a language other than English (i.e., large urban universities in Canada, UK, US, and especially Australia). While Glaser referred to this second category as a “native-language setting,” I refer to this context type as an internationalized setting to more fully acknowledge the linguistic diversity that characterizes internationalized campuses and host communities and to move away from the problematic label of native-speaker.  Short-term programs hosted in internationalized universities are increasingly attracting attention from SA researchers, and with good reason. According to the Erasmus impact study (European Union, 2014), students reported that over 60% of all exchange programs in Europe are delivered in English and over half the world’s international students are hosted at institutions in English-speaking countries (Mitchell et al., 2017). While most statistics on international mobility report the number of students in degree-granting programs rather than in short-term programs like the JC program discussed in this study, there have also been reports of increases in the number of students enrolled in short-term programs hosted in English-speaking countries. For instance, the number of Japanese undergraduates participating in programs lasting one year or less grew from 26,451 in 2011, to 27,390 in 2013 (MEXT, 2015). The Japanese government also aimed to double that number by 2017 (see discussion of Go Global in Mock, Kawamura, & Naganuma, 2016). These trends have meant that there has been an increase in the amount of work examining SA experiences and perspectives for English learners in internationalized contexts (Benson, Barkhuizen, Bodycott, & Brown, 2013; Davis, 2007; Doerr, 2015; Ellwood, 2011; Glaser, 2017; Jackson, 2008, 2011; Taguchi, 2011; Tanaka & Ellis, 2003). In the sections 34  that follow, I identify two key aspects of language learning in internationalized universities and discuss their relevance for understanding the results of the research reported in this dissertation.  2.3.1 English as a global language An especially important facet of language learning in internationalized SA contexts concerns the instrumental value that is often assigned to English vs. other languages. As the de facto language of globalization (and neoliberalism), English is not only viewed as a tool for connecting with specific cultural others (i.e., native speakers) but rather as a gateway to communication with an international community (Canagarajah, 2012; Pennycook, 2017). Pennycook (2017) contends that, rightly or wrongly, English is often positioned by people and policymakers alike as a neutral language that extends beyond culture to allow people all over the world to communicate with each other. As such, knowledge of English tends to be perceived through a neoliberal lens as a commodity to be acquired to ensure success in and today’s international marketplace (Phillipson, 2008).  In discussing the role of English in SA, Mitchell, Tracy-Ventura and McManus (2017) explain that given its associated power and prestige, SA students have significantly more incentive to learn English than they do to learn other languages. This statement is particularly true for those students wishing to pursue graduate education. According to the Times Higher Education ranking, three quarters of the 100 most highly ranked universities give instruction in English (https://www.timeshighereducation.com/), making English not only the language of travel and of business, but also the language of the educated elite. As a key gatekeeper to university entrance, English also plays an important role in shaping who can and cannot have access to sought-after educational resources in countries where English is not the national language. For example, in Japan, English is a required examination subject for entrance into high 35  school and university (Outline of the National Center for University Entrance Examinations, 2017) 12 and large businesses often rely on English language tests for hiring and promotion (Kubota, 2011). Indeed, the students in this study were required by their home university to take a standardized English exam before, during, and after their sojourns. Such intensive testing schemes are rarely observed for programs involving languages other than English (Doerr, 2015).  An additional consequence of viewing English as a means for international (rather than national) communication is that the emphasis on cultural authenticity (and thus on the native speaker) is often somewhat mitigated in programs hosted at internationalized universities. Program advertising may instead emphasize SA as an opportunity to network with international students from all over the world. For example, the popular international education website www.hotcoursesabroad.com/usa/ indicates on its homepage “[t]he diversity of cultures and cities in the United States is an invitation to expand your academic horizons” (n.p.). The Government of Canada’s (2014) branding strategy for the Canadian higher education sector also emphasizes Canada’s reputation for multiculturalism, calling Canada “a welcoming, safe and multicultural country offering high-quality education at an attractive price” (p. 7). The program in which this study was conducted also specifically mentions local diversity and multiculturalism in its online advertising. Thus, the perceived role of English as a bridge for intercultural communication may partially disrupt the ideological connection between SA and the native speaker. Indeed, in this study, the connection between intercultural communication and international experience was reproduced in various ways in interviews with both PWU and JC peers, pointing to the                                                  12 According to the National Center for University Entrance Examinations, students can also take another foreign language to satisfy the requirement to enter university. However, English is the language selected by default. It is also the only language test which includes a listening component.  36  importance and power of discourses that construe English as a global or international language in internationalized SA contexts.  2.3.2 Campus diversity  Given the large number of international students who pursue education in English-speaking institutions, many internationalized universities in the UK, the US, Canada, and Australia have highly diverse student bodies. This is especially true in Canada, where universities have been particularly aggressive in their international recruitment policies. According to Anderson (2015), between 2000 and 2011, the number of international students in Canada grew over 99% (from 89,532 to 178,491), at which point 8.2% of all tertiary enrolments were international students (more than double the percentage of international enrolments in the US). Since 2011, the Canadian government’s internationalization strategy has set a target to double international student numbers in Canada by 2022 (Trilokekar, 2015).  While discussions of international student recruitment frequently reference its economic benefits for host institutions, Haigh (2014) also points to a notable increase in idealistic institutional discourses that position international students as contributors to campus multiculturalism (see, for example, the 2017 special issue of the Journal of International Education on the topic of global citizenship). He points out that many institutional initiatives on internationalized campuses are meant to foster interaction and intercultural understanding through the promotion of collaboration between international and domestic groups (e.g., Reid & Garson, 2017). From these “intercultural” encounters, both international and domestic students theoretically have the opportunity to become global citizens, by “learning to live together sustainably [as] moral cosmopolitans” and extending the “notion of the ‘we’ to include those global others” (Haigh, 2014, p. 13). In this sense, inbound SA students are treated in institutional 37  discourses as contributors to (rather than consumers of) cultural diversity and, crucially, as resources for domestic students’ learning (Trilokekar, 2016).  In reality, of course, international students are not integrated so seamlessly into these internationalized student communities. Schartner and Cho (2017) found evidence to suggest that domestic students perceive internationalization efforts as strategies for revenue generation rather than as opportunities for intercultural learning. In their study conducted in the UK, when asked about the cultural exchange aspect of internationalization, focus group participants expressed cynicism, highlighting the financial motives for international student recruitment and the lack of support for international students and for staff who work with them.  In the last decade, in Canada, debates around the merits and challenges of internationalization have also appeared in the mainstream media (e.g., Quinn, 2012). For instance, in 2010, Mclean’s magazine, a prominent source for information about higher education in Canada, printed a story in which English-speaking students in Ontario claimed to avoid certain universities because of their reputations “for being too Asian” (Findlay & Köhler, 2010, n.p.). Research on international students’ experiences in Canada has also identified implicit hierarchies in the ways in which different international groups are treated, and noted the ways in which certain less visible groups (i.e., Caucasian sojourners) tend to receive less support (Kenyon, Frohard-Dourlent, & Roth, 2012).  At PWU, where this study was conducted, international students accounted for approximately 20% of the student population and, at the time of the study, issues around student diversity and integration were being highlighted across campus in a variety of ways.13 For                                                  13 In order to protect the identity of the university, I have not provided precise details or media links.  38  instance, there had been a recent protest jointly organized by domestic and international students when the government decided to raise tuition fees for international students (which were already much higher than those of domestic students). A new campus magazine featuring stories from students of diverse backgrounds was also published that year. Finally, there had been controversy in the media about a bridging program that accepted students who did not meet the mainstream English language proficiency requirements at PWU, yet which did not accept domestic students with similar levels of language proficiency.  What all these events and various media publications demonstrate is that internationalization, international students, and English proficiency/learning are sensitive issues frequently discussed on internationalized campuses like PWU. Indeed, this sensitivity was apparent in many of the PWU peer interviews, where explicit evaluations of JC peers’ proficiency were often avoided, as were explicit mentions of terms like “native-speaker.”    2.3.3 Summary In the second half of the chapter, I have discussed two interrelated issues related to English learning on internationalized campuses. Section 2.3.1 highlighted the role of English as a means to becoming a member of the global, educated elite. Section 2.3.2 described how the global desire for English has produced university environments in English-speaking countries that are highly linguistically and ethnically diverse. For SA students on highly internationalized campuses (as in the case of the JC peers in this study), these phenomena mean that they are more likely to encounter a wide variety of English-speaking interlocutors during their sojourns. In addition, given the emphasis on cultural diversity and intercultural learning on these campuses and in light of discourses that view English as a global (rather than national) language, SA students are also more likely to value their encounters with those diverse fellow students (or 39  perhaps less likely to prefer interactions with native speakers of English). However, local debates around integration, diversity, and internationalization on these campuses may also lead to complicated tensions between members of various communities. As the findings of this dissertation will show, these discourses and tensions were all relevant to varying extents for understanding participants’ descriptions of their motivations and experiences with peers at PWU.  2.4 Conclusion In this chapter, I have described the context of this study in terms of the assumptions and discourses that are commonly associated with both SA and English. Specifically, I have addressed the assumed benefits of SA for language learning and discussed the ways in which this study contributes to reifying or challenging those assumptions. I have also endeavoured to distinguish the specific features of SA sojourns that occur in English-speaking environments, and particularly at institutions with highly diverse student bodies. By providing a critical overview of trends and assumptions in the SA literature, I have laid the foundation for discussing my findings on peer interaction at PWU in a more reflexive and critical way. I will refer to these findings and assumptions throughout the results chapters of this dissertation in order to better interpret peers’ discussions of their goals, expectations, and experiences during their sojourn. In Chapter 3, I turn to a discussion of the theoretical framework that guided this study.   40  Chapter 3: Theory and analytic focus 3.1 Introduction In the introduction to this dissertation, I described this study as one that investigates how talk-about-language can be accomplished between peers in an internationalized SA context – the JC program at PWU. This characterization of my research focus reflects my understanding of language learning as a fundamentally social enterprise that is jointly constructed and mediated by local norms and ideologies. In this chapter, I provide an overview of the theory and research that I have used to think about and investigate English language learning and peer talk in this context. I first present the main tenets of language socialization (LS), the theoretical framework that guided this study. Second, in order to locate the contributions of this research, I review literature that has adopted an LS approach to understanding SA interaction. Finally, I discuss the aspects of the LS process that constitute the focal points of this research and define several key constructs, namely: language ideologies, expert and novice categories, and stancetaking practices.   3.2 Language socialization Language socialization, broadly defined, is the process by which novice language users move toward expert status through participation in local communicative practices (Ochs, 1996). From an LS perspective, learning to perform valued linguistic practices involves interacting in various ways with experts to access opportunities for learning (for example, via explicit explanations, negotiation, observation, opportunities for language use and feedback, evaluation, and construction of new norms). Thus, LS perspectives place interaction at the heart of the 41  socialization process, considering it to be “a major, if not the major tool” (Ochs, 1986, p. 3) for the development of social and cultural knowledge and resources.  3.2.1 Language in LS LS views language and culture as inseparable, mutually constitutive constructs and the use of language as situated cultural practice (Ochs, 1996; Watson-Gegeo, 2004). LS research identifies firstly, the cultural and linguistic practices that have come to be valued by members of a particular community (socialized practices) and secondly, how those practices are learned and negotiated, both implicitly and explicitly, by those seeking membership in that community (or socializing practices). From an LS perspective, cultural and linguistic practices not only accomplish communication aims but are also used to accomplish pro-social aims such as creating and maintaining relationships with others (Joaquin & Schumann, 2013). LS also acknowledges that the ways in which speakers configure interactional resources (e.g., grammatical markers, vocabulary, laughter, pitch, pronunciation, silence) in social activities are constrained by local values and norms while simultaneously reproducing, resisting, or sometimes transforming those same values and norms.  The notion that language practices produce and reproduce systems of cultural value is closely linked to the concept of indexicality (Ochs, 1996). According to the principle of indexicality, specific linguistic choices are linked to cultural systems of value and function as indexes, which can be defined as “interpretative leads between what is said and the social occasion in which it is being produced” (Blommaert, 2005, p. 11). According to Ochs (1986), “language socialization can be seen as unfolding understanding of the indexical potential of particular linguistic forms and the skill to apply that understanding to construct situations with other interlocutors” (p. 419). The range of social meanings that may be indexed is immeasurable, 42  including but not limited to “practices and theories for acting, feeling, and knowing, along with their material and institutional products” (Ochs, 1996, p. 409). When individuals use language, they draw on their awareness of previously established cultural meanings, a process known as indexical presupposition (see, for example, Silverstein, 2003), and simultaneously create a new context from which such presuppositions can be drawn in the future, termed indexical entailment. Thus, language use is understood to contribute to the reification, reproduction, and transformation of cultural systems of value (Fader, 2012; Kulick & Schieffelin, 2004; Riley, 2007).  A central aim of LS research is to make explicit the tacit indexical links between a community’s linguistic forms and cultural meanings. Only then can scholars investigate the processes by which novices learn to navigate those layers of indexical meaning. Research in LS has focused on linguistic forms indexing affective and epistemic stance (e.g., Cook, 2012), moral stances (e.g., Fader, 2012), politeness (e.g., Burdelski, 2012), and language ideologies (Mori, 2014) to name only a few (these concepts will be expanded on later in the chapter). The practical implication of this language-culture connection is that they must always be investigated in tandem and with reference to their contexts of production.  3.2.2 Learning in LS LS takes engagement as the primary means by which novices learn to perform language practices. This interaction-based notion of learning differs from input-based theories of language acquisition in that it privileges meaningful interaction in and exposure to situated activities over quantifiable exposure to linguistic forms (De Leon, 2012). In LS, engagement, or participation, is also an indicator of competence. Whereas cognitive models measure competence in terms of accurate reproduction of grammar and lexis, competence in LS, which draws on early work by 43  Gumperz and Hymes (1972), is viewed as the ability to participate in the practices of a given community “in recognizably social ways” (Howard, 2012, p. 343). However, as socialization is an ongoing and negotiated process, competence is not viewed as something that can be attained and stored: the rules for participation are constantly changing as individuals grow up, join new communities with ever-changing norms, and take on new roles. Thus, competence must constantly be re-established, novices must continually be re-socialized, and experts must adapt as circumstances and cultures change. Conceiving of competence in this way reflects LS’s focus on language as a tool in “the human quest for belonging, connectedness, and affirmation in family, community, and society” (Howard, 2012, p. 341).  While interaction with community members can often result in novices’ socialization into community practices, it is important to highlight that LS approaches acknowledge that uptake and participation are also mediated by social factors such as identities, expectations, and ideologies as well as level of exposure and engagement. Indeed, in contexts of second language education and migration, such as SA, novices may be unable to access opportunities for participation or are discouraged from participation, leading to a number of possible negative outcomes including “ambivalence, defiance, resistance to or rejection of the target language, culture, or community (or aspects thereof), or prematurely terminated or suspended L2 learning” (Duff, 2012, p. 6). For this reason, LS research on additional or second language learning in particular (Duff & Talmy, 2011) has highlighted how socialization into a given community’s practices is also not always “successful,” often despite opportunities and desire for interaction with local experts.  44  3.2.3 Multi-directionality in LS For this study, a particularly important aspect of the LS framework is its emphasis on the collaborative and co-constructed nature of learning and participation (Duff & Talmy, 2011). All participants, novices included, shape the socialization process to differing degrees and novice and expert roles are viewed as fluid and co-constructed within and across interactions.14 In this sense, LS has integrated recent concerns in sociolinguistics and applied linguistics that prioritize multi-directionality, agency, and fluidity (Duff, 2015; Duff & Doherty, 2015). LS research in diverse multilingual contexts also acknowledges that norms for participation are not monocultural but rather are subject to ongoing joint negotiation, creation, and transformation (Duff, 2015). This is particularly true in contact zones like internationalized universities in which participants coming from different countries, communities, and linguistic backgrounds co-create norms to suit their communicative needs (Canagarajah, 2012; Kalocsai, 2009). To address the collaborative nature of socialization, I investigate the talk and perspectives of both JC and PWU peers, thereby providing a multifaceted picture of the norms and practices surrounding peer interaction in this context.  3.2.4 Methods in LS Theorizing language use as a collaborative and socially mediated achievement has important epistemological implications. LS work studies language in its interactive, cultural, and socio-political context (Ochs, 1996; Watson-Gegeo, 2004). To generate the necessary data to gain productive insight into local meaning-making practices, most LS scholars agree that to                                                  14 What constitutes agentic action within interaction is a contested issue in the field of LS, one which I do not have the space to review here (Duff & Doherty, 2015; Fogle, 2012; E. R. Miller, 2014) 45  investigate development or change over time, longitudinal, qualitative, and ethnographic approaches are appropriate for investigating socialization processes (Duff, 2008b; Kulick & Schieffelin, 2004; Watson-Gegeo, 2004). For Duff and Talmy (2011), LS research ideally ought to “document changes in language and other social practices, explain development in terms of socialization, and involve close analysis of a rich primary data record derived from participant observation, documents, and audio- and/or video-recordings, among other methods” (p. 105). LS researchers also approach data from a participant-relevant perspective to examine how participants themselves orient to and co-construct local meanings. In other words, expert and novice roles are not pre-assigned and fixed, rather the LS analyst examines how participants use language in ways that make local expert and novice identities relevant and seeks to provide a rich description of participants’ practices. In the next section of this chapter, I discuss how LS has been taken up in SA research, including key findings and methodological insights.    3.3 LS research in SA In recent years, a number of scholars have highlighted the utility of the LS framework for investigating language learning in the SA context (Kinginger, 2012, 2017; Shively, 2018; Wang, 2010). Kinginger (2017) in particular has advocated for an LS approach to address fundamental unanswered questions in SA: “why do some students prosper while others do not? What are the precise qualities of study abroad experiences as they relate to language learning?” (p. 1). Wang (2010) has similarly emphasized the benefits of focusing on situated learning processes (i.e., moment-to-moment interaction), as LS approaches do. She contends that by focusing on the qualitative aspects of interactions over the sojourn, LS is poised to provide crucial insight into the variable outcomes that have often been reported for SA language learners as well as the 46  particular affordances of interaction with various cultural others (Kinginger, 2009). Shively (2018) also notes that LS’s focus on participation is a useful frame for conceptualizing learning in SA, since sojourns are often designed to provide novel opportunities for participation in activities in TL communities (e.g., via homestays, clubs, service learning).  A relatively large number of studies in SA have adopted LS perspectives broadly-speaking. These usually have taken the form of single or multiple case studies that have examined SA students’ language contact, social networks, or community participation and often highlight barriers to participation, individual learners’ trajectories, and identity and academic development (e.g., Isabelli-Garcia, 2004; Jackson, 2008; McGregor, 2012; Pellegrino, 2005; Trentman, 2012; Zappa, 2007). In general, these studies have provided evidence of the challenges that SA students face forming relationships with local TL speakers and the complex ways in which encounters with the community prompt SA students to reflect on their national, gender, and learner identities (for an overview of these studies, see Kinginger, 2012). Taken together, they provide evidence that proximity to TL speakers does not always result in meaningful interactions with TL speakers and thus these studies serve to problematize the ideologies of immersion discussed in Chapter 2. Based on findings from the LS literature in SA, Shively (2018) notes that socialization opportunities rely on the dispositions, ideologies, identities, and relationships of all participants. While the studies cited above provide rich insights into SA students’ perspectives and the barriers and opportunities they encounter, they do not include recorded naturalistic interactions and thus provide only limited insights into the specific affordances of host community interactions for language learning.   47  3.4 LS studies of interaction in SA  As indicated in Chapter 1, my aim in this dissertation is to investigate peer language socialization by analyzing one type of language event in particular: talk-about-language. Therefore, in the section that follows, I review studies that incorporated analyses of SA student interactions, highlighting key findings and methodological approaches. These studies, which are relatively few in number, bring various discourse analytic methods to bear on recordings and videos of interactions collected throughout the sojourn, most often by the participants themselves.15 These studies have investigated socialization into performances of gender and politeness indexed by specific language markers (e.g., sentence final particles), broader ideologies, (e.g., cultural notions of taste), and strategies for socializing novices (e.g., foreigner and teacher talk). Table 3.1 presents a list of SA studies in reverse chronological order that have investigated LS through the analysis of recorded interaction in SA.  Table 3.1 SA Studies with Recorded Interaction Study SA participants TL/context  Socializing practices Socialized practices, ideologies, identities Kinginger and Wu (2018)  Cardellio (2016) 2 US students,  Chinese, China   3 US students, Italian, Italy Informal talk with roommates   Encounters with service learning providers Personal narratives Humour   Directives  *Kinginger, Lee, Wu, and Tan (2014) 3 US students, Mandarin, China Meal-time conversation about food with host families Linguistic and gestural resources for presenting food, ideologies related to food  *Diao (2013, 2016) 4 US students, Mandarin, China Explicit discussion of particles with roommates Affective sentence-final particles                                                  15 Iino (2006) refers to this technique as “remote observation method” (p. 156). 48  Brown (2013) 3 UK/1 Australian, Korean, Korea Interactions with various local Korean speakers Korean honorifics (contaymal and panmal) *Shively (2013) 1 US student, Spanish, Spain Interactions with host parents and a local peer Interactional resources for engaging in humour (e.g., humourous revoicing) *Cook (2012) 9 US students, Japanese, Japan Participation in talk using epistemic stance marking Epistemic stance marking (deshoo) during meal-time talk with host family *Shively (2011) 7 US students, Spanish, Spain Service encounters, reflective journals, and in-class discussions about language use Pragmatic moves and modals in service encounters *Ishida (2010) 4 US students, Japanese, Japan Informal peer conversation Japanese modal expressions (ne/kamo) during peer talk *Cook (2006) 8 US students, Japanese, Japan Negotiating folk beliefs in homestay meal-time interaction Nihonjinron ideologies *DuFon (2006) 5 US students, Indonesian, Indonesia Talking about food in meal-time interaction with host families Beliefs and ways of talking about taste post SA *Iino (2006) 30 US students, Japanese, Japan Interaction with host family Nihonjinron ideologies *Dufon (1999) 6 US students, Indonesian, Indonesia Interaction with native speakers and correction practices  Experience questions, greetings, terms of address Stewart & Talburt (1999) 6 US students, Spanish, Spain In-class discussions of racialized and gendered experiences Gendered identities Wilkinson (1998; 2002) 4 US students, French, France Initiation-response-evaluation sequences in homestay interactions Learner identities Siegal (1996) 4 US students, Japanese, Japan Participation in institutional talk Politeness and gendered speech  Note. *Studies which explicitly cite LS as a framework 49  The common thread across these studies is a focus on how participants are socialized into context-specific linguistic practices, through context-specific linguistic practices.16 Typically, they also account for local systems of value by studying participants’ beliefs and expectations, and how they relate to broader societal discourses. By connecting transcripts of naturalistic data with broader cultural and sociopolitical ideologies, these studies make more concrete the link between the language SA sojourners encounter and the wider systems of sociocultural value to which they belong. They also document how the sojourners themselves come to a deeper understanding of the layers of indexical meaning attached to different forms in different contexts. For instance, DuFon (2006) examined how she and four other learners of Indonesian came to a new appreciation of food through practices in which food was discussed as a source of pleasure and in terms of its health-related qualities. Over time, SA students in Dufon’s study came to understand how to participate in meal-time talk appropriately (e.g., to be critical when food was not to their liking) and in some cases, promoted or maintained these practices upon returning from their sojourns.   3.4.1 Methodological insights   The studies listed in Table 3.1 all used multiple data sources in conjunction with recorded interactions. Their use of multiple data sources allowed the researchers to make connections between interactionally accomplished events and students’ narratives in interviews and journals. For instance, Shively’s (2011) pragmatics-focused study of service encounters in Spain demonstrates the insights to be gained from multiple data sources. Shively presented data in                                                  16 Increasingly researchers are examining other forms of semiotic practice, particularly gesture. However, in this dissertation most interactions were not video-taped and thus an analysis of gesture was not possible.   50  which one student, Jared, inappropriately uses the expression que tal (i.e., how are you) with a Spanish shopkeeper. She was then able to connect this incident to Jared’s later metapragmatic reflections in his journal in which he recognized the shopkeeper’s unease and declared his intention to act differently in subsequent encounters. In this way, Shively captures how Jared learned the rules of appropriate participation in ways that would not have been observable through recordings or journals alone. She was also able to track Jared’s subsequent use of que tal in interactions following his revelation.  An additional advantage of this discourse-focused LS approach is the ability to track changes in the frequencies with which specific linguistic choices are made by different interlocutors (e.g., SA students vs. native-speaking hosts or peers). This also contributes to understanding the extent to which the linguistic practices of local speakers are being taken up by SA students. Diao’s (2013) study of gendered sentence-final particles in Mandarin is an exceptional example of this. Using weekly recordings of SA students’ interactions with their Mandarin-speaking roommates, she was able to observe the relative frequencies of different particles, finding that two US participants in particular showed evidence of moving toward patterns that more closely resembled those of their Chinese roommates. In addition, she was able to capture several instances in which the student and roommate explicitly negotiated when such particles should be used and observed how the information was taken up by the SA students both within the interaction and in the recordings that followed. In this way, Diao was able to observe the non-linear trajectories of the socialization process as it unfolded.  Finally, by drawing on transcripts of interaction rather than student reports alone, these studies provide more space for examining the actual linguistic contributions of the hosts, locals, peers and experts with whom SA students interact. It also allows the researcher to analyze how 51  SA students impact the practices of the experts with whom they interact. Cook (2006), for example, analyzed dinnertime conversations between SA students from the US and their Japanese hosts and found it was not only the students who negotiated their own cultural assumptions but also members from the host families who through the dynamic process of co-telling transformed aspects of their beliefs. The author notes that discussions of folk beliefs related to nihonjinron (theories of Japaneseness) in particular were an “opportunity space for [all] the participants to co-construct shared perspectives and emotions” (p. 147). Thus, LS research of this kind better accounts for the dynamic co-construction that occurs when relationships are built and helps to avoid the pitfalls of presenting interaction partners as language dispensers or the static characters in student narratives. 3.4.2 Key findings In general, these studies have shown that through participation in interactions with local speakers, SA students learn to engage with local speakers and negotiate new forms, often in agentive ways. The changes that occur in SA students’ language use are often subtle (e.g., use of gendered particles or affective stances) and would have likely been difficult to capture via traditional assessment methods, making this research especially valuable for understanding the affordances of interaction for language development in these contexts. These studies also have the benefit of providing rich descriptions of the learning contexts, relationships, and experiences of students. Not only do the researchers examine how interactions are accomplished, they attend to the ways in which locally circulating ideologies around issues of language, gender, and taste etc. serve to mediate what is and is not treated as appropriate or desirable. In so doing, they illuminate what is learned by students at the micro level as well as identifying the community factors and discourses that allow for that learning to take place at the macro level.  52  An important contribution of these studies is that they provide concrete evidence to problematize the use of native standards as the yardstick by which to measure language learning outcomes in SA. Diao (2013), for example, found that no student, regardless of their level of engagement with the community, was observed to use gendered particles consistently in “native-like” ways. While this result may be due to students’ emerging competence or lack of emphasis on such particles in textbooks, evidence from other studies also suggests that students resist native norms, particularly if they conflict with students’ values or identities. Siegal (1996) recounts the case of Mary, a graduate SA student in Japan, who used less humble language in her L2 Japanese than would normally be appropriate with a professor in order to preserve her self-image as a scholar. Siegal explains that “when learners study abroad it is necessary to consider the conscious and unconscious desires of the learner to maintain her image and the resulting language use which might deviate from native speaker norms” (p. 240). This view is also supported by findings in Kinginger (2008) of a student who was extremely successful in engaging with the French community but purposefully chose not to use the formal address pronoun vous (i.e., you), explaining that it did not fit his personality.  Findings from this literature also point to the fact that hosts often interact differently with SA students than they do with other members of the host community. Much of the SA literature assumes that students will experience interactions similar to those engaged in by groups of native speakers. However, the studies discussed here find that locals modify their practices to communicate in ways they perceive are more appropriate for SA students’ status as foreigners and learners (Kinginger, 2015). Iino (2006), for example, found that Japanese host families switched from their home Kyoto dialect to Tokyo dialect when speaking with students in order to provide a more standardized model. In addition, he found that hosts treated mistakes as “cute,” 53  while “very good Japanese was an object of wonderment and laughter” (p. 166) and was perceived by hosts as forced and inauthentic. Modified foreigner talk was also discussed in Wilkinson’s (1998, 2002) study of French homestay encounters. Using conversation analysis to investigate short encounters between host-family members and SA students, Wilkinson found evidence of initiation-response-evaluation sequences (Mehan, 1979) common to Western classroom routines. Although she had only recorded a small number of interactions, the prevalence of these routines in what were meant to be “natural” conversations was striking, leading her to conclude that the talk SA students experienced more closely resembled teacher-student discourse than informal discussions with friendly locals. This provides further evidence that interactions in SA cannot be conceptualized as opportunities for exposure to TL input but are in fact negotiated encounters in which both the experts and novices co-construct opportunities for learning based on the expectations, identities, and knowledge they bring to the interaction. 3.4.3 Gaps in the literature While these studies bring much needed attention to various forms of interaction in SA there is more work to be done in expanding the range of contexts, types of interaction, and interlocutors that are investigated from an LS perspective. For instance, all the studies cited involved US (or English-speaking) students learning languages other than English. As discussed in Chapter 2, English is often treated as having a different status and function than other languages and the ideologies that mediate interaction in English are likely to differ from those that mediate interactions in languages such as French, Spanish, Mandarin, or Japanese. To date, the SA studies that incorporate LS and analyses of recorded interaction have focused primarily on homestay interaction. Only a few studies have investigated peer interaction (Brown, 2013; 54  Diao, 2013; Ishida, 2010; Kinginger & Wu, 2018; Shively, 2013). In addition, the host members with whom SA students interact in these studies are all native speakers and little to no attention has been paid to the ways in which multilingual or more proficient TL-speaking others mediate students’ language development in these contexts. Finally, all these studies have emphasized interview and self-report data from SA students only. That is to say that the perspectives of host members are rarely attended to systematically, and if they are, they are interpreted through the reports of SA students. As discussed in Chapter 1, Kinginger (2009, 2012, 2017) has repeatedly called for more attention to the ways in which host member perspectives shape the affordances and interactions in which SA students engage.  This study contributes to addressing these gaps in a number of ways. By investigating SA interaction in an English-speaking context with Japanese L1 students, this research provides a much-needed non-US perspective on LS in SA and contributes to the growing body of literature on SA in English-speaking contexts (e.g., Jackson, 2008). Secondly, by investigating JC peers’ interactions with a range of multilingual peers, it contributes to expanding the notion of “host community” beyond conceptions of the native speaker and seeks to complicate the notion of hosts as a monolingual homogenous group. Finally, by systematically attending to the perspectives of PWU peers through analyses of interview data as well as primary interaction data, this study seeks to more clearly highlight that LS in SA is a multi-directional, collaborative, and negotiated process.   3.5 Focal aspects of peer language socialization in this study In this dissertation, I am interested in peer language socialization at PWU and specifically the norms that are indexed, reproduced, and resisted by JC and PWU peers in talk-about-55  language. In her seminal paper on LS, Ochs (1996) explains “in all societies, members have tacit understandings of norms, preferences, and expectations concerning how situational dimensions such as time, space, affective stance, epistemic stance, social identity, social acts, and social activities cluster together” (p. 417). According to Ochs, to gain insight into how particular instances of interaction, such as talk-about-language sequences, unfold, LS researchers should take into account these multiple dimensions in order to shed light on the socio-cultural meanings that are indexed in talk and the value those meanings carry within the community. Duff (2007) also highlights the importance of triangulating perspectives and insights obtained through a variety of data sources at both the macro-level and the micro-level, which she maintains are crucial if scholars are to move beyond simple description toward explanation.  In this study, in order to more clearly conceptualize the relationships between macro-level social processes and micro-level practices in peer language socialization, I turn to the recent transdisciplinary framework developed by Douglas Fir Group (2016). In their framework, they identified three interconnected levels of social processes that shape additional language learning: the micro level, which refers to specific instances of social activity and the cognitive and semiotic resources mobilized to participate in that activity; the meso level, which refers to the institutions, communities, organizations, and their associated identities which can “provide or restrict access to particular types of social experiences” (p. 24); and finally, the macro level, which refers to “large-scale society-wide ideological structures with particular orientations toward language use and language learning” (p. 24). The framework, which was intended as a broad and transdisciplinary representation of second language acquisition, fits nicely with the assumptions and findings of LS research, which emphasize mutually constitutive relationships between language use (i.e., social practice at the micro level), communities (the meso level) and 56  culture more broadly speaking (the macro level). I have reproduced a simplified version of these mutually informing and embedded aspects of language learning in Figure 3.1.  Figure 3.1 Multifaceted nature of language learning and teaching. Adapted from Douglas Fir Group (2016, p. 25) To gain insight into how talk-about-language was initiated and managed in interaction between peers, I have chosen to focus on one dimension at each of the three levels. At the macro level, I attend to how language ideologies mediated the ways in which JC and PWU peers assigned value to different peers, to peer interaction, and to English language learning; at the meso level, I analyze the expert and novice identity categories that were made relevant when the participants and I discussed language learning through peer interaction; and at the micro level, I investigate how participants engaged in stancetaking (particularly epistemic and deontic stances) that made those expert and novice categories relevant in talk. I discuss and define each of these dimensions in the following sections. As the descriptions will show, across all three dimensions, 57  I use theory and approaches that represent a discursive constructionist epistemology, which views interaction (interview and naturalistic) as a privileged location for observing the ways in which semiotic resources operate to make stances, categories, and ideologies relevant in talk. In this sense, I have deliberately selected compatible and mutually informing approaches that represent my understanding of interaction as situated and co-constructed social practice.       3.5.1 Language ideologies  The term language ideologies has been defined as “representations, whether explicit or implicit, that construe the intersection of language and human beings in a social world” (Woolard, 1998, p. 3). Ideologies mediate the process of meaning-making by serving as recognizable rationales, or interpretative frameworks, for the valuation of others’ or one’s own actions (Gal, 1998; Silverstein, 1992). In other words, language ideologies are circulating systems of value related to language and language use that are used by individuals or groups to frame their understanding of social action and have long been a focus of LS research (Garrett & Baquedano-López, 2002). While language ideologies are used by individuals in situated ways in talk, they are not the property of individuals in the same way that “beliefs” for example are usually theorized. Verschueren (2012) aptly describes this distinction in his volume on investigating ideologies in research:  Ideas, beliefs and opinions in as such do not make ideology. Simplifying a bit, they are merely 'contents of thinking,' whereas ideology is associated with underlying patterns of meaning, frames of interpretation, world views, or forms of everyday thinking and explanation. Thus the ways in which beliefs, ideas or opinions are discursively used, i.e., their forms of expression as well as the rhetorical purposes they serve, are just as 58  important for ideology as the contents of thinking for which these three terms serve as labels. (p. 7) Mertz (1998) explains that language ideologies “may be very explicit as when speakers overtly discuss aspects of language use” (p.151), as my participants did in this study. Alternatively, they “may appear more subtly, for example as a set of meta-level structural linguistic features indicating what kind of speech is occurring (or ought to occur)” (p.151).  Given that language ideologies serve as sense-making resources in interaction, it stands to reason that they play an important role in LS. Mori (2014) explains that in interaction, “language ideologies can serve to undermine, challenge, or support particular political alignments and social identities” (p. 154). In their review of LS research on language ideologies, Garrett and Baquedano-López (2002) have also observed that language ideologies “intersect in complex and interesting ways with local notions of cultural and group identity, nationhood, personhood, childhood, and language acquisition as a developmental process” (p. 354). Of particular interest for this study, they note that language ideologies are especially salient in multilingual contexts, where differing notions of what language is, as well as the relationship between individuals, culture, and language may come to the fore. These authors explain that in these contexts and others, ideologies are multiple, dynamic, and partial in nature and can thus be drawn upon and locally produced in contradictory ways. According to Miller (2009), it is by investigating how individuals position themselves and others “unproblematically with respect to language identities, without needing to provide accounts, explanations, or justifications (signs of interactional trouble) in ongoing talk” (p. 324) that we can investigate how dominant language ideologies are perpetuated and transformed as natural shared beliefs. 59  There are two categories of language ideologies that are of most relevance for this study. The first category is related to ideologies around the value assigned to specific languages, and in the case of this study, English. There is a small but growing body of SA research that focuses on ideologies related to language codes. For instance, several researchers have investigated ideologies of Japanese uniqueness, or nihonjinron, and the ways in which those ideologies mediated SA students’ experiences (e.g., Iino, 1996, 2006; Siegal, 1996). These studies found that in host family talk, ideologies of nihonjinron served as resources for evaluating SA students’ efforts to learn English. Specifically, hosts and community members espoused beliefs that non-Japanese people were unlikely to master the complexities of the Japanese language, expressed low expectations of SA students’ levels, and avoided providing feedback on features such as the use of honorifics, which they viewed as particularly difficult to master. De Costa’s (2011) and Park and Bae’s (2009) work with SA students at high schools and elementary schools in Singapore also examined ideologies related to specific linguistic codes. De Costa’s (2011) work in particular highlighted ideologies around the global value of English and the ways in which the concept of English as a tool for international communication and mobility mediated students’ descriptions of their future careers and contributed to the development of students’ cosmopolitan identities. As I demonstrate in Chapter 5, in this study the value of English as a language associated with mobility and multiculturality served as a key discursive resource in participants’ descriptions of their peer relationships.   The second category of ideologies relevant to this study is shared understandings about who language learners are, the kind of people that interact with them, and how language learning can best occur through those interactions. To my knowledge there is no specific term to refer to this type of ideology, although Riley (2012) coined the term language acquisition ideologies in 60  reference to work on beliefs around first language acquisition. The last fifteen years especially have seen an increase in work exploring how language ideologies shape the interactions of additional-language learners and teachers, particularly in English language classrooms (Mökkönen, 2012; Mori, 2014; Olivo, 2003; Razfar, 2005; Razfar & Rumenapp, 2012). These studies often examine teachers’ and students’ beliefs and attitudes toward the use of multiple languages in English language classrooms and how classroom practices (e.g., gossiping, disciplining, etc.) as well as classroom policy (e.g., English-only in the class) reproduce, resist, or transform those beliefs. Doerr’s work (2012, 2013b, 2015) on discourses that value experiential learning and immersion in SA is another example of this type of language ideology. As discussed in Chapter 2, Doerr addresses the ways in which ideologies of immersion were produced in guidebooks and interviews and the consequences of those productions for stakeholders’ understandings of host, home, and other in SA. For this dissertation, I explore the ways in which participants produced explicitly ideological descriptions of the value of English when describing their language learning goals in interviews. I then examine how those language ideologies were reproduced in locally situated and contingent ways as they described their expectations about how language learning can or should happen between peers.  3.5.2 Expert and novice identity categories Studies of social identities have long been a cornerstone in LS research given that “participation in socializing interactions fundamentally implicates identity, as individuals accommodate, resist, subvert, and or transform the acts, stances, and activities that constitute particular social identities” (Duff & Talmy, p.108). In recent years, the amount of SA research focusing on identity has grown exponentially based on the notion that identity development is a key outcome for all SA (i.e., not just sojourns involving language learning, for a review, see 61  Kinginger, 2013). In general, this research has privileged narrative (e.g., Benson, Barkhuizen, Bodycott, & Brown, 2013), sociocultural (e.g., Jackson, 2008) and poststructuralist (e.g., Anya, 2017) approaches to identity. These studies often yield insights into the personal development of a small number of SA students during their sojourns or seek to explain the power of institutions in shaping the range of identities available to sojourners (Block, 2007). In this study, the focus is less on JC peers’ identity development over the course of the sojourn (although it is discussed briefly), rather the main focus is on understanding the types of identities, or social categories, JC and PWU peers were able to legitimately take up in their moment-to-moment interactions.  For this reason, I view identity through the lens of membership categorization analysis (MCA) – a relatively novel approach to identity in SA.17 Unlike structural or psychosocial approaches to identity, the MCA notion of identity, which has been described as identities in practice or identity in interaction, views it as something one does rather than as something one is (Antaki & Widdicombe, 1998). Identity categories are understood as both locally occasioned and consequential in talk. As an ethnomethodological approach, MCA also takes a profoundly emic approach to identity: identity categories are not determined a priori but are considered first and foremost as locally-relevant members’ matters (i.e., matters that are negotiated and established in interaction by those participating in the conversation, or members). Identity in MCA is a form of social action in which categories and their attendant features are used to make meaning in situ. As such it is grounded in interaction data and provides insight into how participants use categories to engage in local sense-making practices.                                                   17 I know of only a few studies that have drawn on MCA to examine identity in SA. Wernicke-Heinrichs (2013) used it to gain insights into discursive resources for constructing teacher identity for a group of French language teachers sojourning in France. Another study used MCA to examine intercultural identities in talk between Japanese SA students and their host families (Greer et al., 2013). 62  Using MCA provides insight into the activities, characteristics, and attributes that participants treat to as usual or normative for certain identity categories, and how intersubjective understandings of those identity category constructions can produce complex meaning in talk. For example, Goodwin and Kyratzis (2012) reviewed studies that examined categorization practices in the talk of young peers. They contend that by examining practices such as gossiping, storytelling, and assessments, it is possible to shed light on how peers use evaluative stances to “locate and reference the peer group’s notion of culturally appropriate moral behaviour” and “negotiate their alignments to one another and position one another in the local social group” (p. 367). They explain that research on interaction which combines ethnographic insights with micro-analytic approaches such as MCA: permits investigation not only of how the local situated activity is organized but also how actions and stances (Du Bois 2007; Jaffe 2009) taken across a range of interactions are consequential for participants’ lives and help to construct more enduring forms of social organization. (p. 372)  Given my interest specifically in ideologies around how to “do” expert and novice in language learning between peers, it seemed appropriate to draw on insights from MCA to attend to the situated production of expert and novice identity categories (more details about doing MCA are provided in Chapter 4).  3.5.3 Stancetaking An additional aspect of peer language socialization of interest in this study is related to stance. According to Kockelman (2004), stance is “a new name for what is often called the speaker’s attitude, view, or evaluation” (original emphasis, p. 130). Examples of stancetaking might include when a JC peer indicates that a particular English word is unknown to them (i.e., 63  an unknowing stance), or a PWU peer evaluates a JC peer’s English use as “cute” (i.e., a positive, or perhaps condescending, affective stance). According to Ochs (1996), linguistic resources that index stances are fundamental resources for accomplishing social activities (such as talk-about-language) and for indexing membership in identity categories (such as expert or novice). Ochs argues that investigations of stancetaking should be central to LS research because of the privileged role of stance in constructing social life. She explains that learning to perform stancetaking and interpreting the stancetaking of others is a socialized (or learned) practice – it involves making connections between linguistic and embodied stancetaking resources (e.g., lexis, grammatical markers, intonation) and their subtle social meanings. The term stance itself seems to have gained popularity after its use in Biber and Finegan’s (1988, 1989) publications comparing various grammatical and lexical stance markers across genres of written and spoken English. However, evaluation (Labov & Waletsky, 1967), positioning (Harre & VanLangenhoeve, 1991), evidentiality (Chafe & Nichols, 1986), appraisal (Martin, 2000), assessment (Heritage & Raymond, 2005), and epistemic and affective stance (Biber & Finegan, 1989; Ochs, 1996) are all terms that have been used to describe interactional resources that are used to evaluate objects in talk. In this work, I draw on Du Bois’s (2007) understanding of stance as a discursive construction occasioned by previous talk and that has consequences for intersubjective alignment and affiliation. Du Bois defines stancetaking as: a public act by a social actor; achieved dialogically through overt communicative means (language, gesture and other symbolic forms), through which social actors simultaneously evaluate objects, position subjects (themselves and other), and align with other subjects, with respect to any salient dimension of the sociocultural field (p. 163).  64  What sets Du Bois’ discursive approach to stance apart from most of the other conceptions of evaluation cited above is his attention to the situated co-production of stances. By theorizing stance as a discursive construct, researchers more fully acknowledge the collaborative nature of interaction, which fits nicely with the assumptions of LS. This approach to stance is also compatible with the notion that stances can become associated with particular identity categories or ideologies. Bucholtz and Hall (2005) maintain that over time, more durable macro-identities may emerge through repeated patterns in stancetaking which connect particular types of evaluations (e.g., correcting the English of others) to a particular identity category (e.g., expert English speaker).18 They assert that “the interactional positions that social actors briefly occupy and then abandon as they respond to the contingencies of unfolding discourse may accumulate ideological associations with both large-scale and local categories” (p. 591).  For this dissertation, I explore the range of resources used by peers and SA students to accomplish stancetaking in a specific discursive context: instances in which peers or SA students evaluate an “object” related to language (e.g., lexis, pronunciation, form, rules of use). Given that I am interested in how peers “do” novice and expert, two categories of stancetaking are most relevant to my study. The first, epistemic stance, refers to stancetaking which makes relevant more- or less-knowing subject positions and has been researched and discussed extensively (see the recent 2018 special issue of Discourse Studies edited by Heritage). The second, deontic stance, refers to stancetaking which indexes rights or authority to control the flow of talk and is                                                  18According to Bucholtz and Hall, Du Bois has also spoken of accumulated associations between stances and positions under the term “stance accretion”. However, it appears that this was only ever discussed in a particular conference presentation delivered in 2002 and I was unfortunately unable to locate it.   65  relatively new to work in second language interaction. I provide a brief description of each below (these are discussed further in Chapters 7 and 8). Epistemic stance  Epistemic stance concerns the ways in which members orient to and evaluate knowledge claims and it is thus directly implicated in the ways in which expert and novice roles are negotiated in interaction. Ochs (1996) has defined epistemic stance as stancetaking which displays “knowledge or belief vis-a-vis some focus of concern, including degrees of certainty of knowledge, degrees of commitment to truth of propositions, and sources of knowledge, among other epistemic qualities” (p. 410). Investigations of the interactional consequences of epistemic stance examine how interactants use linguistic and other embodied resources to claim or reject authority over knowledge (such as knowledge of the English language) and in turn, how these stances serve to establish their epistemic status relative to those of other interactants (Heritage, 2011, 2013; Heritage & Raymond, 2005; Stivers, Mondada, & Steensig, 2011).  Researchers investigating epistemic stance in educational contexts are interested in the ways interaction constructs students as “unknowing” and teachers or more advanced peers as “knowing” (Jakonen & Morton, 2015; Kirkham, 2011; Koole, 2012; Koshik, 2003; Melander, 2012; Sert, 2013; Sert & Walsh, 2013). Jakonen and Morton (2013) investigated peer group work in a primary school in Sweden, showing how epistemic search sequences (i.e., questions) were resolved through sequences of stancetaking by peers, focusing on the ways in which such interactional work could be affiliative or disaffiliative. Findings were then contextualized with reference to Swedish school practices in which group work was highly valued. For the present study, I am interested in if and how peers claim epistemic authority over English knowledge and the kinds of stancetaking that would be associated with those claims.  66 Deontic stance Deontic stance is a relatively new concept in discursive approaches to stance. The deontic aspects of interaction refer to members’ “entitlements to impose actions on their co-participants” (Stevanovic & Peräkylä, 2014). Deontic stancetaking often occurs in tandem with stances in which participants display knowledge (i.e., knowing epistemic stances). For example, in a prototypical hierarchical institutional relationship between members of the language teacher and student categories, the teacher asserts power by exercising his or her deontic rights to initiate corrections in interactions (e.g., Kurhila, 2005). The teacher therefore expresses both a knowing epistemic stance as well as a deontic stance: the teacher has the language knowledge as well as the right to impose the action of correction. However, it is not always the case that category members have both epistemic and deontic authority.  While it is relatively commonsense to assume an asymmetrical distribution of deontic rights between teachers and students, it is much less evident how deontic rights and responsibilities might be distributed between members of more symmetrical categories, such as peers or friends. Indeed, research has shown that peers often engage in complex stancetaking practices in order to mitigate face threats and avoid overt displays of authority (Jenks, 2013; Kurhila, 2005; Tsai & Kinginger, 2014). For instance, Kurhila (2005) found that NS peers often pass on opportunities to correct the language of less proficient users. In online interactions between more and less proficient TL speaking peers, Jenks (2013) and Tsai and Kinginger (2014) found that peers often produced compliments to minimize asymmetries and attend to the face needs of peers. In addition, as I reviewed the data for this study, I noticed that PWU peers often produced lengthy rationales and explanations when describing instances in which they engaged in talk-about-language and that they often explicitly referred to deontic rights around 67  who should or should not initiate talk-about-language. For this reason, both epistemic and deontic rights became focal aspects of this study.  3.5.4 Summary In sum, the three focal dimensions of LS investigated in this study are language ideologies, expert and novice identity categories, and epistemic and deontic stancetaking in talk-about-language. As the Douglas Fir Group (2016) model discussed at the beginning of this section shows, these dimensions are not separate but are interconnected and produced, reproduced, resisted, and transformed for local purposes in specific instances of interaction. For each of these dimensions, I have described my specific understanding of the construct, in each case adopting an emic discursive approach which views these dimensions as dynamic, contingent, and locally occasioned.  3.6 Conclusion This chapter has provided an overview of LS as this study’s lens for understanding language, learning, and development. It has pointed out that, from an LS perspective, learning is a multidirectional and collaborative process in which all participants play an active role. I have also provided an overview of the ways in which LS has been integrated into research in SA, the insights that have been gleaned from that research, and the gaps that remain. In particular, I pointed to the need for LS research in the SA context to include a broader range of interactions and contexts and located the contribution of this study within that gap. Finally, I described my discursive constructionist approach to the dimensions of LS that form the focus of this study and highlighted how attention to each provides insights into peer language socialization in this context.  In the next chapter, I introduce the study design and data generation procedures.  68  Chapter 4: Methods 4.1 Introduction Case study allows researchers to reveal new perspectives on processes, entities, phenomena, or experiences. Dyson and Genishi (2005) explain that “[i]t is the messy complexity of human experiences that leads researchers to case studies in the qualitative and interpretive tradition” (p. 3). In Chapter 3, I showed through a review of the literature that peer relationships and their role in TL development are the types of messy, unpredictable phenomena to which these authors refer. I also pointed out that SA researchers have advocated greater attention to the processes that mediate SA students’ engagement with peers. Therefore, for this research, I have opted for a multiple case study design in order to gain a more nuanced understanding of how talk-about-language can be accomplished between JC and PWU peers.   4.2 Case study According to Miles and Huberman (1994), a case is “a phenomenon of some sort occurring in a bounded context” (p. 25) – it is both the method of research as well as the final product. Case studies are characterized by “boundedness or singularity, in-depth study, multiple perspectives or triangulation, particularity, contextualization, and interpretation” (Duff, 2008a, p. 23). The qualitative multiple case study reported here is what has been called an instrumental case (Stake, 2000). Its objective is to explore, describe and/or explain a phenomenon rather than focus on the biographical particularities of any one participant. In this case study report, the phenomenon of interest is peer language socialization in SA, and more specifically, peer LS related to talk-about-language. A key aspect of case study is its typical use of different forms of data and perspectives to generate and discuss findings. For this study, I examine peer LS via 1) 69  JC peers’ reports of their experiences with PWU peers over the course of a semester, 2) an investigation of the discourses surrounding how JC and PWU peers should (and should not) contribute to language learning, as reported in interviews, and 3) a close analysis of recorded conversations which include talk-about-language.  4.3 Research context This study was undertaken at a highly internationalized Canadian university I have called Pacific Western University (PWU). The research focuses on the out-of-class experiences and interactions of three Japanese undergraduate students participating in the Japan-Canada (JC) program – an SA program. In what follows, I describe PWU as well as the objectives and features of the JC program.  4.3.1 PWU  PWU is a large western Canadian university with over 50,000 undergraduate and graduate students, about one fifth of whom are classified as international students.19  As such, PWU has been ranked as one of the most international institutions in North America (according to the Times Higher Education Rankings). In addition to students registered as “international students,” a large number of students that attend PWU are generation 1.5 (Rumbaut & Ima, 1988), meaning they immigrated to Canada at a young age and completed their secondary school education in Canada. The university prides itself on being a diverse and inclusive institution and has a well-funded diversity initiative program. In addition, nearly 25% of the student-led clubs have a specific ethnic, linguistic, racial, or cultural affiliation (e.g., Russian Club). Courses at                                                  19 I have avoided providing a source or exact statistic to maintain anonymity. The information about PWU was gathered through publicly available sources such as the university website. 70  PWU are generally delivered in English, which is an official language of Canada and the dominant language of the province in which the university is located. The campus is outside of the downtown area and exists as its own university town with grocery stores, restaurants, recreation facilities, and leisure activities all accessible directly on-site. There are over half a dozen large student residences where housing is guaranteed for most exchange students as well as first-year students. 4.3.2 JC program Each year, between 70 and 100 undergraduate students journey from Japan to PWU as a cohort to attend the full-time program for two semesters (September to December, and January to April). To enter the program, which is designed especially for JC students and delivered in English, students must obtain a minimum score of 450 on the TOEFL ITP (Test of English as a Foreign language, Institutional Testing Program version). This score is substantially lower than the score required to attend courses for mainstream PWU students (around 570).20 Participants in this study reported that the lower English requirement was one reason that JC students chose this program over the many other programs offered by other partner universities, which required higher scores.  At the time of the study, JC program operations were centralized in one location on the PWU campus, JC House (pseudonym), which contained the administrative offices, computer lab, dormitories, and classrooms. Within that building, JC students had access to a staff person dedicated exclusively to the JC program as well as a Japanese-speaking program assistant to help                                                  20 The TOEFL ITP is not accepted by PWU because it includes only reading and listening components. Therefore, the number provided here is a rough conversion based on the score that PWU accepts for the internet-based version of the TOEFL. PWU accepts students scoring 90 on the TOEFL iBT. Comparatively, JC participants recruited for this study who had taken the TOEFL iBT reported scoring between 65 and 80.  71  them with questions about courses, housing, health care, and any other concerns they might have. Before the start of the study, I had been involved in the JC program in various capacities for three years. 4.3.3 JC courses The English-medium courses in the JC program are loosely based on social, critical, and functional approaches to language education, including attention to genre and discipline-specific language. During the year in which this research was conducted, classes were held in several buildings that were also used by mainstream PWU students. JC students attended classes for approximately 18 hours per week on topics related to sociolinguistics, culture, geography, and academic writing. Most courses contained 20 students or fewer. JC students who obtained excellent grades in JC courses as well as high TOEFL scores in the first semester of their sojourn were permitted to take one or two mainstream PWU courses (e.g., psychology, sociology, linguistics) in the second semester of their sojourn (about 30% of the cohort).  The JC courses were designed to foster critical thinking about cultural and linguistic difference and touch on topics such as World Englishes, immigration, and the intersection of language and race/gender. As such, JC students discussed issues such native speaker norms, multiculturalism, and stereotypes. Courses did not include explicit grammar teaching or target TOEFL preparation. The JC program structure, therefore, differs from SA programs offered by private language schools or university language centres, which typically provide intensive language courses with an emphasis on native-like use. It also differs from Erasmus or bilateral exchanges in which individual students take courses with domestic students. Finally, it is in stark contrast with the faculty-led short-term summer programs that are increasingly popular in the United States and which are often developed in tandem with a private study tour provider.  72  4.3.4 JC students The students enrolled in the JC program were Japanese undergraduates with various majors in the social sciences and humanities (e.g., sociology, law, international relations) who ranged in age from 19 to 21 and in most cases, had just completed their second year of study in Japan. Almost all had Japanese as their L1 and some had taken a few courses in a language other than English (e.g., Spanish, Mandarin, German). Students’ average English proficiency could be characterized as intermediate, with the bulk of students obtaining scores between 460-510 on the TOEFL IPT.   Among their reasons for choosing this program, JC peer participants in this study cited the length of stay (two semesters instead of one), the slightly lower English proficiency requirements, and their impressions of Canada as a peaceful and multicultural country. They also reported that many JC students had contacted program alumni about their experiences and so they had a reasonably clear idea of what the program would involve when they arrived. For the JC peer participants, the JC program (one academic year) was the longest time they had spent away from home, although they also reported that many of the other JC students had undertaken shorter sojourns in English-speaking contexts with their high schools or summer programs. JC peer participants also explained that some students received supplementary funding from their universities in the form of a bursary to cover travel and living costs, but that most would have to reimburse the funding following graduation.  4.3.5 On-campus residences In 2015, JC students lived in several different on-campus residences where the kitchen, bathroom(s), and living room were shared with three to five other housemates in apartment-style units. JC students were assigned to units with students from other programs (domestic, 73  international, and other exchange students) and were not allowed to share a unit with another JC student. As is the case for most roommates who are randomly assigned at universities, JC peer participants reported that the amount of interaction JC students had with their roommates varied widely – some formed close relationships with their roommates while others did not.  It is important to note that the three residences in which JC students were placed tend to house a high proportion of international and exchange students, often resulting in very culturally, ethnically, and linguistically diverse units. Many JC students also had to welcome new roommates part way through their sojourns. For example, Lisa, one of the JC peer participants in this study, lived in the first semester with four roommates: a Korean exchange student, a Canadian senior undergraduate, an international roommate from Jordan, and a Chinese-speaking generation 1.5 roommate.  In the second semester, Lisa’s Korean and Canadian roommates left the flat (one graduated and one completed her exchange) and two first-year students, an international student from Singapore and a Malayalam-speaking generation 1.5 student, arrived to take their places. From discussions with JC students, I learned that Lisa’s situation was relatively typical.  4.3.6 Language exchanges and other clubs The JC program encouraged students to participate in PWU student-run clubs, which centred around interests such as dance, tea, or food tours. However, JC peer participants generally reported that the clubs were not very active. Instead, JC students turned primarily to free language exchange programs to meet PWU peers: in 2015-2016, over two thirds of the JC cohort participated in some form of language exchange (i.e., trading Japanese language practice 74  for English practice).21 The largest language exchange program, which I will call Pair Talk, was a campus-wide program in which students were paired randomly based on the languages they wished to exchange. Pairs could schedule a time to meet independently for an hour each week or meet at an appointed time in a common lounge, where they would receive guidance on topics from facilitators. Pairs were to spend 30 minutes speaking each of the languages that the participants wished to exchange (i.e., 60 minutes total). In this study, all JC peer participants had a Pair Talk partner.  Another language exchange, which I will call Japanese-English Exchange Group, was targeted specifically to students wishing to exchange English and Japanese and was advertised in the JC program and the Japanese language programs at PWU. As the name implies, the Japanese-English Exchange Group was an informal small group conversation held at an appointed time once or twice a week. Like Pair Talk, 30-45 minutes of the chat were held in English and the other 30-45 minutes took place in Japanese. All JC peer participants attended the Japanese-English Exchange Group at least a few times during the sojourn. JC peers reported that for both language exchanges, participants did not always have the necessary language levels to conduct entire conversations in the TL (e.g., PWU students’ level of Japanese was sometimes too low). In these cases, the participants explained that JC students would teach Japanese words through the medium of English.  Finally, in the second semester, the JC program began offering an informal English class specifically for JC students, which I will call English Conversation Club. The class, which offered cookies and coffee, was facilitated by TESOL professionals in training. Between 15-20                                                  21 This estimate was provided by JC staff. 75  JC students attended regularly, including all the JC peer participants in this study. Members of the English Conversation Club discussed questions related to topics such as happiness and intercultural marriage and received targeted feedback from their teachers-in-training. This English-only group was not open to mainstream PWU students.  4.4 Participants  For this multiple case study, I recruited four JC peers who were willing to participate in interviews about their language learning experiences and record casual conversations with consenting PWU peers (usually their roommates or language exchange partners). Ultimately, I selected three focal JC peers upon which to base this dissertation (selection rationale is provided in section 4.6). 4.4.1 Recruitment  Prior to recruitment, I discussed my intention to investigate out-of-class interactions with the program director, who gave me permission to do so and greatly facilitated the recruitment process.22 I also opted out of teaching during the semester while I conducted the study to reduce any real or perceived conflict of interest (i.e., with students possibly feeling that they needed to participate because I was their teacher). By not teaching, I also hoped to be able to nurture friendly relationships with my JC participants, who ultimately shared many private moments and experiences with me. After obtaining ethics approval from the university review board, I recruited JC peers primarily with a flyer that was distributed at the welcome back orientation as well as through the program listserv and by some instructors in the first week of class (see                                                  22 While the director and instructors were aware that I was conducting research with students from the program, the identities of the individual students have been kept confidential and during the year the students were in the program, no data or findings were shared.  76  Appendix A). The flyer indicated that I was interested in learning more about JC students’ interactions in English outside of the classroom. Once the JC participants had consented, they recruited PWU peers with whom they had regular contact in English. The choice of who to recruit was left up to the JC participants in order to investigate features of relationships that had formed naturally rather than those that had been designed for the purposes of research. As I describe below, however, sometimes the best laid plans do not always turn out as expected. Given the qualitative nature of this study, I do not view these unforeseen turns as hindrances, rather they show that there was something unexpected to be discovered in this context.   4.4.2 JC peers Recruitment for the study began in January 2016, at the start of the second semester of the program. The second semester was selected in the hope that students would have developed relationships with English-speaking PWU peers by the time recruitment began. Throughout January, eight students emailed or approached me, and I explained the study to them in detail either in person or via email and shared the consent form. Ultimately four JC peers, Ami, Lisa, Blue, and Samantha (pseudonyms chosen by the participants), consented to participate in exchange for two $15 gift cards of their choice, given after the first and last interview respectively, and for opportunities to meet with me and practice English.23 In many ways, Ami, Lisa, Samantha, and Blue (three females and one male) were typical JC students: they were all 20 years old, they were all ethnically Japanese and spoke Japanese as their L1, and they all had previously had one or more shorter SA experiences.24 When they                                                  23 I met with each JC participant several times for informal chats, and once took them to a local tourist attraction.   24 When I began working in this program in 2011, it was not the case that most students had previous SA experience.  The proportion of students with prior experience seemingly has increased each year to the point where now it is 77  contacted me about the study, they explained that they were interested in the project because it was just one more way in which they felt they could accomplish their twin goals of making friends and improving their English. Unfortunately, none of the four had developed a strong friendship with any English speaker in the first semester, despite desiring and actively trying to do so. As is relatively common in SA contexts (Pyper & Slagter, 2015), they expressed dissatisfaction with their opportunities for English interaction. They told me they felt this project would encourage them to speak English more regularly. All the JC peer participants knew each other and were aware of each other’s participation in the study. Table 4.1 summarizes each of the JC participants’ characteristics.   Table 4.1 Summary of JC Participants Pseudonym Gender TOEFL iTP Major in Japan Previous SA experience Ami  Female 517 Political science High school trip to Philippines Lisa  Female 463 Sociology High school trip to Canada Samantha  Female 540 International relations Homestay in the US Blue Male 510 Business Summer school in the US, language school in Germany                                                   commonplace. This is likely a product of Japan’s aggressive internationalization strategy and its intensification of English training at the elementary and high school levels (Mock et al., 2016).  78  4.4.3 PWU peers Once I had recruited the JC peers, I asked each of them to recruit one or more PWU peers who would be willing to come in for interviews and allow the JC peer to record their informal conversations (see follow-up emails in Appendix A). At the outset of the study, I defined PWU peers loosely as people of similar age to JC peers and with whom JC peers interacted regularly in English in informal contexts (i.e., outside the classroom). Based on these broad criteria, JC peers were free to recruit whichever PWU peers they chose. PWU peer recruitment was ongoing through the data generation period, with the last peer being recruited in March. In all cases but one, the PWU peers were more proficient in English than the JC peers and had met the language requirements for attending mainstream courses at PWU.25 The exception, Emma, was a fellow JC student recruited by Ami. Ami argued that she and Emma had been using English to complete group work in a JC course in the first semester and that they wanted to continue using English informally in the second semester. She felt the study would be a good reason for them to keep in regular contact. As the analyses in Chapter 8 will show, Emma’s data reveal different patterns related to talk-about-language and thus provides a useful means of contextualizing findings.  Table 4.2 provides the pseudonyms for each of the PWU peers, how they met the JC peer, their country of origin, and the languages they spoke.                                                     25 I did not test PWU peers’ English proficiency. However, from interviews, in my opinion as both a proficient speaker and seasoned English teacher, most PWU peers spoke English expertly, often in native-like ways. 79  Table 4.2 Summary of JC Peer Participants  Focal JC Peer PWU Peer Relationship Country of origin Languages Ami So-Yi Roommate Korea Korean (L1) English (L2)  Paulisper Friend Ontario, Canada English (L1) Singhalese (L1) Tamil (L1)  French (L2) Emma26 Cohort-mate Japan Japanese (L1) English (L2)  French (L2) Lisa Serena Roommmate Dubai, immigrated to Canada age 10 Malayalam (L1) English (L1)  Ed Roommate Bangladesh/ Philippines/ Singapore English (L1) Tagalog (L1) Bengali (L1) Spanish (L2)  Hindi (L2) Kyla Roommate China, immigrated to Canada age 8 English (L1) Mandarin (L1)  Samantha Elizabeth Roommate United Kingdom English (L1)  Arabic (L2)  John Paired language exchange partner China Mandarin (L1) English (L2) Japanese (L2) Blue Bob Group language exchange Hong Kong, immigrated to Canada age 5 English (L1) Cantonese (L1) Japanese (L2)                                                   26 As indicated in the paragraph above, as Ami’s cohort-mate and a fellow Japanese speaker, Emma was different in many ways from the other PWU peers. 80  Once they had selected a peer, the JC peers provided PWU peers with basic information about the study and my contact details. I then shared further details and the consent form only if the PWU peer contacted me or consented to be contacted. PWU peers also received two $15 gifts cards – one after the first interview, and one following the second interview. This strategy led to the recruitment of peers who were relatively easy to talk to, willing to help when asked, and, as Table 4.2 indicates, to some extent multilingual. However, different peers had different levels of ability in different languages. Paulisper and Elizabeth, for instance, reported having only basic knowledge of languages other than English, while Kyla, Ed, Serena, and Bob reported being proficient speakers in all their languages.  4.5 Data generation In this study, I opt to use data generation (rather than data collection) to refer to the processes by which I co-created conversation and interview data with my participants. In selecting this term, I follow other qualitative researchers (e.g., Roulston, 2010; Talmy & Richards, 2011) who seek to make evident their role in shaping their research data. Accordingly, as I describe each type of data and the circumstances of its generation, I will also refer to my own role in its genesis. Despite my use of methods traditionally associated with ethnography such as interviews and naturalistic recordings, this study is not an ethnography. Rather it is a multiple case study that adopts some ethnographic methods and makes use of additional ethnographic details where appropriate. The main data for this study consist of audio recordings made by JC peers, which I refer to as remote observation data, and formal research interviews with both JC and PWU peers. In total, the study comprises approximately 48 hours of recorded 81  data. Additional data include field notes from informal discussions throughout the generation period, as well as focus group data generated in April 2016. 4.5.1 Interviews Interviews are an ideal context for gathering information on how participants experience and perceive their social worlds and report on it to others (Roulston, 2010). I conducted and audio-recorded semi-structured interviews with JC peers three times, at the beginning, middle, and end of the 14 weeks, and with PWU peers twice, once when they were recruited and again at the end of the 14 weeks (exceptions are So-Yi, John, and Elizabeth, who completed only one interview; see participants’ descriptions in Table 4.2). While I often took up the role of researcher with the participants, at various junctures other identity categories, such as native English speaker, local (I am from the area around PWU), and, in the case of the JC peers, as (former) instructor, were also oriented to by the participants in our conversations.  To account for the complex ways that I participated in explicitly and implicitly guiding data generation for this study, I adopt a constructionist approach to analyzing interview data (Talmy & Richards, 2011). This approach treats interview accounts not as objective truth, but rather as rationalizations and retellings negotiated jointly by the interviewer and interviewee. According to Roulston (2010), in the constructionist approach, “interview data provides situated accountings on research topics – that is, particular versions of affairs produced by particular interlocutors on specific occasions” (p. 17).27 Thus, I do not treat interviewees’ rationales and                                                  27 From an MCA or CA perspective, accounts or accounting practices are stretches of talk in which members describe or explain the properties of a given situation. For example, if a participant is asked to choose a restaurant, the participant is likely to name the restaurant and then also describe it or explain its appeal. The additional explanations are accounts. According to Sacks (1992), analyses of participants’ accounts provide insight into the organization of members’ practical reasoning and into the foundations of local social organization.  82  explanations as reflective of their interior selves, but as drawing on shared resources to craft social, rational, and moral messages (Rapley, 2012; Talmy, 2010). Thus, for this study, the interview was used as a tool for gaining insight into how JC peers, PWU peers, and I (the researcher) co-constructed their language learning histories and goals, and crucially, peer relationships and interactions.  I conducted all interviews one-on-one, on campus and in person, usually in a quiet public space in the morning with a cup of coffee (see interview protocol in Appendix B). The interviews took place in English both out of necessity (I do not speak Japanese) and by choice: using English ensured all the data for the study were in the same language, which facilitated comparisons between instances in which JC and PWU peers spoke about their interactions, relationships, and goals, and how they co-constructed those ideas in interaction with each other (Razfar & Rumenapp, 2012). JC peers also viewed these interviews as an additional opportunity to practice English (a major reason for them undertaking the project in the first place). The longest interview took about an hour and a half, and the shortest took just twenty-five minutes.  During second interviews, again individually with JC and PWU peers, I also conducted a playback session in which participants listened to two or three talk-about-language clips from their recorded conversations and reported their impressions (e.g., what they remembered about the incident, how they felt at the time, impressions of JC peers’ English use).28 Playback served two purposes. Firstly, it allowed me in some cases to clarify aspects of the recordings (e.g., who was speaking, if they were referring to a previous event) and thus served as a form of member-                                                 28 This technique is similar to what applied linguistics often refer to as stimulated recall (Gass & Mackey, 2000); however, the notion of “recall” does not fit with my discursive approach to interviews, so I have chosen the more neutral term “playback.” 83  checking. Secondly, in some cases, participants’ explanations were rich in rationalizing talk in which they described why they felt they (as JC peers) had or had not improved their English or justified their interactive choices (e.g., to ask for help or not during communication struggles). A summary of the interview data that were generated is presented in Table 4.3. Table 4.3 Summary of Interview Data Participants Total minutes Total interviews Ami 193 3 • So-Yi • Paulisper • Emma 22 105 116 1 2 2 Samantha • Elizabeth • John 155 27 71 3 1 1 Lisa • Ed • Serena • Kyla 176 76 72 84 3 2 2 2 Blue • Bob 170 101 3 2 Total interview data:  1097 (18h17) 27  4.5.2 Remote observation  Remote observation (Iino, 2006) relies on participants to select and record interactions in informal settings. While this approach is not traditionally used in LS research, which usually privileges direct participant observation, it has been popular in the SA context (Cook, 2006; Diao, 2016; Dings, 2014; Ishida, 2010; Kinginger et al., 2014; Levine, 2009; Shively, 2011; Wilkinson, 1998). For this study, each JC peer used their own recording device (phones or laptops) to video- or audio-record discussions. Participants chose times and locations convenient to them, usually on campus in social spaces or their dormitories. Most interactions were dyadic – 84  they took place between one JC peer and one PWU peer. However, Lisa often interacted with all her roommates at the same time (up to four speakers), which could sometimes prove difficult to transcribe. The JC peers were instructed to complete weekly recordings of a minimum of 15 minutes with each PWU peer they recruited and to meet with me weekly or every two weeks to submit the recordings. In practice, participants recorded much less often for many reasons: school work, vacations, mid-term breaks, and changes in the relationship. While I reminded students regularly to complete recordings, I did not insist too strongly – after all, both peer relationships and study participation were voluntary commitments that relied strongly on the goodwill of others. I did not wish to damage the emerging relationships.  There are several advantages to having participants, rather than the researcher, generate audio or video data. Firstly, it allows data to be generated in spaces where the researcher’s presence might be perceived as intrusive. By asking JC peers to complete the recordings themselves, I ensured that I did not invade their private lives unduly and that they could control which aspects of their lives to share. In addition, while the researcher can only be in one place at one time, the remote method allowed me to obtain data from multiple contexts and networks simultaneously. Finally, with the proliferation of mobile devices, filming and recording with small devices is not only a fairly simple process but it is also an everyday one (Bachmair & Pachler, 2015). Young adults in most cases have already developed the habit of carrying and consulting electronic devices (Surtees, 2013). My participants frequently used the video and audio recording software built into their phones and laptops for purposes not related to my study. Although I was not physically present when participants recorded their conversations, I was symbolically present via the electronic recording device. Participants often greeted me or said goodbye when turning the recorder on and off and they sometimes spoke about me with full 85  knowledge that I would hear their comments. In addition, as mentioned previously, many of the conversations would not have taken place if the JC peers had not asked their PWU peers to meet up for a recorded chat. Thus, the electronic device (as a proxy for me and my project) mediated the participants’ opportunities and interactions in important ways. Rather than perceiving these data as somehow “less natural,” Gordon (2013) and Speer and Hutchby (2003) both explain that researchers can gain significant analytical insights by acknowledging and attending to the ways in which participants treat the recorder. Given the ways in which the recording devices (as proxies for me) shaped the data, I conceptualize my participation in these recordings as that of an unseen over-hearer (Goffman, 1981) and in some cases (particularly for video recordings) as an audience member witnessing a performance. I also prefer the term remote observation to non-participant observation, or participant recordings as a way to acknowledge my role as a distant observer in this context.29 Figure 4.1 provides a summary of recordings for each JC peer over a 14-week period. As the figure shows, recording times varied from the minimum of 15 minutes to 86 minutes. In total, participants generated 30 recordings (approximately 13 hours). For Lisa and Blue, these recordings represent a fraction of the time spent interacting with these PWU peers. In other cases (e.g., Paulisper and Ami), the recordings represent all but three of their face-to-face interactions.  In some cases, the JC peers recorded parts of what appeared to be longer conversations. In other cases, the interaction occurred explicitly for the recording (Blue and Bob, Emma and Ami).                                                  29 Originally, I had also intended to have JC peers complete logs about their interactions (the locations, people involved, etc.); however, it quickly became clear that the participants would not complete the logs consistently or with enough detail for them to be of use. The logs were therefore abandoned. I compensated for this by discussing the recordings with them during informal meetings when I retrieved their recordings and during interviews. 86  Ami So-Yi (20:39)  So-Yi (15:16)  Emma (22:04)  So-Yi (17:38)  Paulisper (18:27) Paulisper (38:28)  Emma (24:19)                            Reading break                                                   Reading break Emma (15:01)  Emma (15:35)       Emma (18:30)  Total: 10 Duration: 206 min.               Lisa   Kyla (20:00)  Kyla & Ed (30:37) Ed (32:36)  All (49:31)  All (35:29)   Kyla & Ed  (36:20)  All (11:09) All (18:14)   Total: 8 Duration: 234 min.               Blue   Bob (33:30)  Bob (30:33) Bob (30:20)  Bob (33:20)         Total: 4 Duration: 129 min.               Samantha  Elizabeth (13:40)  Elizabeth (16:57) Elizabeth (16:22)  Elizabeth (18:27)  Elizabeth (15:05) Elizabeth (20:36)  John  (43:45)  John (86:25)   Total: 8 Duration: 231 min Figure 4.1 Remote observation timeline. The dotted lines represent time, the bolded numbers correspond to the week during which the recording was made.   1     2 3 4 5 6     7      8    9  10   11   12    13      14  1     2 3 4 5 6     7      8    9  10   11   12    13      14  1     2 3 4 5 6     7      8    9  10   11   12    13      14  1     2 3 4 5 6     7      8    9  10   11   12    13      14  87  4.5.3 Other data In addition to the interview data, several other types of data were used for contextualization purposes. I gathered public documents concerning the JC program both on the PWU and the Japanese university websites. Internationalization policy documents and local news articles related to internationalization at PWU and in Canada more broadly were also gathered. I took field notes on the dates and topics addressed in informal meetings with participants, my impressions of their language development, and the circumstances of our interviews. Finally, in the 14th week, I conducted a 90-minute focus group with two PWU peers (Bob and John) and all four JC peers about what constitutes a successful SA sojourn, SA goals, SA strategies, and SA challenges. These data served to contextualize my findings.     4.6 Focal cases  Based on the data generated, I selected three focal cases for the purposes of this dissertation: Ami and her peers, Samantha and her peers, and Lisa and her peers. Although Blue and his peer interlocutor Bob present an interesting case, it will not be discussed in this report for several reasons. Firstly, Blue made fewer recordings of talk than the others. Secondly, Blue’s recorded conversations tended to be principally pedagogical in nature and I came to understand that they were not particularly representative of most of his conversations with Bob. Finally, although Bob described their interactions at length, Blue tended to be less open about his relationships and the nature of his interactions with Bob.  Below I describe the defining features of each of the focal cases that were ultimately selected (also refer to Table 4.2).   88  1. Ami and her peers The first case concerns Ami and her peers, each of whom had very different levels of English proficiency, including: a Canadian native English speaker (Paulisper), a Korean exchange student who was more adept in English than Ami (So-Yi), and an L1 Japanese speaker with English abilities similar to Ami’s (Emma).  2. Samantha and her peers The second case concerns Samantha, who interacted in different contexts: in her residence with her British international student roommate (Elizabeth), and in her Pair Talk language exchange with John, a first-year Chinese international student.  3. Lisa and her peers The third case involves Lisa, who most often had multiparty conversations with her roommates, all of whom were multilingual (Ed, Serena, and Kyla). This case therefore features collaborative stancetaking practices that occurred in group situations.  While I refer to these as separate cases for simplicity, these cases could also be understood as nested within the larger case of the JC program at PWU. Indeed, all the JC peers knew each other and had similar opportunities for peer interaction afforded through the programs and clubs offered at PWU. PWU peers, on the other hand, did not know each other and were all enrolled in different PWU programs. The following case map demonstrates the ways in which the JC peers’ experiences overlapped in the JC program. Circles have been drawn with dashes to indicate that these people and spaces are not autonomous entities; rather they interacted with each other. The three colours represent the three cases – blue is Samantha’s case, purple is Lisa’s case, and green is Ami’s case.    89               Figure 4.2 Focal case map  4.6.1 Case stories To get a sense of the JC peers’ relationships over their sojourn abroad, I began the analysis phase by composing what Kinginger (2008) referred to in her work as brief case histories of each JC peer’s interaction experiences and their SA goals. For this dissertation, I have chosen to present these histories as stories, rather than case reports, in order to emphasize that the details presented are not objective “facts” obtained in interviews, but rather are situated retellings of my and the participants’ interaction during those interviews. The term “story” is meant to highlight my epistemological commitment to viewing data as co-produced and to formally acknowledge my role in its production. As the ultimate storyteller, I am present in these  Lisa         Ami So-Yi          Roommate  Paulisper Acquaintance Ed, Serena, and Kyla Roommates Elizabeth Roommate  John Pair Talk  partner  Emma JC peer PWU  JC PROGRAM Samantha  90  stories via the selection of interview excerpts and the reports of my own impressions of our initial meetings. In writing these stories, I have also deliberately entextualized (Briggs, 2011) participants’ experiences in ways that highlight key themes of interest to this study.  There are several additional advantages to presenting the data in a narrative form. Firstly, it enhances readability and allows for ample coverage of each JC peer’s experiences. Secondly, the narratives are understandable for the participants themselves. As such, Ami, Samantha, and Lisa were each given an opportunity to read and contribute to these stories after they were written. The stories presented in Chapter 5 are thus mutually agreed upon texts about JC peers’ experiences during their sojourns. By writing these stories, I sought to make sense of the general goals JC peers had for their sojourns, to what extent those goals included peer relationships, and their experiences in relation to those goals.  4.7 Analysis  The tools I use for analysis are what have been termed “eclectic” (Miller, 2014); however, the common thread that unites them is close attention to “participants’ ongoing process of interpretation in conversation” and “what it is that enables them to perceive and interpret particular constellations of cues in reacting to others and pursuing their communicative ends” (Gumperz, 1982, p. 5). For this work, I found it especially helpful to draw on a set of questions used by Hinnenkamp (1991) to guide analysis in his discursive case study of intercultural communication. In his work, he uses the following questions to approach talk: “How and why is it possible that the said is sayable, that it can be said that way, that it is permissible to be said that way, and that it can be understood that way?” (p. 93). As an analyst, referring to these questions while I engaged with the data was a helpful reminder that analyses must always be grounded first 91  in what can be observed in the data themselves. The questions highlight the importance of making the familiar strange to explore how mundane conversation is accomplished. These key principles guided the selection, analysis, and transcription phases. 4.7.1 Data management and transcription Qualitative data analysis is by nature an ongoing and iterative process that involves not only the analysis of discourse but also data logging, management, and grouping. For this study, these processes began during data generation. I listened to recordings as they were submitted and flagged particularly salient talk-about-language moments, noted questions and avenues for exploration, and possible foci for analysis. Following JC peers’ return to Japan, I transcribed the interviews and remote observation data using a plain verbatim approach.30 I replaced participants’ names with the pseudonyms they chose. To facilitate the simultaneous visualization of multiple data points, I loaded all transcripts and recordings into the qualitative data analysis software, ATLAS.ti. In ATLAS.ti, I read and re-read the transcripts, anchored them to the audio files, and linked them to field notes around the circumstances of their production. This enabled me to read and listen to the data simultaneously and to easily read the associated notes and contextual details. I also consulted a Japanese speaker to translate the occasional Japanese words that surfaced in the interaction data (notably between Emma and Ami).   Once I had identified the segments of talk that were most relevant for answering the research questions, I re-transcribed them in more detail adding markers for features such as overlap, intonation, and pauses (conventions are listed in Table 4.4).                                                    30 All words and any major noises were transcribed with conventional English punctuation.  92  Table 4.4 Transcription Conventions Symbol Meaning . falling intonation  ! Exclamation ? rising intonation (not necessarily a question) [ ] overlapped speech  =    latching  charact- a restart of sharp cut-off  (0.5)       pauses measured in tenths of second (.) unmeasured pause less than 0.3 seconds oh:   sound stretch  Underline emphasis signaled by pitch or volume WOW yelling or highly exclamatory speech ° ° portions that are delivered in a quieter voice >  <         noticeably faster speech $  $         smile voice “   ”        reported speech ((laugh))    laughter, translations, and transcriber comments Shinkasen   words spoken in a language other than English ((.pt)) Lipsmack Note. All transcripts are presented in courier new font, conventions adapted from Prior (2016) As Prior (2016) has noted, transcription is a process of observation, noticing, and representation. The transcription process is, therefore, a form of analysis in its own right and an important component of my analytic approach. Re-transcribing data according to these new conventions helped me to include important analytic details in the transcripts themselves and provided me with the opportunity to listen to the audio data multiple times, to notice patterns and actions (e.g., accounting practices and repair work). These re-listenings also allowed me to catch nuances that were overlooked in the first transcription.  In the sections that follow, I describe how I approached analysis of the interview and remote observation data. 93  4.7.2 Interview interaction The analyses conducted on the interview data were principally designed to answer the first set of research questions: How do JC peers describe the contribution of peer interaction and/or relationships to their English learning? How do PWU peers describe their contributions to JC peers’ English learning? These questions were designed to focus on the discursive construction of PWU and JC peer roles and to shed light on the ideological resources used to accomplish those constructions. Silverstein (1979) has advanced the view that ideologies are most visible in metalinguistic discourse (i.e., discussions about language and language learning), or secondary rationalizations related to language practices (often obtained via interviews). In the interviews for this study, I elicited metapragmatic/metalinguistic commentary by asking PWU peers to share their impressions of JC peers’ progress as well as the ways in which they supported JC peers’ English learning. I also asked JC peers about their impressions of their learning experiences and what they learned from interacting with PWU peers. Excerpt selection  To analyze the ways in which participants used ideological resources to construct narratives and accounts of their roles in JC peers’ language learning, I first selected excerpts in which participants talked about language and language learning.31 This process involved a first pass through the data in which I tagged segments where participants (including me) oriented to topics around language learning. I tagged sections in which PWU peers were asked to evaluate JC peers’ English or JC peers were asked to evaluate their own English abilities. I also tagged                                                  31 I purposely have used the term data selection rather than reduction to foreground my role in producing the analysis.  94  sections in which participants described their learning or helping strategies, or stories in which language was a focal aspect. For example, in the following excerpt, So-Yi, Ami’s roommate, describes how Ami asks for language help.  Excerpt 4.1 “She's not ashamed”  01 02 03 04 05 06 07 08 09 10 11 12 13 14 15  Victoria: So-Yi:    Victoria: So-Yi: Victoria:  So-Yi:  Victoria: So-Yi:  Victoria:  so (.)what do you like best ab-about speaking with Ami.  mm: (0.9) she is tr- she is trying to (0.5)  I guess she is trying to speak more fluently?  and (0.6) and she asks sometimes sometimes asks me  how to do this or do that  ⸰mhm⸰ and I really like that part? of her? so yeah.  mhm. so you like- why do you like it. (0.9) because like (0.3) she asks wha- asks to me?  because she knows I guess she knows me ⸰mhm⸰= =and she’s not ashamed about that?  and I really like that part. like yeah. ah. that’s great.  (So-Yi, Interview 2, 15/01/2016, 15:45-16:44) Focus on categorization  Once the segments had been selected, I analyzed each instance individually. The analysis was guided by a general orientation to members’ methods – in other words, I was interested in examining the discursive resources that participants used to co-construct expert and novice identity categories and responsibilities in talk with me. Given that interviews are often rich in explicit categorial references (e.g., friend, exchange student) as well as detailed descriptions of activities associated with such categories (Baker, 2004; Potter & Hepburn, 2012; Prior, 2011), MCA’s insights into the ways in which categorization operates in talk provided a useful starting place for an analysis of expert and novice categories in this context.  MCA was first elaborated by Harvey Sacks in the 1960s in tandem with conversation analysis (for a complete collection of his lectures on this subject, see Sacks, 1992). According to 95  Housley and Fitzgerald (2015), “MCA is not so much a method of analysis but rather a collection of observations and an analytic mentality towards observing the ways and methods people orient, invoke and negotiate social category-based knowledge when engaged in social action” (p. 6). MCA examines how categories, category resonant descriptions (Stokoe, 2012) as well as their locally associated predicates (Watson, 1978), such as activities (e.g., teach, listen) and attributes (e.g., knowledgeable, kind), are occasioned and used to perform social action in talk and text. Categorization analysis attends to “how categories are stipulated, how membership in a category is accountable and, particularly, how speakers proffer their category work as common, cultural knowledge” (Baker, 2004, p. 283). While the analyses in this dissertation do not rely solely on MCA, they use the analytic insights described here to make sense of participants’ interview talk. During this phase of analysis, I wrote up and compared analytic memos to look for patterns within and across interviews. Below is an example of an early memo I wrote for the data presented in Excerpt 4.1. Victoria asks So-Yi what she likes most about talking to Ami. This question doesn’t ask “if” but assumes that she likes talking to Ami. So-Yi provides a response to the question unproblematically with very little pause. After explaining that she likes that Ami asks how to do this or that, she provides a further evaluation to close her turn saying “I really like that part of her?” Victoria asks for further specification, of why she likes that part in particular, and So-Yi explains that it's because it means that Ami knows her and she is not ashamed, so it is not just because Ami is an “asker” but because So-Yi falls into the category of someone that can be asked. Thus, there appear to be two categories made relevant here: Ami as the confident novice “who asks” and So-Yi, the type of friend that can be asked. Both the asking and being asked are positively evaluated and provided as evidence for Ami being a likable speaking partner. As the interviewer, I also provide a very positive evaluation of this answer, further recognizing and ratifying her evaluations.  (So-Yi, Interview 1, 15/01/2016, 15:45-16:23) Figure 4.3 "She's not ashamed" memo. Draft analysis after initial reading of the data. 96  Over time, memos like these were refined, added to, and compared. Once a pattern was identified, for example the connection between asking and doing good exchange student, I searched for similar examples and counter-examples in an iterative analytic process of confirmation and retesting. To keep track of different category resources that appeared across the dataset, I created a table in which I recorded the explicit category mentions related to expert and novice identities as well as an associated list of predicates. Finally, the most prevalent categories and predicates were selected for discussion and the segments associated with them re-transcribed in more detail to facilitate in-depth analysis. The results of these analyses are reported in Chapters 6 and 7. 4.7.3 Interaction between PWU and JC peers The analysis of the remote observation data was principally designed to answer the second research question: What role (if any) does talk-about-language play in interaction, how is it initiated, and how is it discursively managed? To identify instances in which participants actively oriented to language-related issues, I proceeded on a case by case basis. I began by selecting segments for one focal participant and searching for patterns across those interactions. I then repeated the same process for the data of the other two cases. Excerpt selection As described in Chapter 1, to identify segments related to explicit language-focused talk, I borrowed the label talk-about-language from Levine (2009). For this study, I operationalized talk-about-language to include multi-turn segments of talk in which participants oriented to a language-related stance object – in other words, segments in which at least two speakers engaged in stancetaking that evaluated language form, meaning, or use; or language learning activities in talk (see Table 4.5 for examples). For this dataset, that included instances of repair focused on 97  linguistic items; spontaneous language-focused explanations; questions on topics related to English; observations about the ways others used English; and accounts/narratives in which language learner identities were used as resources to justify an action or behaviour (examples and details are provided in the method section). A new tag was applied for each new language-related stance object. In the first round of tagging, I used the following criteria:  1) Two or more speakers orient to a language-related stance object. 2) At least one speaker’s orientation to that object lasts two or more turns.32  As I examined the excerpts, it became apparent that talk-about-language took several forms. There were short asides in which a misunderstanding was resolved as well as instances in which a friend explained words to a JC peer, for example. Based on what I observed in the data, I created four further subcategories that reflected the circumstances that occasioned the relevance of talk-about-language sequences: repairs, explanations, reports, and accounts. These categories, while they have some basis in the literature, were principally inferred from a close reading of the data. The first category, repairs, refers to instances in which communicative trouble occasioned the relevance of a side-sequence designed to resolve the language-related problem. The second category, explanation, refers to instances in which PWU peers or JC peers pre-emptively described a language item in the absence of repair. The third category, reports, includes instances in which participants discussed language practices they had observed (e.g., saying “thank you” when getting off the bus). Finally, the account category includes instances in which participants’ membership in the language expert or language novice category served as a resource for warranting actions. Table 4.5 presents an example for each of these categories (see next page).                                                   32 These criteria were selected in part to exclude instances of self-repair.  98  Table 4.5 Examples of Talk-About-Language Sub-Categories Repair Explanation Report Account  Eating dinner in the dorm  15/01/2016, 4:34-5:10 Ami (A), Friend (F),  So-Yi (S)  ------------------------- A: did you know that in     Foodway we can buy        ro:ll salmon. F: oh really?  S: for sashimi? A: yeah. (0.4) sashimi. F: oh really?= A: =mm  F: I wanna buy! A: mm but it is ((mouth     full) F: big? A: mm    (2.5) ((Ami is chewing)) A: freezing= F: =o[h:    ] S:   [frozen] A: frozen! yeah frozen     one=  F: =frozen mm S: ((laugh)) A: freezing mean feel  S: yup A: just feel     [  ((laugh))   ] S: [yeah just feel] A: salmon feel freezing       like=  F: =ah: okay.((laugh))    frozen okay.   Chatting in the dorm 02/29/2016, 9:32-9:58 Samantha (S) Elizabeth (E) ----------------------- S: um yeah but (0.7)     and I went to ramen     restaurant? (1.4)     do you know ramen     (0.6)    S: ramen    (1.5) S: Chinese noodle.    ((laugh))= E: =yea:h yea:h  S: yeah I went (.)     ramen restaurant     (0.7) on Saturday?  E: was that good? S: yeah it was good     but it’s more     expensive than Japan E: oh!             Having brunch downtown 30/01/2016, 2:16-2:54 Ami (A) So-Yi (S) ------------------------- A: I also surprised       whe:n people get off     the bus they a-     every people     say thank you:= S: =yeah     (2.5) S: I do that in Korea?     but I guess I’m the     only one doing that. A: only one? S: only one saying thank     you when you get off     [in Korea] A: [oh::    ] yes yes     yes no- not everyone= S: =n:o=  A: =and especially like      when people get off     (.) like the just     behind the driver     side (0.4) people say     thank you or arigatoo     in Japan but even here     even the backside?     people still say     thank you. (.) it’s     really good   Chatting in the dorm */03/2013, 1:35-2:05 Samantha (S) Elizabeth (E) ------------------------- E: I woke up but my class      started=  S: =mhm E: and I’m like “oh my     goodness”  S: ((laugh)) E: like- S: -it’s okay that has     happened to me too      ((laugh))     yea:h I went there I     went class today and        (1.2)     there so so many     discussions?  E: mhm= S: =and so it’s a little     bit hard for me but       (0.7) I think it’s     good for me to speak     English so ((laugh))= E: =yea:h S: it’s nice opportunity     *Exact date unknown, during the third week of March *Note. Language-related stance object is presented in bold.  99 Across the recordings for all three JC peers (26 recordings, 11 hours 11 minutes), I identified a total of 149 occurrences of talk-about language. On average, a talk-about-language sequence occurred roughly every five minutes across all three JC peers’ data. In some cases, these sequences lasted as few as two turns. In other cases, the language item to which participants’ oriented was taken up and discussed across as many as 50 turns. This would seem to indicate that JC and PWU peers did in fact orient to language and language use relatively frequently in their conversations (or at least those that were recorded for this study). Tables 4.6, 4.7, and 4.8 show the distribution of talk-about-language sequences in interactions for each JC peer.  Table 4.6 Frequencies for Talk-About-Language in Ami's Remote Observation Data PWU Peers (recording time) Repair Explanation  Report  Account  Emma (95 minutes) 22 6 1 0 Paulisper (57 minutes) 14 6 1 0 So-Yi (53 minutes) 10 2 2 1 Totals 46 14 4 1  Table 4.7 Frequencies for Talk-About-Language in Samantha's Remote Observation Data PWU Peers (recording time) Repair Explanation  Report  Account  Elizabeth (101 minutes) 16 3 0 3 John (130 minutes) 17 8 1 0 Totals 33 11 1 3  Table 4.8 Frequencies for Talk-About-Language in Lisa's Remote Observation Data PWU Peers (recording time) Repair Explanation  Reports  Accounts  Serena, Ed, Kyla (234 minutes) 25 4 4 3  As the tables show, while explanations, reports and accounts constituted 45 of the total sequences, as may be expected, the majority of talk-about-language occurred in repair sequences. Repair accounted for 104 of the 149 talk-about-language sequences identified. It became 100 apparent that the majority of talk-about-language sequences arose in situations of conversational repair, as Levine (2009) alluded to in his definition; therefore, the analyses of talk-about-language focus generally on this type of sequence. The concept of repair is further elaborated on in Chapter 8. Discourse analytic approaches to stance As the second research question indicates, I was primarily interested in how participants initiated and managed talk-about-language moments. To ensure a thorough accounting of that management, I approached the data systematically with the following set of questions:  1. What language-related object was being oriented to?  2. What occasioned the relevance of a language-related topic in talk? (e.g., communicative trouble, an observation of language use by others, an experience) 3. Who initiated talk about that item and what resources did they use to accomplish the shift in talk? (i.e., who displays deontic authority and how?)  4. To what extent did the initiation produce a relevant and affiliative response from the other speaker(s)?   5. Who took up the knowing and unknowing positions related to the stance object (i.e., epistemic stance)? What resources did they use to take up those positions?  I produced a set of analytic memos for each tagged sequence and the memos were refined over time. The memos included a summary and identification of the language-related stance object, a sequential analysis of the talk leading up to the initiation of repair, a summary of the stances taken up by all speakers, and finally a reflection on the extent to which the repair was successfully accomplished and if affiliation was maintained (i.e., interlocutors displays of empathy and endorsement of other speakers’ evaluative stances, Stivers, 2011). Below I present 101 a repair sequence between Samantha, a JC peer, and Elizabeth, a PWU peer, and an early version of its associated analytic memo. For each case, I focused on features and patterns in the way the repair sequences were managed. I present the results of this analysis in Chapter 8.  Excerpt 4.2 “Midterm”  01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 Samantha: Elizabeth: Samantha: Elizabeth: Samantha: Elizabeth: Samantha: Elizabeth:  Samantha: Elizabeth: Samantha: Elizabeth: Samantha Elizabeth: Samantha:  Elizabeth:  Samantha:  $we- but we will have midte:rm$= ((laugh)) =what's that? midterm? (.)  oh [me?     ]    [the exam] oh you?  yeah you have. do you have midterm exam.  um next week (0.5)  [I d- ]  [um no] no not next [but-       ]                     [oh oh! just] in general.=  =°yeah°=   =um: (.) I have like papers?  ah okay okay.= =like a lot of papers=  =°mhm° (1.3.) and then like a midterm after reading week?  [but yeah.  ] [mm::! after] (Samantha, Remote Observation, 01/31/2016, 10:37-10:55)  Figure 4.4 Analytic memo for repair sequence "midterm." Draft analysis after initial reading of the data. In this excerpt, the stance object is the word midterm. It appears here that Samantha has understood Elizabeth’s question in line 02, “what’s that” as requesting clarification of the word midterm (i.e., what does midterm mean). This is clear from the way she reacts in line 03 by repeating the word with rising intonation and in line 05, when she glosses it as “exam”. Samantha is thus in some ways taking up the expert role here by advancing an explanation. However, I suspect that Elizabeth is using the phrase “what’s that” as a repair initiator (typical of British English) to request simply that Samantha repeat her initial statement. Elizabeth never does orient to Samantha’s specification or explanation, instead she pursues her own competing line by continuing to clarify different elements (i.e., me or you). It takes them several turns to finally get this misunderstanding sorted out. The end result is successful resolution; however, this repair necessitates a number of potentially face-threatening moves, including interruptions and requests for clarification. While it is clear that the misunderstanding is related to language, Samantha does not appear to overtly take up a learner (or unknowing) stance.  (Samantha, Remote Observation, 01/31/2016, 10:37-10:55) 102 4.8 Interpretation  As Duff (2008) notes in her description of case study methods, an important facet of case reports involves interpretation of the findings, in this case through the lens of LS. While the findings of each chapter individually provide insights into peer LS in this context, the four chapters are better understood as a mutually informing set of findings that illuminate different situational dimensions of peer language socialization. Therefore, to visually represent the relationship between the different facets described in each chapter, I will use the transdisciplinary framework of second language acquisition developed by the Douglas Fir Group (2016) described in Chapter 3 to summarize findings and, crucially, to make connections across chapters. The framework appears at the end of each results chapter (Chapters 5-8). 4.9 Conclusion In this chapter, I have provided an overview of this study’s method, including research context, recruitment, data generation, and data analysis procedures. Throughout, I have emphasized my epistemological commitment to viewing interaction data (both interviews and peer conversation) as situated co-productions that draw on particular ideological and semiotic resources. In the chapters that follow, I examine those resources more closely via the methods described here. Specifically, after providing an overview of each case in Chapter 5, I examine the co-construction of expert and novice identity categories in interviews (Chapter 6), the negotiation of roles and responsibilities related to talk-about-language learning in interviews (Chapter 7), and finally the collaborative management of talk-about-language in peer conversation (Chapter 8).  103 Chapter 5: Stories of peer socialization 5.1 Introduction Most results chapters in this dissertation foreground detailed analyses of talk rather than general facets of JC peers’ experiences. However, an important feature of LS research is the contextualization of the findings within the broader ideological, institutional, and social context, or the macro level as identified by the Douglas Fir Group (2016). Thus, before proceeding to a detailed analysis of the interview and remote observation talk in subsequent chapters, in this chapter, I present a more global portrait of Ami’s, Samantha’s, and Lisa’s SA sojourns. I then situate these accounts in relation to the context of the JC program, the institutional constraints that shaped these experiences, and finally, other SA case studies. In presenting JC peers’ experiences in this way, this chapter begins to answer the first research question about how JC and PWU peers describe their respective contributions to JC peers’ English learning. It does so by shedding light on the extent to which JC peers described interaction with PWU peers as a relevant site for English learning, the role of peer relationships in their SA goals, and the ways in which PWU peers’ contributions were shaped by institutional factors. The following stories of the participants’ experiences are based on their self-reports to me which I could not independently verify, but which I solicited and therefore co-narrate below. These stories should therefore be considered accounts produced for the purpose of this study. In the interpretation section of this chapter, I also discuss the ways in which these stories draw on and reproduce language ideologies related to assumptions about language learning in SA and English learning more broadly (see Chapter 2). To enhance readability, transcript excerpts are presented in verbatim transcription. 104 5.2 The story of Ami Ami was the first to volunteer for the project. From the beginning, according to Ami, English learning, making friends, and experiencing a broad range of cultural differences were key intertwined objectives. In her first interview, her professed goals for her SA sojourn were “to speak English fluently and make a lot of friends” (Ami, int.1). She also wanted to improve her TOEFL ITP score from 517 to over 550. When asked why she chose to study in Canada, she replied that in Canada, “there is so many people who is not, who are from Asia, like other country and multicultural” (Ami, int. 1) and explained that she chose it over the US for that reason. Ami noted that in the second semester, although two of her roommates were originally from Hong Kong and India, they had come to Canada as children and so were “like Canadians” (Ami, int. 1). Ami explained that their “Asian” ethnicities were: not expected before I came here but now it’s okay because they speak English so fluently, and they are so kind and it’s really easy to take conversation. I don’t know why but it’s really com-, feel comfortable to speak [to] Asian, like there is so many common things like traditional things or food or- it’s really similar and we can find topic easily. (Ami, int. 1)  Ami presented herself as a highly motivated individual. She was actively involved in almost every peer interaction opportunity offered by the institution: she tried to make connections with her assigned roommates, took part in the English Conversation Club organized by the JC program, applied for a Pair Talk partner, and participated in the Japanese-English Exchange Group. She also went regularly to the student recreation centre and explored other PWU clubs related to food. Despite all these efforts, Ami admitted that the first semester had been difficult. She had often felt helpless to understand her classes and roommates and spent a lot of time crying in frustration. She was surprised by how difficult it was to make friends whereas when she was in Japan, she had never had any problems.  105 She explained:  I expected it’s not so difficult things to make friend but for me, it is difficult because I can take a conversation with someone new people but after that, I don’t know how to keep contact with them, I can’t find good topic or good time, good opportunity to meet them, so I’m not so good for to make friend I guess […] so sometimes, I really feel depressed. Yes, I think it is one point I disappointed, not in Canada but for to myself. I didn’t know that I’m not good at to make friends because in Japan it’s really easy to make friend for me. (Ami, int.1) With her roommates, things were apparently not always easy either. She explained that interacting with her British roommate was especially difficult. She attributed the difficulties mostly to her own lack of familiarity with accents, saying:  She speaks British English so, so fast and I couldn’t understand at all the first day. It was my first day to come here and I lose my confident, I thought I can understand English a little bit but I came here and I couldn’t understand what she is saying to me and I feel really like negative? So, from first day, I lose hope. (Ami, int. 1) Ami explained that even after she asked her roommate to slow down or explain words, her roommate sometimes refused to take the time to explain. According to Ami, her Japanese friends and especially the support from regular Skyping with her family and boyfriend in Japan had really helped her get through the first semester.  At the start of the second semester, she was more confident and willing to redouble her efforts to engage with English speakers. She joined the study for that very reason. She said her new confidence for the second semester came from a homestay experience she arranged for herself over winter break in another part of Canada. There, she had reconnected with a Taiwanese friend she had met on a school trip and her friend helped her develop new strategies for speaking English with her host family. When I asked Ami what she did when she felt like she was not confident, she explained the role her friend played in supporting her English development: She’s from Taiwan and she helps me a lot, so I talked a lot to her and yeah, she helps me, support me about English or how to contact with communicate with the homestay family 106 […] Cause she have experience to homestaying for half a year so she’s kind of expert, so especially I asked her a lot about English and she helps me to speak English fluently. (Ami int. 1).  Ami was excited to take control of her situation and speak more English. She had been allowed to take an elective course and looked forward to meeting English speakers there.  Ultimately, she reported that her most successful peer relationship was with her second semester Pair Talk partner, whom she met at the end of January, just 10 weeks before her departure. Her partner was a half-Japanese student who also had a long-distance boyfriend in Japan and shared similar tastes in food and music. She explained that they often switched back and forth from Japanese to English and helped each other learn new words and expressions. For Ami, this relationship was a significant milestone in her English journey. Since her return to Japan, her partner has visited her several times and they remain fast friends.   By the end of her sojourn, Ami explained that even though it was harder than she thought to improve her English, she was satisfied because she did the best she could. She also noted that this research project had helped her achieve her goals:  Because of this project I have to speak English with someone, to do the recording, so it was really good, good time to speak English and this interview, so three times is really good time to improve my English because I have to, I can make sure this is the correct way to use or wrong, and then you also teach me the new words or expression. (Ami, int. 3).   5.2.1 Ami’s peers The peers that Ami recruited for this study did not become her close friends. Even though Ami often tried to meet with them, she noted that each of them interacted with her less or not at all by the end of the study. Ami could not identify a specific falling out that led to this result – in most cases, her PWU peers simply became busy or involved in different relationships or activities.    107 Cohort-mate Emma. Ami decided to recruit Emma to participate in the study because they had completed a group project together in the first semester and had been committed to speaking English together in the class and for the project. Like Ami, Emma was ethnically Japanese and spoke Japanese as her first language. She had lived in Chicago for a few years as a girl. She and Ami also liked to swim together so they decided to get together to speak English before swimming. Like Ami, Emma’s goal for her SA sojourn revolved around interacting in English outside the classroom but she had found it to be more difficult than expected:  The biggest purpose was to improve my English skills by interacting with many people in here but I think my opportunity to interact with local people is less than I had expected because um I’m really working on my assignment. So, I think I’m not good at making time to interact with local people. (Emma, int. 1). They both expressed that the research project was fun but they could not really be themselves when they spoke English together. When asked about the biggest challenge of practicing English with Ami, Emma said that because neither of them were native speakers, perhaps they were less able to “assume” (or infer) each other’s meanings:  When I talk to native speaker I think sometimes they are assuming from the content like how can I say, if I were not making a sense of the sentences probably they will assume because they are native speaker, but we not both native speakers, so sometimes misunderstanding will happen maybe. (Emma, int. 1)  Ami and Emma usually spoke Japanese together, often in group settings with other friends, and only used English for the study recordings. Although they met less often in March and April, the two managed to continue the project to the end.  Roommate So-Yi. A third-year exchange student from Korea, So-Yi arrived at PWU at the same time as Ami. While in some ways So-Yi’s exchange status made her like Ami, So-Yi had met the TOEFL requirements to take courses with mainstream students and was noticeably more proficient in English. So-Yi saw her time in Canada as a break from the stressful academic workload of university in Korea:  108 It’s really good to be here because like when I was in Korea, I was really stressed out for my jobs after graduation but here, I cannot do anything here, so like I can have my time, my spare time and like I can really spend my time thinking of what I really want to do in the future. (So-Yi, int.1) She also viewed her sojourn as an opportunity to be “comfortable with speaking” English (So-Yi, int.1). She explained that social life and making friends were her main goals and that most of her friends at PWU were Korean. In her first semester, she found Canada fun but by the second semester she had grown homesick and sought the comfort of Korean friends. She described Ami as her “first Japanese friend” (So-Yi, int. 1) and admired the way Ami confidently tried to use English even though she made mistakes. At the beginning of the first semester, they were on good terms and So-Yi was planning to visit Ami in Japan. However, at our second interview in March, Ami explained that So-Yi had been spending time exclusively with her Korean friends and Ami felt she could not participate. Soon after, Ami stopped recording conversations with So-Yi and asked that I not invite her for a final interview.  Friend Paulisper. A first-year student from central Canada, Paulisper met Ami while they were completing an assignment for Ami’s elective course. The two struck up a conversation and agreed to go to a Korean restaurant where Ami asked her to participate in the project. Paulisper was a Canadian student who had immigrated from Sri Lanka when she was a child. She only spoke English fluently and had not travelled much outside of Canada. She had been trying to learn Korean because of her affinity with Korean pop culture and had tried with limited success to make friends with a Korean exchange student during the first semester. When Ami asked her to participate, she viewed it as a good opportunity to try again to befriend an exchange student – she explained that most of her friends were from Canada and America. However, as the semester progressed, and assignments came due, the two found it harder and harder to make time to meet. Although Paulisper came in for a final interview in April, the two had not met since mid-March 109 and communicated only infrequently via social media. Ami still continued to message Paulisper with language questions until the end of the semester:  Sometimes she will ask me questions like about, because I told her if she had like grammar questions she could ask me, and she asked me once and I’m trying to remember how to explain it, hopefully it came across because I’m just kind of guessing, I’m not like an English expert (Paulisper, int. 2).  Paulisper regretted not staying in touch and admitted she probably did not message Ami as often as she should have.  5.2.2 Summary of Ami’s experiences Ami’s experiences with peer relationships were characterized by repeated attempts at making deep and sustainable connections with English speakers, most of which were apparently short-lived or unsuccessful. Despite taking full advantage of the considerable opportunities for peer contact that were offered by the institution, it was not until she reconnected with her Taiwanese friend in Ontario, who was an experienced homestay student, that she felt she gained the confidence to ask for help and communicate with English speakers. She thus redoubled her efforts and, by chance, was paired with a Pair Talk partner who shared her desire to improve her language skills as well as her other interests (e.g., boyfriends, food, and shopping). She viewed “making friends” as an important goal of her sojourn and felt she had gained confidence through her interactions with her Pair Talk partner (who did not participate in the study). Thus, the story of Ami’s peer socialization could be characterized as a challenging but ultimately successful one.   5.3 The story of Samantha Samantha approached me about the project because Ami, her good friend and fellow JC student, suggested it. Samantha had the highest TOEFL score (550) of the three focal participants. When asked what she thought of English, Samantha said she associated English 110 with the word “world” and described the value of English in terms of its value on the job market (Samantha, int. 1):  Victoria:  so if you had to describe English what would you say about the language what do you think about English in general. Samantha:  English one word? Victoria:  sure! let’s start with one word Samantha:  I think it’s mmmm world. I think cause most of people speaking English, and in Japan we have to, without English skill it’s so hard to get job, we can get job but it’s not so good pay, so I think it it’s so important to speak English also in Japan.   Her desire to learn English therefore appeared to be connected to her view of English as a desirable skill in the job market. Samantha’s main goal for her SA sojourn was to improve her English-speaking skills and TOEFL score. However, by the end of the sojourn, she said she had shifted her objective towards prioritizing English friendships, saying: “I just try to hang out with friend and speak English more and more here because when I go back to Japan I will miss a lot of time to speak English” (Samantha, int. 3). She also hoped to study immigration and viewed Canada’s “multiculturalism” (Samantha, int. 1) as an excellent experience for achieving her goal. When I asked Samantha if it was multicultural in the way she expected, she replied:  After I came here, I was so surprised because there are so many race and so many language, I thought it’s multiculturalism country but more like English dominant but it, many people speak French or Chinese or other languages and English, Japanese and yes, it’s so nice. (Samantha, int. 1) However, like Ami, in the first semester, she had a disappointing experience with peer relationships and PWU clubs.  When I was in Japan I didn’t think I struggling in my study abroad because I’m not person who always struggling? [...] but in the first time, I was, I think it was so hard to find opportunity to speak English. (Samantha, int. 1) 111 She explained that her roommates were “independent” (Samantha, int. 1) and that they did not often speak or eat together in her unit. In the first semester, she reported being friendly with a Korean roommate, who was kind and helped her settle in. However, her Korean roommate was only at PWU for one semester and she moved out of the unit in December. This is how Samantha described her first term experience:  I was more together with Japanese my friends, but I know, I know it’s not good but I need, so but when we speak Japanese with my friend, we can talk more like serious and worries about study abroad or something. Speaking Japanese is not bad I think but I need to speak more English. (Samantha, int. 2) Also, like Ami, she reported mixed experiences with native speakers of English:  English native speakers talk so fast and I’m so not comfortable because I cannot. That’s my fault I think because I have to speak more English and I have to improve my English more but I think it’s a little bit hard for me […] I feel like they doesn’t get used to hear non-native speaker English, it’s my thought, but I think so. (Samantha, int. 1) For winter break, Samantha arranged a homestay in the United States with some family friends and said she felt her English improved more there than at PWU. She explained that her improvement stemmed from her total immersion: there were no other Japanese speakers there, so she spoke English exclusively for almost an entire month. At the start of the second semester, she was determined to maintain the gains she had made during the break and felt more confident about seeking peer contact. She was also excited to be taking a mainstream elective course, and ultimately made what she considered to be a good English-speaking friend through a group project in that course.  A month into the second semester was a turning point for Samantha: firstly, she and John, her Pair Talk partner, began meeting and secondly, she was accepted into a hip hop dance group. Samantha found interactions with the members of the dance group the most satisfying since, she explained, they were mostly “internationals” who understood the challenges she faced (Samantha, int. 2). She interacted with them multiple times a week, both on and off the dance 112 floor, and they often spoke about topics that she did not regularly discuss with Japanese friends, such as religion. In this context, she felt comfortable asking questions and was also supported by a fellow JC student who attended and was also interested in speaking English with the hip hop club members.  By the end of her sojourn, Samantha said she had gone from speaking English only about 30% of the time to speaking English 80% of the time. She described her experiences in terms of a major shift in attitude:  Of course English skill is so changed, but I think the attitude, my attitude to speak English is so changed. So, I feel like more open for speaking English, so as I said I try to go outside to speak English. (Samantha, int. 2) She was ultimately pleased with her progress but felt that it was not enough and that she would have to work hard in Japan to maintain her gains.  5.3.1 Samantha’s peers The two peers that Samantha recruited for the study did not become close friends with her; however, they remained friendly throughout the semester. The most interesting facet of these relationships was how Samantha viewed their utility for language learning. Although her roommate was a native English speaker, Samantha felt that conversations with her were not always very helpful. In contrast, she described the conversations with her Pair Talk partner John, an international student from China, as being of greater pedagogical value.  Roommate Elizabeth. Samantha initially tried to recruit all her roommates via their unit Facebook chat but Elizabeth was the only one who responded. Elizabeth, a British international student, arrived in the unit in January, having just returned from an exchange in northern Africa. She was quiet and friendly and liked the idea of helping Samantha out even though she admitted that in January, she did not know Samantha well. She said she thought that Samantha was “pretty good at English” (Elizabeth, int. 1) and was happy to help her.  113 Throughout the semester they made grand plans to see some local tourist sites, to join a club, and to go to the gym together, but as the semester wore on, the timing never seemed to line up and they mostly just chatted in their unit. At the beginning of the semester, the two met often but by the end of the semester, Samantha had become close friends with the members of her dance group and spent less and less time with Elizabeth. When asked about the usefulness of talking with Elizabeth, Samantha had trouble identifying anything useful and explained: I don’t know but I think Elizabeth’s little bit shy? And so, I sometimes I have, I think it’s a little bit difficult to find some topics […] our faculty’s different and we, I don’t know, life is so different from Elizabeth and me. (Samantha, int. 2)  Despite numerous attempts to contact her, Elizabeth declined to come in for a final interview. Pair Talk partner John. John was paired randomly with Samantha at the end of January through the Pair Talk program. He was a first-year international student from Shanghai enrolled in the science program and was therefore younger than Samantha. He had originally wanted to teach Mandarin but was paired with Samantha to teach English and decided to make the best of it. Much to Samantha’s surprise, at the beginning, John treated Pair Talk as a relatively formal teaching opportunity and asked her to speak on specific subjects and offered careful correction. While she was shocked at first, Samantha admitted that she found the feedback very helpful and she was careful to focus on specific aspects of her pronunciation and grammar. John had a strong interest in education and strong feelings about the importance of making special efforts to integrate international students. He felt he shared a common culture and sense of humour with Samantha since they were both from Asian countries. In describing Samantha, John explained:  She’s pretty […] like shy. Okay, so the conversation between us is normally initiated by me, she doesn’t really find topics unless she unless I ask her. (John, Int. 1) 114 I was surprised to hear this assessment of Samantha, since she was one of the more outgoing JC students I had met. Although they had been interacting since January, Samantha did not recruit John until March, by which point they had become friends and more often discussed music, movies and Pokémon rather than engaging in the explicit types of language teaching Samantha described in the interviews. Because of his late recruitment, John only came in for one interview.  5.3.2 Summary of Samantha’s experiences Overall, Samantha’s peer interaction experiences mirror Ami’s in many ways. She experienced initial difficulties but was helped by her Korean roommate. After a positive homestay experience, she gained confidence and began initiating contact more often.33 Finally, through her dancing abilities, she gained access to a community of English-speakers that she could meet with regularly and ultimately felt that relationship was of great benefit to her English development. Like Ami, she pointed to a change in attitude in which she began to expressly seek out and create opportunities for connecting with English speakers outside the classroom. Also, like Ami, her closest connections were with multilingual students. Ultimately, Samantha’s goal shifted toward prioritizing out-of-class interaction over gains in TOEFL scores and this shift was also accompanied by a change in attitude about seeking English contacts. 5.4 The story of Lisa Overall, Lisa faced different challenges than Ami and Samantha. Lisa, by her own admission, was less academically inclined than the other two JC students. She received the lowest TOEFL score (463) and was not allowed to take electives. She told me she did not enjoy                                                  33 Both Ami and Samantha organized their homestays independently through providers or through family friends. No homestay options were provided through the JC program.  115 reading and was embarrassed by her TOEFL result. However, her oral English was relatively fluent and she said she loved speaking English from a young age – ever since she won a prestigious English competition in elementary school. She hoped one day to be able to translate for her parents when they took trips together. Lisa described English as a “tool” for communicating with the world:  I think it is really really good tool, ah because, ah we have all own native language like Spanish or French or Japanese or something but if we um- English is common language I think so um we can share my opinion or feeling by using English, even [with] another country’s people, so I think it is really great thing. (Lisa, int. 1) She chose Canada because she had already come to Canada for a trip in high school and felt that it was a “peaceful country” (Lisa, int. 1), in comparison to the US. Her choice had also been limited given her low TOEFL score. Her goals were to “speak English more smoothly” and “to make a lot of friends” (Lisa, int. 1).  Like Ami, Lisa engaged in every opportunity offered to her for peer interaction: Pair Talk, the Japanese-English Language Exchange, English Conversation Club, and even an English meet-up group hosted downtown off campus. She had a very difficult time for the first month of her sojourn, but her Japanese friends and a Korean roommate helped her through the most difficult period. She found that discussing her feelings of homesickness with her Japanese friends helped her feel less alone. She also explained how her Korean roommate helped her gain access to important information for managing her daily life at PWU.  At the first time I came here, I couldn’t speak English at all and I couldn’t understand English at all so I was so depressed because I wanted to speak. I really like talking with another people but […] I couldn’t conversation with my roommates. […] we need to decide the chores role or something but I couldn’t understand so, their English speaking was so fast for me […] but it is really really important thing to share like kitchen or bathroom or something, so I asked to my Korean roommate […] and she’s also exchange student but she can speak English then, so but she understood me, my feeling. (Lisa, int. 1) 116 At the beginning of the new semester though, that roommate moved out and two new roommates, both of whom were new to PWU, moved into her unit. Lisa explained that they got along well and she was happy to develop such a close relationship with them so quickly. So, unlike Ami and Samantha, Lisa faced few difficulties cultivating an English-speaking social life after the first month. She often went to house parties, interacted with other JC students’ roommates, and met up with exchange partners outside of the appointed times. She had a reputation amongst the JC participants and her roommates for being loud, cheerful, and goofy. She explained that her bigger difficulty was managing her time between being with her English and her Japanese friends. She explained they were both very important to her and she wanted to “cherish the balance” (Lisa, int. 2). Despite the many English relationships that she built, for Lisa, her relationship with her roommates was the most fulfilling and comfortable.  By the end of March, she felt she was speaking less English than she had been at the beginning of the semester. With the end of term coming up, her Japanese friends wanted to travel and celebrate, and she also had many exams, so she spent less time with her roommates. However, she ultimately felt very satisfied with her experiences and her progress, and felt that in the second semester especially, she had improved. She attributed her success to attending conversation groups where she could get some structured instruction and to her informal chats with her roommates.  5.4.1 Lisa’s peers Lisa recruited three of her four roommates, who despite having just met Lisa, eagerly volunteered. As a group, they would often tease each other and gossip about boys, food, classes, and dieting. The conversations recorded amongst the roommates were more casual and personal than other recordings submitted by the other focal participants. All three friends noted that Lisa 117 often asked for help with her English, whether it was for assignments or Instagram posts, and that they admired her hard work and dedication to English learning.  Roommate Ed. Before attending PWU, Ed had been living in Singapore where she had attended an international high school. Her mother was Filipina and her father Bengali, so she spoke both their native languages and had an expert command of English. Ed arrived at PWU at the beginning of the second semester and described herself as a bubbly person who loved to tease and make trouble. Lisa was the first roommate that Ed met when she moved into the flat and Lisa helped her through her homesickness:  The first week I was really homesick, I always went to her, I was like “I really want to go back” and then she’s like, “me too.” But yeah, that’s sort of how we bonded because we were both like quite homesick. (Ed. Int. 1) She explained that in the flat, there was a running joke that she and Lisa were “party girls” (Ed int. 1) since they were the only two that drank alcohol in the flat. She explained that her time at the school in Singapore had allowed her to meet people from all over and from different cultures. That had allowed her to learn a lot and get used to speaking to English learners, like Lisa.  Roommate Serena. Born in Dubai, Serena grew up living less than an hour’s drive from PWU. She would often go home on weekends and worked with local charity organizations. She was extremely health conscious and tried a number of diets throughout the semester, which ended up being a focus of numerous recorded conversations. Like Ed, it was Serena’s first semester at PWU and Lisa was her first PWU contact. Serena described Lisa as “sweet” and explained that “she just gets everyone together” and “is ready to go out and try something new” (Serena, int. 2). Having attended some years of elementary school and all of high school in Canada, she also had a native-like level of English; however, she sympathized with those who 118 had difficulties speaking it. She said she admired Lisa for working so hard to improve because she knew how difficult it could be: For me personally, because I moved here when I was 10 and I was teased a lot for my accent and stuff, so I think what I do learn from her is just to not be embarrassed, you know, just go with the flow, it’s a process to learn and just have fun with it I guess. (Serena, int. 2) For Serena, the lesson she would take away from being friends with Lisa was “don’t be afraid to just ask or get help” (Serena, int. 2). Roommate Kyla. A third-year student and residence advisor for the floor, Kyla arrived in the unit at the same time as Lisa, in September. Kyla was born in China but immigrated to the region around PWU at a young age and thus spoke English with native-like proficiency. She also frequently spoke Mandarin with her friends and family. Like Serena, she often went home on weekends and volunteered for local charity organizations. Kyla was focused on academics: she did not drink or swear and routinely chastised Lisa and her roommates for doing so. She said her first impression of Lisa was that she was someone who “tried hard” (Kyla, int. 1) and that Kyla liked that about her. She felt that Lisa could be quiet sometimes, but that Lisa had improved and spoke much more when just the two of them interacted.  5.4.2 Summary of Lisa’s experiences Lisa’s case shares commonalities and differences with the previous two. Like Samantha and Ami, she actively took advantage of the peer interaction opportunities organized by the institution and was guided in her initial few months by a Korean roommate who helped her settle into life at PWU. However, Lisa had relatively less difficulty making connections with other students and formed a close relationship almost immediately with her roommates. She also prioritized speaking English and making friends as the goal of her SA sojourn from the first interview, rather than other more academically focused goals such as increasing her TOEFL 119 score. In the end, Lisa viewed her sojourn as a success and felt especially proud of the relationships she had formed with her roommates. In addition, Lisa’s reports of her experiences place more emphasis on the need to balance Japanese and English relationships and on the benefits of both kinds of interaction and relationships. Finally, the story of Lisa most clearly emphasizes how empathy was an important component of the relationships she formed. Peers who became close to Lisa expressed empathy for the challenges of being a language learner and were all previously English language learners themselves.  5.5 Characterizing cases These descriptive stories allow a clearer picture of the nature of these cases of peer socialization. Via institutionally provided opportunities such as on-campus housing, clubs, and language exchanges, Ami, Samantha, and Lisa described situations in which they established regular contact with English-speaking peers and formed at least one satisfying relationship. For Lisa, her university residence provided the necessary environment for making social connections; for Samantha, it was a PWU club involving hip hop dance; and for Ami, it was by participating in a PWU language exchange. At each of these sites, JC and PWU peers engaged in joint activities, such as cooking, dancing, and language learning. By the end of their sojourns, all three students felt they had achieved their twin goals of improving their English and making friends. Given the overall positive outcomes described by the students, these students’ experiences appear to represent successful peer language socialization in an SA context.  5.6 Situating JC student cases A key feature of case studies involves their contextualization (Duff, 2008). In what follows, I contextualize these three cases by considering first the extent to which the cases 120 presented are similar, how these cases might compare to other cases within the JC program and PWU, and how these cases compare to others cited in the SA literature. In the final section, I situate these accounts in relation to the common assumptions of SA (see Chapter 2) and highlight the ways in which JC peers’ accounts draw on and reproduce language ideologies.   5.6.1 Comparison across cases  While there are of course important individual differences across these three cases, the accounts of JC peers’ experiences across the sojourn were relatively similar. Despite having clear goals that prioritized peer contact from the outset of their sojourn (i.e., the goal to make friends), all three students represented their initial few months as a challenge. For Ami and Samantha, their perceptions of themselves as outgoing and socially adept people were threatened by their lack of success in forming close relationships. They explained that they had not previously felt like the type of people that normally “struggle” (Samantha interview 1). Ami especially was frustrated by her inability to make friends, when in Japan it had been so simple. Their stories highlight a mismatch between their expectations of interaction and real access to interaction in English once they arrived. The stories also all highlight the finding that the contributions of peer interaction are not homogenous or stable. These three JC peers had both positive and negative experiences with different PWU peers. They all felt that certain peers, namely international or multilingual peers, were especially supportive and empathetic. While all three students reported enrolling in multiple programs that would theoretically offer opportunities for peer interaction and relationships, they were only able to form meaningful connections via a select few of these opportunities. Additionally, the meaningful relationships that each JC peer did form all emerged via different institutionally managed programs (i.e., language exchange, dance club, university 121 residence). Thus, there did not appear to be one program or opportunity that served these three students globally better than other opportunities. The striking similarities between the goals and ultimate successes of these three JC peers is likely not a product of happenstance but rather a product of research design. As mentioned in Chapter 2, this project assumed that JC peers would seek out interactions in English beyond the classroom and that such interactions would be valuable learning resources. The recruitment materials explicitly targeted JC students who were motivated to improve their English skills and expand their opportunities for peer contact. Therefore, the goals that students identified were in some ways a prerequisite to participation in this study.  However, not all aspects of JC peers’ experiences were anticipated by the original design. The recruitment strategy for this study originally assumed that given the focus on peer interaction, the JC peers who volunteered in the second semester would already have formed connections with English speakers in the first semester. However, as the JC peers’ stories highlight, that was usually not the case. Rather, even these three highly motivated students had difficulties accessing and maintaining English-speaking relationships in the first four months. This is an important finding since it challenges the common assumption that the reason SA students have difficulty interacting in the host community is because they are lazy, unmotivated, or shy (Pellegrino, 2005). Instead, it suggests that making such connections is difficult even for resourceful and motivated students.    5.6.2 Cases within the context of the JC program and PWU While these three students cultivated satisfying peer relationships in English, not all JC students formed friendships with English speakers – or even desired to do so. Samantha explained that about one in four students in the program cared more about partying with 122 Japanese friends than learning or practicing English with peers outside the classroom. In other words, according to Samantha, forming relationships in English was not a priority for those students. In other cases, such as the case of Ami’s JC colleague Emma, academic obligations were prioritized over informal interaction outside the classroom. Emma explained that she felt she had little time to attend the various conversation groups because of her heavy workload and only managed to connect with her Greek roommate near the end of the second semester.  For other students attending PWU on exchange, English learning might also be deprioritized for other reasons. For example, So-Yi, who was on exchange from Korea for the same period as Ami, explained that her TOEFL score was already high and that while she wanted to be more comfortable using English, that was not the main objective of her sojourn. Instead, she sought to have fun, make friends, and temporarily escape the intense academic pressures of the Korean university system. According to Ami, by the second semester, So-Yi had only Korean friends and spoke Korean predominantly in their shared apartment.  Given these diverse stories, it is important to highlight that the cases reported in this dissertation pertain to three highly motivated and resourceful students who often prioritized English learning over other competing goals (e.g., academics, relationships with co-nationals, and travel) and viewed interaction outside the classroom as a privileged site for achieving their aims. 5.6.3 Connections to the SA literature  As noted in Chapter 2, high expectations for social contact, and the ensuing disappointment when those expectations are not met, are both common findings across the SA literature (e.g., Jackson, 2008; Kalocsai, 2009; Meier & Daniels, 2013; Mitchell, 2015; Yang & Kim, 2011). For the close friendships they did create, Ami, Samantha, and Lisa as well as their 123 peers highlighted the JC peers’ non-language-related contributions to the relationship. Previous work in SA (Goldoni, 2013; McGregor, 2012; Trentman, 2013b) has also examined how possessing social capital that is valued in the local community can allow students to gain access to TL relationships. Trentman (2013), for example, explored the Arabic language use of two US students studying in Egypt who gained access to a community of Arabic speakers by joining a local rugby team. Trentman noted that both students were valuable team members since they had played college rugby prior to their sojourns. For those students, that community therefore became a significant site of peer socialization while the remaining 29 students participating in the study had relatively little access to local communities. Trentman also highlights the fact that for these students, just signing up for extracurricular activities was not enough to promote sustained engagement since many other students on the exchange had not experienced the same level of success. Goldoni (2013) also reported the cases of five (out of 44) US exchange students who, by pursing interests or passions, were able to meet Spanish speakers with similar interests in Spain. She mentioned, for example, Leslie, Emma, and Scott, who drew on their background in the Christian church to scaffold their participation in a Bible study group that met weekly. However, the number of students who were able to mobilize exceptional talents or interests to gain access to relationships was manifestly quite low.  Ami’s, Lisa’s, and Samantha’s experiences appear to mirror those of the successful students in Trentman’s and Goldoni’s work insofar as the relationships formed around a mutual interest in a particular activity and the JC peers could contribute meaningfully to that community with non-linguistic knowledge. Samantha, for example, possessed the necessary dancing skills to contribute to group performances – she was an experienced dancer who was able to pass try outs, show up regularly for practices, and participate in the group’s public performances. Lisa 124 contributed to her roommate community in several ways. Her roommates, Ed and Serena, described how she served as a guide to the residence by showing them laundry facilities, for example. She also served as a focal point for gossip in the house since she was willing to openly share embarrassing stories about her love life. In Ami’s case, she had a long-distance Japanese boyfriend and so did her Pair Talk partner. In other words, sustainable relationships in these narratives were described as those to which all members contributed in a variety of ways. The students’ stories also reinforce the findings of previous literature (Diao et al., 2011; Kinginger, 2008; Knight & Schmidt-Rinehart, 2010) that access to opportunities for meaningful interaction and relationships are not automatically generated as a result of proximity to TL speakers through shared accommodation. Given the parallels across cases and with the SA literature, it is probably fair to say that these case studies represent typical experiences of SA students who prioritize language learning through peer contact and feel their sojourns have been worthwhile.  Finally, these accounts present distributed and dynamic portraits of each JC peer’s individual social networks. As Zappa-Hollman and Duff (2015) point out, relationships both within the local context (i.e., roommates, exchange partners, homestay friends) as well as transnationally (e.g., family and boyfriends) can serve to support or undermine students’ language learning experiences and goals at various junctures. In each story, there is one peer who emerges as a broker in moments of linguistic difficulty. In the sociology literature, brokerage is defined as “the process of connecting actors in systems of social, economic, or political relations in order to facilitate access to valued resources” (Stovel & Shaw, 2012, p. 141). Lillis and Curry (2006), who examine brokering in academic literacy, conceptualize brokering as mediation where brokers “are involved in helping people interact with written texts” (p. 12). Lillis and Curry identify several categories of broker, including nonprofessional brokers whose 125 involvement “can be characterized by their personal relationship to the author and their serendipitous knowledge of English” (p. 14) and language professionals, who include English-language specialists. In each story, the JC peers named at least one nonprofessional broker who, while not necessarily an English expert in the conventional sense, was able and willing to facilitate interaction between JC peers and other English speakers. In both Samantha’s and Lisa’s cases, it was a roommate (a Korean exchange student). In Ami’s case, it was a Taiwanese acquaintance, who was an experienced homestay student. All three JC peers described their brokers as kind people who offered useful interpretations of the practices of more expert members following a challenging encounter. In Ami’s case, her Taiwanese friend explained the behaviour Ami was experiencing in her homestay. For Lisa, her Korean roommate took the time to explain chore rotations. In each case, these peers spoke English more expertly than the JC peer but were not native speakers and had not grown up in Canada. In this sense, they were not only relative experts in English but also had more expertise in navigating the SA experience. The JC peers also identified interactions with me, a language professional, as helpful for their language development. For example, in Ami’s interviews, she pointed out that I sometimes taught her new words in our conversations. In this sense, I was also an important broker for these participants.  Wenger (1998) identifies brokers as those who have “multimembership” (p. 109) and who can, by virtue of their knowledge of more than one community, bridge gaps for peripheral participants. These international peers appeared to have developed such multimembership and were thus able to recognize the language struggles of the JC peers and offer useful advice to scaffold their participation. The language brokerage literature has tended to focus on generation 1.5 students and the ways in which they broker interactions between their parents and institutions (Morales & Hanson, 2005). However, several studies have also examined brokerage between 126 peers (Bolden, 2012; Talmy, 2005). Duff (2007), for example, has observed that Korean international students at a Canadian university also found Korean generation 1.5 students to be the most effective linguistic and academic socializers. According to JC peers’ accounts, these peer socializers’ empathy for their challenges as well as their ability to accommodate to JC peers’ English levels (e.g., by speaking more slowly) at moments of significant difficulties made them key resources for ongoing participation in peer interaction. All three JC peers explicitly recognized the importance of these students in their experiences and the special accommodations these peers offered.  5.7 Interpretation  Language ideologies serve as powerful interpretative resources that mediate the ways in which we evaluate the language use and behaviour of ourselves and others. According to Jaffe (2009), “[l]anguage ideologies are reflected in explicit statements about language (in metalinguistic discourse); they are refracted in practices that orient towards or draw upon ideologies as resources, and are also embedded as presuppositions of discourses” (p. 391). In response to interview questions about the value of English and their reasons for learning it, all three students produced a rich array of explicit metalinguistic discourses. Two key ideologies emerged as resources for the JC peers when constructing the stories of their sojourn experiences: neoliberal ideologies related to English as a global language and ideologies of immersion. 5.7.1 English as a global language As discussed in Chapter 2, neoliberal ideologies related to language learning extend market-based based principles to the acquisition of languages, and particularly to the acquisition of English, framing language learning as a means for upward social and economic mobility (Kubota, 2016b; Shin, 2016). JC peers’ descriptions of their reasons for learning English aligned 127 with this understanding of English in several ways. Firstly, JC peers produced accounts that framed English as tool for “global economic competitiveness” (Wee, 2006). Samantha, for example, discursively linked upward career mobility and the ability to speak English, explaining that “without English skill it’s so hard to get job” (Samantha int. 1). JC peers also used the words “tool” and “world” to describe English. Secondly, JC peers’ accounts constructed the role of English primarily as an inter-ethnic lingua franca. Lisa, for instance, explained that her skills in English would open up opportunities for world travel and the experience of different cultures saying, “English is common language […] we can share my opinion or feeling by using English, even [with] another country’s people” (Lisa, int. 1). Together, these explicit statements about English reproduce neoliberal notions of English as a tool for gaining access to an international world elite (Kubota, 2016a) and equate English language learning with self-improvement  (Warriner, 2016).34 Ideologies of English as a global language also served as resources for valuing particular aspects of their SA experiences. One area in which this phenomenon can be observed is in JC peers’ stances on multiculturalism. Both Ami and Samantha specifically use the term “multicultural” to justify their choice to come to Canada. In valuing this multiculturalism as a site for English learning, these students implicitly reify the notion that English serves as a tool for accessing multicultural communities rather than a tool for communicating solely with native speakers. These discourses were again reproduced in the ways each student described the value                                                  34 Ideologies which connect language with culture (and particularly nationality) were still strongly present in JC peers interview accounts. However, these ideologies were usually reproduced in relation to languages other than English (e.g., in relation to their “Japaneseness”). This is similar to the way Wee (2006) describes the differences in the ways that Singaporeans value English versus “mother tongues” such as Chinese and Malay. According to Wee, mother tongues in Singapore are viewed as “repositories of cultural value” (p. 350) while children “must learn English so that they will have a window to the knowledge, technology, and expertise of the modern world” (p. 349). 128 of using English with international and multilingual friends. Samantha, for example, explicitly commented on the international composition of her dance club friends, and explained that she felt comfortable asking questions because of their empathy and history as English learners. Ami noted that using English with her Asian roommates was easier since they had more in common and could find topics to talk about more easily. Lisa emphasized how she could learn about the cultures of other people who spoke Spanish or French through the use of English.  While native-speaking peers were mentioned, they were often not described as positive contributors to language learning. In fact, the JC peers’ narratives often highlighted specific encounters with native speakers as unpleasant or unproductive, often because the native speaker failed to accommodate to the JC peers’ level or recognize their communication difficulties. Ami’s reports about her British roommate point to how that roommate seemed to be unwilling or unable to take on a mediating role by providing explanations or slowing down her speech. Similarly, Samantha’s reports about native speakers not “being used to” non-native speech imply that it is the native speakers that are unwilling to do the work to listen and understand. Multilingual and international peers, on the other hand, were described as understanding and supportive. Although Ami and Samantha both recognized in their accounts that their own lower proficiencies were barriers for their comprehension of fast native speech, their stories also attribute some responsibility to the native speaker, who must accommodate to their level. It appears, then, in this context that neoliberal ideologies that constitute English as a tool for connecting with a range of cultural others (and not just native speakers) operated as a resource for producing positive evaluations of interaction with international and multilingual peers.35                                                  35 In most cases, the peers with whom JC peers connected were of Asian ethnicity, pointing to the potential relevance of a “pan-Asian, transnational, multilingual” community (Duff, 2010, p. 182). Indeed, Samantha 129 While it is common for SA students across contexts to spend the bulk of their time with international students (cf. Coleman, 2015) and to describe interactions with international students as easier and less threatening to their sense of self-efficacy in English (cf. Kaypak & Ortaçtepe, 2014), it is less usual for them to describe interactions with international students as better for their language development than interaction with native speakers. However, some research conducted in English-speaking contexts has noted that students may come to value more international encounters (cf. Kalocsai, 2009). In their work on the individual networks of practice of undergraduate Mexican SA students sojourning for up to one year in Canada, Zappa-Hollman and Duff (2015) report on one student, Isabel, who initially was disappointed with her lack of contact with Canadian students but eventually came to view international students from countries other than Mexico as just as good if not better friends than Canadians might be. She explained that being around international students made her more open-minded, whereas the younger Anglophone Canadian students in her classes were inexperienced and wasted a lot of time when completing group assignments.  JC peers’ positive attitudes toward diverse international others were likely shaped to some extent by discourses encountered in program advertising as well as the course content of the JC program. While this study was not designed to examine the impact of classroom learning on peer interactions, it is also important to acknowledge the potential role of curriculum in shaping JC peers’ attitudes and experiences. For instance, JC classes encouraged JC peers to explicitly question the native speaker standard through reference to topics such as World Englishes and English as a lingua franca. In their classes, JC peers often discussed                                                  specifically referenced the Asian ethnicities of her dance club members and how that shared background contributed to their ability to communicate easily with each other.  130 multilingualism and were asked to describe their multilingual repertoires. Research by Morita (2013) demonstrated that addressing topics such as World Englishes in the classroom can have an impact on the ways that students describe the value of English. Her study showed that Japanese students studying World Englishes in Japan tended to view English as a tool of intercultural communication with speakers of a range of first languages and resisted the association of English with native speakers exclusively.  The JC peers’ emphasis on multiculturalism and lack of emphasis on native speakers also appears to reproduce discourses present in the Japanese language policies of interculturalism and internationalization. A language-education policy analysis conducted by Liddicoat (2007b) found that English was viewed in Japan as the “unquestioned international language” (p. 36). Liddicoat explains that in Japanese, the very nature of the word internationalization, kokusaika, implies the meeting of different countries and frames the use of English in international contexts as an intercultural affair. This emphasis on internationalization and English in Japanese higher education has only intensified in recent years. Canada has also adopted a nation-wide marketing strategy for higher education which promotes both peace and multiculturalism as benefits of studying in Canada (Government of Canada, 2014) and this policy discourse has trickled down to strategies for program advertising which often include words like “global” and “multicultural.” In addition, JC courses included critical multicultural perspectives. Thus, a number of policies and contextual factors appear to have converged to produce a context in which these JC peers, drawing on circulating norms that connected English with intercultural communication, produced stances in which multilingual peers were evaluated as especially positive contributors to their language development.  131 5.7.2 Immersion  Ideologies of immersion construe language learning as occurring through intensive exposure to (and interaction in) the TL and culture. As described in Chapter 2, in SA, this ideology is most often manifest through a prioritization of “learning by doing” (Doerr, 2015). In other words, SA students are expected to acquire the TL through participation in mundane experiences without clearly set goals or supervision. In many ways, ideologies of immersion were embedded in the design of this study. The study sought to investigate the role of out-of-class interaction with peers and specifically frames that interaction as an opportunity for language learning broadly conceived. By asking JC peers about their contact with English-speaking others and the value of that contact for learning, I reinforced and reproduced the link between mundane experience with others and language acquisition.  However, as I have mentioned previously, discourses and ideologies are co-constructed resources for achieving talk: in interviews, JC peers often oriented to and reproduced ideologies of immersion when describing their sojourn goals and experiences. For instance, JC peers’ goals for improving their English tended to be non-specific (although in Interview 2, Samantha explained that she wanted to improve her ability to say numbers quickly) and they construed contact with English-speaking peers (and not classroom learning) as the principal means by which they could improve their English in a general sense. This was clear from the way in which all three JC peers identified improving (spoken) English and making friends as intertwined objectives for their stays. In addition, JC peers linked the opportunity for informal interaction in English to their limited time in Canada, excluding the possibility that such opportunities could also be readily available in Japan. Samantha explained that her priorities had shifted at the end of sojourn toward using English with friends because she would not be able to use English that way 132 upon her return. Doerr (2016) explains that immersion ideologies portray difference (linguistic and cultural) as inherently tied to the host community. She goes on to note that this valorization of difference exoticizes SA destinations and leads to homogenized understandings of both the home and host countries. For this reason, SA students often fail to perceive or acknowledge their home communities as multicultural and multilingual spaces that, in many cases, can provide ample opportunities for contact in the desired TL, in this case, English.    While ideologies of linguistic immersion were certainly reified through the research design and the JC peers’ accounts, ideologies of cultural immersion in the traditional sense were less evident than is often the case in SA research. As Kubota (2016b) has pointed out, research in SA has usually found that students expect to be surrounded by “locals” in the form of idealized native speakers whose ethnicities, characteristics and experiences match circulating stereotypes about what “authentic” speakers should look, sound, and act like. This SA expectation is underpinned by language ideologies that connect place, language, and cultural, national, or ethnic identities (Iino, 1996; Jaffe, 2009; Wee, 2006). However, as the discussion above highlights, JC peers were not overly concerned with these traditional notions of authenticity. Instead, by the beginning of their second semester, they reported placing high value on relationships with members of international and multilingual communities. I would argue that in some senses, encounters with multicultural others was indeed a kind of authentic cultural immersion from the JC peer perspective. Given the program advertising and their contact with students from previous years, JC peers reported that they expected an “authentic” Canadian experience to consist of immersion in multi-ethnic, multilingual, and multicultural communities. In other words, JC peers equated Canadian culture with multiculturalism. The JC peers also contrasted Canada as a “multicultural place” with Japan as a relatively culturally homogenous 133 space. In this sense, JC peers reproduced the host/home dichotomy that Doerr (2013, 2016) describes as being at the heart of ideologies of immersion.  5.7.3 Summary Ideologies of English as a global language and immersion that were produced in the JC peers’ narratives correspond to what the Douglas Fir Group (2016) identified as macro level ideologies. I have thus mapped them onto the diagram that was introduced in Chapter 4.   Figure 5.1 Language ideologies in JC peer stories. Model adapted from the Douglas Fir Group (2016)   These ideologies are widely circulating resources that JC peers and I drew on when describing, interpreting, and storying their sojourns and their experiences of peer interaction. As the coming chapters will demonstrate, they are also important macro-frames for making sense of how JC and PWU peers describe their roles in JC peers’ language learning and engage in talk-about-language. As described in Chapter 2, these ideologies are neither novel nor uncommon in SA – they appear in various forms across the SA literature. However, by identifying how such 134 ideologies operate, as researchers, we open them up to critical scrutiny and are better able to examine how they are reproduced at various levels of discourse and interaction.  5.8 Conclusion In this chapter, I have provided an overview of Ami’s, Samantha’s and Lisa’s experiences of peer socialization and argued that these three students represent somewhat successful cases within the JC program. I have also highlighted how each student experienced difficulties in accessing and maintaining stable peer relationships, despite taking advantage of the many institutionally organized programs available. In presenting these students’ stories, I have provided context for the analyses in chapters that follow and situated these cases alongside other possible cases within the program, PWU, and the broader SA literature. In the final section of this chapter, I also argued that the JC peers’ views of English as tool of intercultural communication and their understanding of Canada as a multicultural country might have enabled students to place additional value on the contributions of a range of cultural others, and to deprioritize the role of native speakers in this context. In Chapter 6, I explore the identity categories used to describe JC and PWU peers when they are oriented to as language experts or novices and the ways in which these categories reproduce ideologies of English as a global language and immersion in situated ways.  135 Chapter 6: Expert and novice peer identity categories 6.1 Introduction The previous chapter highlighted how JC peers emphasized the role of supportive peers with greater TL expertise who accommodated their linguistic needs and provided understanding at crucial moments during JC peers’ sojourns. It also pointed out that JC peers appeared to be highly motivated language learners who explicitly sought out and created opportunities for interaction with PWU peers. In this chapter, I more closely investigate the discursive construction of the identity categories associated with these supportive experts and motivated language learners. Thus, this chapter contributes to answering the first research question about how JC and PWU peers described their respective contributions to JC peers’ English learning.  The analyses presented showcase how JC peers, PWU peers, and I, as the interviewer/researcher, oriented to particular actions or attributes (or predicates, in MCA parlance) as normal or desirable for expert and novice peers in this SA context. I present data that show similarities in how participants (PWU peers, JC peers, and I) across the three cases oriented to the possible identity categories available to JC and PWU peers when language learning was made relevant in our talk. Through this analysis, I aim to gain insight into “local ethnographic categories” (Bucholtz & Hall, 2005, p. 587) related to English expertise in peer relationships. I argue that the ways in which participants constructed and oriented to expert and novice peer categories drew on and reproduced ideologies related to English as a global language as well as lay assumptions about processes of language socialization in SA. The findings in this chapter also demonstrate how, by categorizing themselves and each other in specific ways, JC and PWU peers were able to accomplish social actions, and particularly, produce positive evaluations of each other, within the contexts of the research interviews.    136 6.2 Focus on categorization  As described in Chapter 3, identity categories and their attendant features are used to make meaning in situ. In other words, identity categories are discursive resources that allow people to accomplish social actions in talk. For example, imagine the following hypothetical interaction:  A: I need to get my essay proofread, do you know anyone good?  B: Yes, my friend Anne can help you, she’s a native speaker.  In this context, B’s use of the category native speaker to describe Anne as a candidate “good” proofreader relies on shared cultural knowledge: in this case, a particular language ideology in which members of the native speaker category are endowed with the attribute of expert knowledge of English, and thus, can engage in competent proofreading. Via explicit mentions of categories, or by alluding to activities or characteristics (i.e., predicates) that strongly suggest those categories, speakers are able to produce inferences in talk that allow them to accomplish acts such as identifying, blaming, or explaining (Jayyusi, 1984; Sacks, 1992) – or in the case of the example above, providing a rationale, or warrant, for an assessment (i.e., of Anne as a good proofreader).36  What allows categories to be deployed as resources for meaning making is speakers’ shared cultural knowledge of what those categories represent. Jayyusi (1984) refers to explicit category mentions as category-concepts, or “individual descriptor designators” (p. 23) and maintains that they are usually “already culturally available” (p. 20). Pomerantz and Mandlebaum (2005) further explain that “participants incorporate explicit relationship categories                                                  36 All categories are indicated in italics in order to highlight that these are labels, or discursive resources, used to talk about participants in situ. They do not necessarily reflect the ways in which participants were categorized in this research (i.e., JC peers/PWU peers) or the ways participants choose to identify themselves more generally.  137 anticipating that recipients will draw on their understanding of the activities, motives, rights, responsibilities, and/or competences associated with incumbents of the category” (p. 152-153).37 However, there is much more to examining categorization practices than simply identifying explicit category labels such as native speaker. Pomerantz and Mandlebaum (2005) explain that analyzing explicit category mentions is insufficient for identifying commonalities in how they can be used interactionally. Rather, the scope of inquiry must include categorization practices more broadly. That is to say that the analyst must examine how speakers engage in sense-making through discursively configuring, or ordering, social categories as well as how speakers claim, attribute, and resist inclusion (or incumbency) in those categories (Day, 1998). Thus, studying categorization includes taking into account how features (Jayyusi, 1984), or predicates (Watson, 1978) – attributes, characteristics, knowledge, or actions – become discursively tied or bound to categories and the social actions that speakers accomplish by doing so. It also includes examining how such predicates can be used to categorize speakers as incumbents in an identity category and the interactional consequences and inferences that are made publicly available via those categorization practices. Finally, it includes analyzing instances in which speakers are held accountable for performing (or not performing) actions in ways that are consistent with category expectations. In this chapter, I analyze categorization practices in order to shed light on the ways in which expert and novice peer categories were reproduced, resisted, and transformed in our interview talk.                                                   37 Here the technical meaning of incumbent coincides with the lay meaning of the word as “one that occupies a particular position or place” (https://www.merriam-webster.com/dictionary/incumbent). The use of this term (as opposed to member) highlights the temporary and contingent nature of a given participant’s status (e.g., incumbents in political positions are usually appointed for a limited term).   138 6.3 Research categories  As Potter and Hepburn (2012) note, it is during research processes such as recruitment that “particular category memberships can be made central to the research” (p. 7). Thus, to present a more reflexive account of my research findings (Byrd Clark & Dervin, 2014), in this section, I briefly present the ways in which the study materials, and particularly the recruitment documents, deployed specific categories as well as the attributes and characteristics that were associated with those categories.  Two categories of particular note that were used to recruit participants for this study were study abroad students and peers or friends. For instance, the excerpt below comes from a study invitation sent to PWU peers. In it, the e-mail explicitly describes activities in which members of the study abroad student and peer categories are expected to engage:  My main focus is on how the interactions study abroad students have outside the classroom contribute to their English learning. So I’m interested in how your English conversations with study abroad students help them to reach their language learning goals. I also want to know what you learn from the conversations and the strategies you have when you communicate with study abroad students. (Original emphasis, Excerpt from study invitation sent to peers, see Appendix A)  A quick reading of this passage readily reveals that members of the study abroad category are attributed specific goals and motivations, including being involved in “English learning” and having “language learning goals.” This implies that the business of being a study abroad student is principally concerned with language learning and that their English language skills are in need of improvement. Conversely, peers (the category of participant to whom the letter is addressed) are assumed to have “strategies” and to be involved in “contributing” to SA students’ language development. Similarly, as the following excerpt shows, recruitment letters for JC peers construed interactions outside the classroom as primarily involving English learning and framed peers as potential “helpers.”  139 my main focus is on how the interactions you have outside the classroom contribute to your English learning. So I’m interested in how your English conversations with your friends, roommates, clubmates and other peers helps (or does not help) you reach your language learning goals. (Excerpt from follow-up e-mail sent to JC students, see Appendix A) In this sense, the categories of expert and novice were pre-assigned to participants at the outset of the research process, with novices (i.e., study abroad students) being construed as active English language learners and peers being constructed as supportive helper-experts.  This is not inconsequential. Likely because of the ways in which the recruitment materials and interview questions were designed, participants often oriented to understandings of study abroad students as principally engaged in language learning. However, as I highlighted in Chapter 5, JC peers indicated that many JC students did not prioritize goals related to learning English and also did not seek to improve their English through out-of-class interactions. Likewise, So-Yi, a PWU peer who would likely be categorized as a study abroad student from an institutional perspective, did not describe her interactions as opportunities for her to improve her English. Thus, the relevance of this pre-existing implicit expert-novice framework likely constrained the categories that were oriented to in talk. In the sections that follow, I analyze interview excerpts that reveal the ways in which novice and expert peer categories were co-constructed in situated ways in relation to English language expertise and other kinds of knowledge, and how these categorization practices served to accomplish specific actions in talk.  6.4 Novice peer identity categories in interviews As discussed above, during the data generation phase of this study, I most often categorized JC peers in the novice/language learner category by referring to them as study abroad students and by treating their ambitions to learn English as commonsense. However, in interviews, neither PWU peers nor JC peers used the term SA student. Rather, they tended to use alternative institutional categories such as exchange students or internationals as resources for 140 categorizing JC peers as language novices.38 In particular, in PWU interviews, these category-concepts and the characteristics discursively associated with them, served as resources for providing positive evaluations of the JC peers as exceptional members of these categories. Their accounts also demonstrate how the participants and I oriented to a number of commonsense assumptions about who exchange students are (e.g., from countries where a foreign language is spoken, language learners), the characteristics they should have, and the behaviours they are expected to display. These behaviours and characteristics are discussed in the sections that follow.  6.4.1 Lower English proficiency One key consequence of PWU peers’ mentions of novice categories such as international or exchange student is that such mentions generally made available an inference that category members had recognizably lower levels of English proficiency. In other words, as the two excerpts that follow will demonstrate, mentions of these categories served to indirectly topicalize JC peers’ (and others’) English abilities.  In Excerpt 6.1, Ed’s categorization of Lisa as exchange student is a resource used to indirectly index language proficiency. In this interview excerpt, I am asking Ed about how she met Lisa. Ed, an international and multilingual student from Singapore, moved into Lisa’s residence at the beginning of the second semester. Ed explains that she met Lisa for the first time                                                  38 It is important to note that participants did not refer to each other solely using the categories described in this chapter. In fact, participants most often described their relationships in terms of being simply “friends” or “roommates” (see discussion in Chapter 1). Participants also used labels such as “Japanese” or “Asian” in interviews to categorize JC peers; however, they overwhelmingly did so to highlight cultural or national rather than linguistic attributes. For example, So-Yi, Ami’s Korean roommate who was also at PWU on exchange, referred to Ami specifically as her “Japanese friend” (So-Yi, Peer Interview 1, 01/15/2016), pointing out that she had not had the opportunity to meet Japanese people in Korea. These alternate categorizations demonstrate the importance of studying how categories are used in situ to accomplish specific actions. 141 when Ed lost her key and Lisa opened the door for her. Note the explicit mention of the exchange student category in line 08.39  Excerpt 6.1 “I didn't realize she was an exchange student” 01 02   03 04 05 06 07 08 09 10 11 12 13 14 15 Victoria:    Ed: Victoria: Ed:    Victoria: Ed:    Victoria: Ed: and how did you: how did that first meeting go:  how was your first week. ((22 seconds omitted, Ed explains she lost her key and Lisa let her into the room)) so when I went i:n she was like really friendly:  ⸰mhm⸰ and she’s o:h like (0.3) “hi:” and everything   and then I sort of (0.4) cause sh- (.) um (0.3)  she didn- she was really good? I didn’t realize  she was an exchange student [sort of?]                                                                       [mm:     ] so (0.5) yeah and she was really friendly  and I was like “oh what program” and then (0.3)  she said she was studying English (.)  for her like exchange  ⸰mhm⸰ and yeah.  (Ed, Peer Interview 1, 1/21/2016, 6:37-7:29)                             Following a 22 second stretch of talk in which Ed tells the story of why she was knocking on her own residence door, Ed describes her first impressions when Lisa opened the door. She explains that Lisa was “like really friendly:” and said “‘hi:’ and everything,” at which point Ed self-repairs twice (sh- (.) um (0.3) she didn-) before finally arriving at the description that Lisa “was really good?.” Given the two self-repairs and the high rising terminal intonation, which can convey a sense of novelty and contrastive emphasis (e.g., Levon, 2016), this utterance can be heard as Ed’s initial and somewhat surprised evaluation of Lisa’s L2 proficiency. Ed then elaborates on her initial assessment explaining that at first she “didn’t realize” that Lisa “was an exchange student, sort of?.”                                                  39 Transcription conventions are listed in Chapter 4, Table 4.4.  142 Jayyusi (1984) explains that “[m]any requests for explanation are answered (and answerable) by the provision, invocation or attribution of a category incumbency to the person for whose action or behaviour an explanation was being sought” (p. 27). In this case, Ed’s provision of a category (i.e., exchange student) makes available an inference that normally exchange students would not be “really good” like Lisa is (line 07), their greetings would be less proficient, and this lack of proficiency would therefore make them recognizable as exchange students. By reporting Lisa’s explanation of “studying English for her exchange” in subsequent (lines 12-13), Ed’s account thus binds language proficiency as a predicate to the exchange student category and treats her initial evaluation (i.e., “really good”) as referring to language proficiency specifically. In addition, by mentioning that the purpose of the exchange is language study, Ed’s story also binds English learning as a predicate to the exchange student category.40  A notable aspect of Ed’s story is the difficulty she displays in producing her evaluation of Lisa’s English (lines 06-08). For most of the narrative (i.e., the omitted lines), Ed maintains the flow of talk (i.e., few hesitations); however, in line 06 when she begins to produce her assessment of Lisa she self-repairs twice, saying “cause sh- (.) um (0.3) she didn-.” Her hesitations combined with her use of the indirect reference to English abilities through the mention of the exchange student category may point to the sensitive nature of publicly evaluating a peer’s language skills. It may also indicate that performing an evaluation of a learner’s language ability in front of her (Lisa’s) English teacher (myself as the interviewer), is not a simple a matter. The fact that Lisa’s proficiency is discussed indirectly in the excerpt may also                                                  40 In topicalizing language proficiency in this way, Ed seems to be orienting to the overall topic of the interview (i.e., Lisa’ language learning). Indeed, the availability of the inference related to Lisa’s English ability may simply come from the omnirelevance of language ability and language learning as the main topics of the interview. 143 point to the sensitivity around discussing international students and English language norms at PWU more generally (see discussion in Chapter 2).  In Excerpt 6.2, it is the mention of the international student category (rather than the exchange student category) that serves to topicalize English language ability.41 In this excerpt, Paulisper, one of Ami’s PWU peers who was from central Canada, mentions the category international students when discussing the topic of “language barriers.” The excerpt begins as I ask Paulisper if there is anything she learned from Ami. Excerpt 6.2 “Mostly my friends are from America” Part 1 01 02 03 04 05 06 07 08 09 10 11 12 13 Victoria:   Paulisper:      Victoria: Paulisper:  Victoria: and is there anything that you feel like  you’re really learning from (.) Ami. (0.9) u:m (.)culture. (0.5) m yeah.  (1.0) >I think it’s just like< (.)  mostly with the: the who:le like I’m-  cause I also have like my roommate is also from u:m (0.4) well she’s from China  mhm but >well I mean she went to< an international school so he- her English is already pretty good but  mhm (Paulisper, Peer Interview 1, 24:04-25:43)  After a relatively long pause, Paulisper’s initial response in line 04 identifies “culture” as what she has learned from Ami. While this response may not be surprising since Paulisper had few Japanese friends in her circle, given that both peers took the same psychology class and loved Korean food, she could have just as easily identified knowledge of psychology or Korean                                                  41 The international student category was used explicitly to refer to JC peers by a several other PWU peers including Elizabeth (Samantha’s roommate) and John (Samantha’s language exchange partner) While, So-Yi and I also mentioned the term international student, we did so in reference to the membership device “institutional student status” (e.g., juxtaposed with domestic, and exchange), usually at the beginning of the interview, rather than in accounts concerning language use or learning. This points to Pomerantz and Mandlebaum’s (2005) argument that the social significance of categories and collections is not static but can be configured in different ways in different instances of talk. The label is also common one at PWU, where a large proportion of the student body is officially categorized as “international.” 144 restaurants as something she has learned from Ami. However, as the review in Chapter 2 noted, the notion of culture learning is intimately tied to encounters between SA students and host nationals and thus cultural difference is a predicate that is associated in this excerpt with members of the exchange student category.  Following another long pause, Paulisper begins to expand on her response saying “it’s just like” and “mostly with the whole like” but does so with significant difficulty and several false starts (lines 06-07) and it is unclear what “the whole” might refer to. At line 08, she says that her roommate is also from China. Her use of “also” seems to indicate that Paulisper initially treats her roommate from China and Ami as members of the same category. However, Paulisper quickly treats that categorization as problematic by identifying what distinguishes Ami from her roommate – her roommate’s English is already pretty good because she went to an international school (lines 11-12). Thus, Paulisper’s response treats Ami as having been potentially categorizable in the same category as her roommate, a non-native English speaker from China (i.e., an international student), but rejects this categorization on the grounds that her roommate’s English is pretty good, implying of course that Ami’s is not. In this way, Paulisper construes language proficiency as a defining predicate tied to particular kinds of international students, like Ami.  In line 13 and onwards (Excerpt 6.3 below), Paulisper’s account pursues this category contrast, categorizing members differently based on whether or not there is a “language barrier.” In the first few lines, Paulisper continues to elaborate on her account of her and Ami’s friendship.    145 Excerpt 6.3 “Mostly of my friends are from America” Part 2 14 15 16 17 18 19 20 21 22 21 22 23 24 25 26 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58  Paulisper:             Victoria: Paulisper:  Victoria: Paulisper:         Victoria: Paulisper:     Victoria: Paulisper:     Victoria: Paulisper:     Victoria: Paulisper:  I think it’s like (0.8) I think that like (.)  the thing is I feel like- actually about the friends thing? now that (.) I think about it (.)   u:m I think it’s it’s (0.3) it’s somehow like a little harder to make friends with u:m people who come from abroad like international students (.) like maybe not all of them but I still feel like sometimes maybe (.) it can be a little harder because (.) I think that the:y (.) like maybe they feel that there is a language barrier?  and like it’s so much easier to like it’s s- like it’s so easy to just talk to people (.)  c- when it’s like easy right?=  =mhm= =like and you just fall back on speaking English o:r (.) your native language right  =mhm= =so you’re more like likely to make friends with people who can speak your native language so then like (.) the full meaning of what you’re saying can like come across right? so: maybe like (0.6) I feel like ma:ybe some- cause I actually don’t have that many international like  (1.2)  when I say international mostly my friends are like from America?= =mhm. =so:  (0.8)  so like $I mean technically that’s international?$ but the:y they still speak like English and there’s no like language barrier there?= =mhm =so I think that sometimes like (0.5)  I feel like (0.6) maybe it’s (0.6)  it’s a little more difficult to like reach out to like people who are (0.5) who don’t like necessarily speak English as a native language maybe? mhm so I feel like maybe that like  (1.2)  maybe this will (.) will help (.)  ⸰ me have more friends⸰ (.) mm. (0.4) oh.= =in general.  (Paulisper, Peer Interview 1, 24:04-25:43) 146 The explicit mention of international students at line 09 binds the characteristics that came before (i.e., being from abroad and being hard to make friends with) to international student category. Although Paulisper’s utterances are hedged (see line 17, “somehow,” line 18 “a little harder,” and line 20, “not all,” line 21 “sometimes maybe”), Paulisper ultimately pursues her original categorization, saying that in most cases, it still is harder because “they feel that there is a language barrier,” thereby binding “language barriers” as a predicate of the international student category, as well. The remainder of this excerpt continues to reproduce the notion of communication difficulties as a predicate bound to international students and the notion of “easy” (lines 23-24) communication as a predicate bound to speakers of “your native language” (lines 28 and 31) (i.e., members of the native speaker category). The relevance of lower language proficiency as an accountable characteristic of the international student category is also evident in the way Paulisper takes special care to explain that Americans are only “technically” international (line 42), since there “is no language barrier there” (line 44) and “they still speak English” (lines 42-43). Paulisper’s account thus reproduces the ideological connection between the international student category and low language proficiency in multiple ways. It also explicitly evaluates incumbency in the international student category as a social disadvantage since category membership is discursively tied to a troublesome language barrier that apparently impedes members’ capacities for forming friendships.  While Paulisper’s explanation of the difficulties inherent in communicating with international students may seem somewhat belaboured, it is worth considering the social action that is accomplished via Paulisper’s categorization practices. Indeed, the apparent reason for her lengthy explanation does not come until the very end of the excerpt in lines 52-55, where Paulisper explains in a small voice that perhaps her friendship with Ami, as an incumbent of the 147 international student category, will allow her to expand her friendship circle, which currently consists mostly of Americans. In other words, Paulisper’s extensive categorization practices serve as an additional warrant for why Ami is a potentially valuable friend – because by pushing through the language barrier with Ami, Paulisper may learn how to “reach out” (line 55) and make friends with more international students. Her accounting work seems to suggest that because of the communication difficulties that are bound to membership in the international student category, her professed intention to make more international friends is not the norm for most incumbents of the native speaker category, who generally prefer “easy communication.” In sum then, Paulisper’s account constructs two category sets, each of which include two moral versions of category members: international students with vs. without language barriers, and native speakers who prefer easy communication vs. those who are willing to push through the language barrier. Paulisper’s categorization work therefore appears to reproduce local understandings of international students as novice English users and treats willingness to interact with them as a predicate bound to only a select number of members in the expert category. Like Ed’s account in Excerpt 6.1, Paulisper displays significant difficulty producing this account, which is filled with false starts and hedges, perhaps pointing once again to the sensitive nature of evaluating a peer’s language abilities or of producing negative evaluations of others in interviews.  6.4.2 Willingness, openness, and asking In addition to treating lower language proficiency as a commonsense attribute of members belonging to the novice category, PWU peers also constructed ideal members of these categories as students who were particularly open, outgoing, and willing to initiate conversation. As noted in Chapter 4, in many of the interview questions, I often asked PWU peers about the 148 strategies and success of JC peers as language learners. PWU peers obligingly produced accounts in which they described how language learners in an SA context should act, often using idealized moral accounts in their evaluations of the efficacy of JC peers’ language learning behaviours or strategies. For example, in their interviews, both Kyla and Serena (Lisa’s roommates) produced accounts in which they described what members of the exchange student category should and should not do to accomplish language learning. In this sense, they produced contrasting moral versions (Talmy, 2009) of exchange students in their evaluations of Lisa as a good exchange student. In this section, I present two excerpts in which PWU peers explicitly described the ideal behaviours and characteristics of good language learners in SA and demonstrate how these descriptions were used as resources to produce positive evaluations of JC peers.   In Excerpt 6.4, Kyla’s account establishes Lisa’s willingness to talk and initiate conversation as a predicate of the good exchange student category. Just prior to this excerpt, I had asked Kyla if she had any recommendations for incoming SA students. She replied that “talking to other people is definitely a key aspect” because “that’s the only way you can really improve” (Kyla, Interview 2). Kyla thus treats my question in lines 01-02 as a request to build on that characterization of what successful SA students do.  Excerpt 6.4 “Lisa kind of takes initiative”  01 02 03 04 05 06 07 08 09 10 11 12  Victoria:  Kyla:    Victoria: Kyla:      so what do you think makes Lisa so: successful (.)  in her efforts. I think the fact that’s she’s willing to do it?  like (.)she’s willing to like go out and have like conversation (0.4) random conversation with other people?  or just like kind of being [((inaudible))]                                                                                         [⸰mhm⸰         ]  I think that’s really good? cause like (.)  I’ve also heard of like (.) other (0.5) like people telling me that like maybe (0.3) an exchange student  >they would just< lock themselves up in their rooms  and like wouldn’t even say hi and stuff and like (0.4)  149 13 14 15 16 17 18 19 20 21 22 23 24  Victoria: Kyla:        Victoria:  that kind of discourages other people from trying to (.) talk to that person it’s [like]                                                                                       [⸰mhm⸰] you can try o:nce you can try twi:ce  but if they keep doing that you’ll just sto:p (.)  but like the fact that Lisa kind of takes initiative      >you’d be like< “okay yeah” she wants to talk  and we’ll like with her like we’ll like (0.4)  be able to like connect and stuff but like (.)  just being open-minded and just talk  like I think that’s the first step you need.  mhm  (Kyla, Interview 2, 4/5/2016, 34:28-35:14)  The initial part of Kyla’s answer (lines 03-05) is an account based on Lisa’s specific qualities – namely that Lisa is willing to talk to other people, which Kyla assesses as being “very good.” Kyla then contrasts Lisa’s behaviour with the behaviour of other hypothetical exchange students who would “lock themselves in their rooms” and not “say hi” (lines 11-12), thereby establishing these latter actions as predicates for this second, less social type of exchange student. Kyla then evaluates these behaviours as discouraging (line 13), thereby categorizing this group of exchange students as members of the negative or undesirable version of the category and highlighting the consequences for students who exhibit these behaviours (i.e., people will not interact with them anymore). After describing members of this unwilling exchange student category, Kyla proceeds in lines 18-23 to explain why Lisa cannot be categorized that way – Lisa “takes initiative” (line 18) and is therefore not treated in the same way as other exchange students who are perceived to display unwillingness. Kyla concludes her evaluation of Lisa by saying that “being open-minded” and “just talking” are the first steps to being a successful exchange student. This excerpt demonstrates how, by describing contrasting moral versions of good and bad exchange students, Kyla is able to produce an exceptionally positive evaluation of Lisa (her friend and roommate) as a successful exchange student.  150 Similarly, in Excerpt 6.5, So-Yi, Ami’s roommate, highlighted the importance of “trying” and not being ashamed to make mistakes when she described what she liked best about speaking with Ami. In line 01, I ask So-Yi what she likes best about speaking with Ami. Excerpt 6.5 “She’s not ashamed”  01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17  Victoria: So-Yi:      Victoria: So-Yi: Victoria:  So-Yi:  Victoria: So-Yi:  Victoria:   so (.)what do you like best ab-about speaking with Ami.  mm:  (0.9)  she is tr- she is trying to (0.5)  I guess she is trying to speak more fluently?  and (0.6) and she asks sometimes sometimes  asks me how to do this or do that  ⸰mhm⸰ and I really like that part? of her? so yeah.  mhm. so you like- why do you like it. (0.9) because like (0.3) she asks wha- asks to me?  because she knows I guess she knows me ⸰mhm⸰= =and she’s not ashamed about that?  and I really like that part. like yeah. ah. that’s great.  (So-Yi, Interview 2, 15/01/2016, 15:45-16:44) In lines 04-05, So-Yi explains that what she likes best about Ami is that “she is trying to speak more fluently.” So-Yi’s response, which topicalizes language proficiency by explicitly mentioning “fluency,” treats my line 01 question as being about Ami’s speech and the manner in which she uses English with So-Yi. By referring to “fluency,” her response also projects Ami’s incumbency in the language novice category. An interesting aspect of So-Yi’s response is that she explains that what she likes is that Ami is “trying.” She thereby emphasizes effort and intention rather than the quality of Ami’s English. After her initial evaluation (i.e., she’s trying to be more fluent), So-Yi expands on her response by providing an example of “trying,” explaining that Ami “asks her how to do this or that” (lines 06-07). The examples So-Yi provides treat Ami as having lower language proficiency in English (she is not yet fluent) and as actively engaged in 151 language learning (i.e., she is trying to become more fluent and she asks questions). When I follow up by asking “why do you like it” (i.e., the fact that Ami asks) in line 08, So-Yi explains that Ami is not ashamed of asking (line 15). “Not being ashamed” bears a striking resemblance to the predicates mentioned in Kyla’s account (Excerpt 6.4), particularly those of “being open” and “just talking.” As such, “not being ashamed” can also be heard as a predicate bound to the good exchange student category (and the good language learner category). Thus, although So-Yi does not explicitly mention the (good) exchange student category in this account, So-Yi’s response serves to categorize Ami as a member of that category.  An especially interesting feature of both this excerpt and Kyla’s (6.4) is the way in which they link attributes, or personality characteristics, of a good exchange student (e.g., being open-minded, willing, and unashamed) with particular actions or stances (e.g., taking initiative and asking).42 For instance, in her account, Kyla first describes Lisa as willing to initiate talk (“willingness” being a characteristic that is assumed about Lisa). She then describes other exchange students who would not even “say hi.” In this way Kyla’s account establishes the act of initiating conversation through greetings as an indicator of the willingness and open-mindedness that good exchange students must have to succeed in their language learning goals. Similarly, So-Yi first explains and positively evaluates Ami’s practice of asking questions (another way of initiating talk) and then discursively constructs that practice as an indicator that Ami is unashamed. Thus, via categorization practices, both Kyla and So-Yi’s evaluations of Lisa and Ami discursively establish a connection between invisible attributes such as effort (i.e., trying)                                                  42 Elements of these moral categories could be observed throughout the PWU peer interview data. Indeed, PWU peers often mentioned predicates associated with the positive version of this category, including being open, outgoing, and intentional about interacting in English with people on campus and negatively evaluated behaviours, such as interacting only with co-nationals or studying, rather than seizing opportunities to socialize in English. 152 and openness and the observable actions of “asking” and “initiating talk.” Gal (1998) refers to the process by which specific aspects of language use become associated with social groups as iconization. She explains that “linguistic differences that index social contrasts are reinterpreted as icons of the social contrasts” so that “the ideological representation fuses some quality of the linguistic feature and a supposedly parallel quality of the social group and understands one as the cause or the inherent, essential, explanation of the other” (Gal, 1998, p. 328). In these interview excerpts, there is evidence of iconization processes at work: exchange students who initiate talk or ask questions were ascribed the qualities of being willing and motivated students who took advantage of the opportunity to learn English through authentic interactions. The ideological connection established between asking and trying and willingness/desire to learn also appears to reproduce language ideologies that value experiential learning in immersive contexts (Doerr, 2015). This is an important point to which I return in the interpretation section at the end of the chapter.  6.4.3 Summary In PWU peer interviews, the categories international/exchange student constituted discursive resources for indirectly indexing lower English proficiency and were generally discursively tied to a desire to learn and to specific actions, such as “initiating talk” and “asking questions,” that members might take to improve their proficiency. An additional interesting aspect of categorization practices across these excerpts is that they were typically deployed for producing positive evaluations of JC peers’ English. Indeed, by using explicit category mentions like exchange student, participants were able to produce inferences that allowed me as the interviewer to assume JC peers’ lesser proficiency and to avoid explicit negative evaluations. Instead, participants’ accounts placed an emphasis on these JC peers’ willingness to learn. In 153 Table 6.1, I list the categories that were explicitly mentioned in reference to members of the novice category across PWU peer interviews as well as the predicates most commonly bound to members of those categories.  Table 6.1 Novice Peer Categories and Predicates Categories Predicates International student International friend Exchange student Exchange friend   • Come from countries other than Canada such as China and Japan (but not America or the UK) • Are non-native speakers with recognizably less proficient English  • Are in the process of improving or studying English • Are more difficult to communicate with  • Are confident/not ashamed to ask questions • Seek interaction, are willing, and open-minded   It is notable that the categories most often used to categorize JC peers as English novices (i.e., exchange and international students) are labels used by PWU to classify students for administrative purposes. When a category is clearly named, and particularly when it is institutionally-relevant like the categories discussed here, it is an indication that the category holds cultural relevance in the community in which it is used (Jayyusi, 1984). The criteria for claiming membership within institutionally-relevant categories are often shared to some extent across the community – in this case PWU. That this was the case is evident from the ways in which these labels were readily recognized by the participants and myself as referring to language ability, even when language ability was referenced only indirectly. It was only when international students were described as having higher proficiency (e.g., in the case of Lisa and Ed’s first meeting, or Paulisper’s American friends) that additional accounts to explain their level of proficiency were produced. In other words, exchange students, or novice peers, are expected 154 to have hearably/recognizably lower English language proficiency and to be involved in actively working towards improving that proficiency. These findings provide valuable insight into the ways in which PWU peers expected JC peers to behave as members of the novice peer category (i.e., displaying lesser proficiency and trying despite challenges) as well as some limited evidence of how such behaviours should be evaluated in this context (i.e., positively).  6.5 Expert peer identity categories in interviews As mentioned in Chapter 1, finding appropriate ways to refer to PWU peers was an ongoing challenge in this study. PWU peers spoke a wide variety of languages and had been exposed to English in different ways, making categorization by language ability a complicated prospect: some peers had learned English at home and were English-dominant (e.g., Paulisper and Elizabeth), others had learned English at school when they immigrated to an English-speaking country at a young age used different languages at home and at school (e.g., Serena, Ed, and Kyla), and yet others had learned English through language classes in their universities abroad and were dominant in a different first language (So-Yi, John, and Emma). The PWU peers that participated in this study also had a wide range of institutional statuses at the university – all were undergraduate students; however, some were international degree-seeking students, some were exchange students, and others were domestic students. In essence, there was no clear way to categorize PWU peers from an institutional, geographic, or linguistic perspective.  Following previous SA research on the centrality of the native speaker in SA, it might be reasonable to assume that the expert peer category would be closely connected to the notion of the native speaker (see Chapter 2) and that native speaker status might feature as a key predicate in JC peers’ categorization practices. Indeed, Paulisper’s interview talk (Excerpt 6.3) explicitly 155 bound predicates of language expertise, such as having no “language barriers” and “easy” communication, to the native speaker category (i.e., people who “speak English as a native language”, line 50). However, in many of the interviews including Paulisper’s, members of the native speaker category were not necessarily evaluated as good members of the expert peer category. As the data presented below will demonstrate, desirable moral versions of the peer expert category were more often constructed through references to peers’ willingness to overcome communication difficulties and to their previous experience with international students (i.e., learners with “language barriers”). In the sections that follow, I analyze the predicates that served to make members recognizable as good expert peers.  6.5.1 Experience with international students The first predicate mentioned in most JC peer interviews was related to acceptance of and experience with non-native speakers. For example, in Excerpt 6.6, Samantha describes how she prefers to speak English with people who “can accept her,” and who have gotten “used to the non-native speakers’ English” (lines 09-14).  Excerpt 6.6 “They accept me to speak”  01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 Victoria:   Samantha:       Victoria: Samantha:      so what kinds of situations do you like E-  speaking English the best.  (1.7)  mm::  (1.2)  ah::  (1.0)  I think that I realize is  ah people who knows that (0.4) mm: (0.9)  non-native speakers? mhm they can- they- I think (0.3) ah: (0.9.)  they accept me like (0.5) $not good$ ((laugh))  I’m not good at speaking English but they accept me to speak and they (0.3) hear me more like mm.  (2.0)  kindly like $I don’t know how to say but$ the mm::  156 18 19 20 21 22 23 24 25 26 27 28 29       Victoria: Samantha:   Victoria: Samantha:  some people doesn’t use- doesn’t get used to the:  non-native’s speakers English? so the:y I think that the:y (1.6) they also (0.4) mm had to hear the-  hear non-native speakers’ En[glish?]                             [mhm:  ] so I think (1.0) mm:: yeah it’s so (0.5)  depressing and I: thi:nk (0.8)  to speak with Elizabeth is so (0.5) comfortable?= =mhm.= =$and I also speak with you is$ so co- comfortable and (0.4) I think yeah      (Samantha, JC Interview 1, 1/29/2016, 42:33-43:47)  In lines 01-02, I ask Samantha about the situations in which she likes to speak English. After a long hesitation and several stretched fillers (lines 03-07), Samantha provides a response in which she describes the category of people with whom she likes to speak English. Her hesitant response treats my question as asking “who do you like to speak English with.” Specifically, the category of people with whom she enjoys speaking are “people who knows the non-native speakers” (lines 09-10) and who “accept her to speak” and hear her “kindly” (line 17). These characteristics can be heard as predicates of the good expert peer category since they refer specifically to members’ communicative abilities rather than say, their shared hobbies or interests. It is likely that Samantha categorizes peers in this way because of the general relevance of language and language learning in the interview and the specific mention of “speaking English,” which is treated here as “practicing English.” The rest of Samantha’s turn in lines 18-29 serves to warrant her claim that those who are “used to non-native speakers” is a relevant category for answering my question at line 01: because according to Samantha “some people” (line 18) are not used to non-native speakers. In line 20, after projecting this contrast between people who are and who are not used to non-native English, Samantha evaluates conversation with expert peers in this latter category as 157 “depressing.” Samantha then furnishes examples of incumbents in the good expert peer category – Elizabeth (her roommate) and myself – with whom conversations are described as “so comfortable” (lines 26/28). Through these evaluations, much in the same way that peers constructed idealized moral versions of exchange students, Samantha’s account also constructs contrasting moral versions of peer experts in terms their familiarity with non-native speech and their willingness to accept it.  A crucial aspect of this excerpt is the way in which the mentions of native/non-native categories are used to accomplish categorization in the good expert and good novice categories. By explicitly mentioning acceptance of non-native speakers, Samantha’s response makes relevant her own incumbency in the non-native speaker category. Her incumbency in that category is then further reinforced through her mentions of predicates associated with it, such as when she explains that “I’m not good at speaking English” (line 14) and produces hesitations and explicit unknowing stances (e.g., “I don’t know how to say,” line 17). As discussed in the previous section, displaying recognizably lower English proficiency is an accountable predicate of membership in various novice peer categories (e.g., exchange student). However, what is perhaps more interesting is that Samantha’s mention of non-native speakers also makes the contrasting category, native speaker, potentially available as a resource for categorization in this context. Both Elizabeth and I, who she mentions as examples of good experts at the end of the excerpt, are native speakers in the conventional sense (me of Canadian English and Elizabeth of British English). However, while the native speaker category may have some relevance here, Samantha’s account makes it clear that native speaker status is not the main criteria for inclusion 158 in the good peer expert category. Rather experience and kind acceptance are constructed as the defining attributes of members in the good version of this category.43  Excerpt 6.7 involving Ed, JC peer Lisa’s roommate, also explicitly references experience with English novice peers as a predicate bound to members of the good peer expert category. In this excerpt, I ask Ed about the challenges she faces when talking to Lisa and she replies that she is used to the way that Lisa speaks.  Excerpt 6.7 “Maybe cause I'm used to it”  01 02 03 04 05 07 08 09 10 11 12 13 14 15 16  Victoria Ed:  Ed:     Victoria: Ed: Victoria: Ed:   Victoria: is there any challenge when you’re talking to ah Lisa?=                                                                              =Lisa?  (0.6)  no: it’s just like (.)  some- like I don- maybe because I’m used to it?  cause I’ve had a lot of friends (0.4) whose visited? who (.) not like tour visited but who came to (0.7)  Bangladesh or Singapore  mm               just to learn English?                                                      mm: and um (0.4) when they did it sort of I sort of  >I don’t know< I got used to the way they’re talking so (.) whenever I talk to Lisa it’s sort of the same? mm. (Ed, Peer Interview 1, 01/21/2106, 17:26-17:53)  By topicalizing “challenges,” my initial question potentially projects Lisa’s incumbency in the novice category through reference to a predicate commonly bound to novice members (i.e., the language barriers mentioned in Paulisper’s excerpts, Excerpts 6.2 /6.3). In line 04, Ed answers the question by saying there was no challenge talking to Lisa. This initial “no” response does not seem to orient to the novice category potentially made relevant by my question. However, in the next lines, Ed extends her response, explaining that she has a lot of friends who                                                  43 Here it is worth mentioning that in JC peers’ course work, they were often asked to reflect critically on the label “native speaker” and that this may have also contributed to Samantha’s avoidance of the term. 159 came to learn English in Singapore or Bangladesh and that speaking English with Lisa is the same as speaking English with them. By explicitly mentioning predicates that make the exchange student category relevant (i.e., actively learning English), Ed’s utterances serve to project her friends’ membership in that category and treat my question as being about exchange students. In addition, by subsequently equating her interactions with those friends with her interactions with Lisa, Ed’s description also collects Lisa in that same exchange student category. Therefore, Ed’s categorization practices do orient to and reproduce Lisa’s incumbency in the novice category (thus as a person whose English is still developing and who could have challenges).  However, Ed’s response also does other important interactional work: by being dismissive of the potential challenges involved in interacting with Lisa, Ed also does “being accepting,” a predicate that was associated with the good peer expert category in Samantha’s excerpt (6.6). Ed’s story about her friends from Singapore, which is offered as a justification for why interactions with Lisa are not challenging, foregrounds Ed’s experience communicating with many different exchange students. Ed’s response implies that although there would normally be potential for communication difficulties (because Lisa is an exchange student), because of Ed’s experience, those challenges are mitigated. I respond to Ed’s story about her experience with an enthusiastic “mm:” at line 12, thereby treating Ed’s response as especially relevant and interesting. It would seem then that I both recognize and align with Ed’s warrant and treat it as projecting her incumbency in the good peer expert category. The upshot of our interactional work is that through complex categorization practices, Ed’s utterances successfully take up the categorizations made relevant in my question while simultaneously avoiding a negative assessment of Lisa (i.e., she does not topicalize challenges) and producing a positive version of Ed as an experienced and supportive communicator. Thus, these two excerpts 160 demonstrate that 1) JC peers produced positive evaluations of PWU peers by referring to their experience with language learners and their acceptance of non-native speech and 2) PWU peers sometimes “did” these predicates in interviews by producing positive and supportive evaluations of JC peers that minimized their language difficulties.   6.5.2 Friendliness An additional categorization resource that was used in connection with members of the good expert peer was the explicit mention of JC peers as friends. Following work in MCA on standard relational pairs (e.g., Sacks, 1992), the category friend is typically understood as one part of a standard relational pair, that of friend-friend. In this pair, the mention of a third party as a friend conventionally generates the inference that the speaker is also an incumbent of the friend category (i.e., if Lisa is Ed’s friend, Ed is also Lisa’s friend). For instance, in Ed’s excerpt (6.7), she not only categorized the students from Singapore and Bangladesh as novices, but also referred to them as “friends,” making her membership in the friend category an additional resource potentially available for interaction. Indeed, PWU peers regularly referred to JC peers as their friends even though in the lay sense, their relationships (especially Ami’s and Samantha’s relationships) often did not conform to conventional norms of friendship (e.g., seeing each other for social reasons, shared interests). Instead, the category friend (and the associated predicate of friendliness) appeared to serve as a discursive resource for generating inferences related to what good friends usually do, such as being friendly, supportive, and helpful.  For example, in Excerpt 6.8, Elizabeth utters the friend category twice when describing why she has agreed to participate in the study, first when referring to her many international friends and again when referring to Samantha, her roommate and a JC peer. This excerpt begins 161 as I am pursuing the reason why Elizabeth agreed to participate in the study when her other roommates did not.    Excerpt 6.8 “A lot of my friends are international”  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Victoria:   Elizabeth:    Victoria: Elizabeth:       Victoria:  Elizabeth:  Victoria: Elizabeth: Victoria: and so what made you want to agree to do the project. (1.3) yeah like (0.5) it sounded interesti:ng  >like quite a lot of< my friends are all like internationals [like]                                                                                     [mm  ] kinda like um (0.5) China: Japa:n (0.5)  so yea:h.  (0.8)  I just >so yeah I thought it would be< interesting to kind of watch her English  just impro:ve. (1.1) so you’re kind of curious about he:r English progress  yea:h and like (0.7) and like she’s my friend so like I want to help her out too. Yeah=  =yea:h. awesome. (Elizabeth, Interview 1, 01/27/2016, 7:40-8:08)  After a long pause in line 02, in lines 04-09, Elizabeth initially answers my questions by explaining that the project seems interesting because she has a lot of international friends from places like China and Japan. Here, much like in Ed’s excerpt (6.7), Elizabeth responds by describing a particular category of friend that is relevant to the study – international people from countries in Asia whose English needs improving. She thereby treats the research project as one that involves members of that category and projects Samantha’s incumbency as a novice.  162 Elizabeth’s utterance also accomplishes an additional social action: it provides a warrant for why she is a good person to participate in the study. As previously observed, experience with novice peers was routinely bound as a predicate of the good peer expert category; thus Elizabeth’s mention of her international friends makes relevant her incumbency in the good peer category and serves as a rationale for her suitability as a research participant. She concludes this line of thought at line 09 saying “yeah” with falling intonation. I pass my turn and following an almost one second pause, Elizabeth provides an additional rationale for her participation, saying that it would be interesting to watch Samantha’s English improve. Elizabeth’s statement implies that she would be able to observe and recognize Samantha’s language development, projecting the potential relevance of English expertise for understanding Elizabeth’s interest in the study. Indeed, I treat her answer as making expertise relevant by requesting clarification of whether she is “curious about her English progress” (a predicate bound to language expertise). While Elizabeth minimally aligns with my question at line 17, her “yea:h” is produced with continuing intonation and a “and like,” treating my recast as incomplete. She then specifies that she is also participating because Samantha is her friend. Here, the mention of the friend category, provided as one reason for participation, generates inferences related to what good friends normally do for their friends – in this case, as she explicitly states, they help each other when help is needed. That she is successful in projecting a positive image of herself as a good friend is evident in the way I assess her account using an upgraded positive assessment (“awesome,” line 19). Thus, in this excerpt, it was not until Elizabeth produced a rationale based on friendship that I align with her and evaluate her positively.  Similar to other excerpts, Elizabeth’s response manages to recognize a JC peer’s incumbency in the novice category but does so in a very positive way, topicalizing Samantha’s 163 potential improvement and the value and interest of having friends like Samantha. In addition, even though Elizabeth’s incumbency in the English expert category is made relevant in the excerpt, her expertise is minimized through references to her additional/complementary role in Samantha’s life as a friend. This minimization of the differences and difficulties between JC peers and PWU peers was evident across all the interview data and points to the delicate negotiations required to “do language learning” in supposedly symmetrical and informal interactions with peers.  6.5.3 Summary In these excerpts, I have shown how both PWU and JC peers routinely used categorization practices to establish the characteristics of good expert peers in this context. Predicates most often tied to the expert peer category involved previous experience with international or exchange students, being used to the speech of language learners or non-natives, and being accepting, kind, and willing to interact with international students. These characteristics are summarized in Table 6.2. Table 6.2 Expert Peer Categories and Predicates  Category Predicates Friend  • Speak relatively more proficient English  • Have novice friends • Are used to less proficient English • Understand the difficulties faced by novices  • Accept communication difficulties with novices  • Are friendly  An interesting feature of these interviews was that expert peers who were assessed positively by JC peers were reported to have special experience not possessed by default by all expert English-speaking peers. Ed, for example, described her experience with student visitors in Singapore to legitimate her claim that she could understand Lisa’s English. In her excerpt, 164 Samantha explained that some people “had to get used to non-native English,” implying that people are not naturally able to understand non-native English but must accumulate different experience to do so. Even Elizabeth accounted for interest in interacting with Samantha as a product of her having a lot of international friends. Thus, the good expert peers were not treated as the norm but as an exceptional or special category of peer. Being a good expert peer was also projected as harder and requiring more effort. Paulisper’s (Excerpts 6.2/6.3) descriptions of the difficulties related to language barriers are the clearest example of this. Her account depicts a world in which native-speaker interaction is easier and more naturally conducive to sharing feelings while interacting with internationals requires extra experience and reaching out. Thus, being a good peer for international students was construed by PWU peers as well as JC peers as a learned practice or disposition rather than as dependent on language proficiency.  Performing or claiming membership in the good peer expert category was generally evaluated positively by me and the participants, for example when I reacted to Elizabeth’s offer to help Samantha as “awesome” (Excerpt 6.4, line 14). PWU and JC peers also constructed people who accept and are willing to interact with internationals as especially kind, helpful, and accepting. This can be seen in Samantha’s account, for example, in which she praises her PWU peers for accepting her even though her English “isn’t really good” (Excerpt 6.6, lines 05-06). In other words, in these accounts, expert peers are people who go above and beyond what would normally be expected in interaction because they are both more kind and more experienced.  6.6 Interpretation Categories and categorization are fundamentally ideological: they draw on circulating commonsense notions and produce and reproduce local moral orders in situated ways (Jayyusi, 1984). Likewise, the novice and expert peer categories in this context predictably drew upon and 165 reproduced several aspects of the language ideologies and assumptions apparent in JC peers’ stories, in particular ideologies related to immersion and to English as a global language (see Chapter 2 and Chapter 5 for more discussion of these).  6.6.1 Good novices as experiential learners Firstly, predicates bound to the novice category, and particularly the “good” version of the novice category, reproduced ideologies of immersion in several key ways. As discussed in Chapter 2, ideologies of immersion emphasize the importance of contact between SA students and the host community, and construe interactions with host members as a privileged site of TL learning. This ideology is clearly reproduced in the ways in which “asking,” “taking initiative,” “saying hi,” and “just talking” were bound as predicates to a good novice peer. If the goal of language learning in SA is for the student to maximize their exposure to the TL, then students who seek out TL contact by starting conversations and asking questions would naturally be praised as conforming to the expected ways in which a student should seek to benefit from an SA experience. In addition, as Doerr (2015) explains, discourses of immersion prioritize experiential learning through mundane encounters and operationalize otherness, difference, and dissonance as necessary conditions for valued forms of learning in SA. Thus, the predicates surrounding openness, trying, and initiative that were mentioned in the interviews also reproduce assumptions about the value of experiential learning by positively evaluating students’ risk taking and willingness to experience difficulty without being “ashamed” (Excerpt 6.5). 6.6.2 Good experts as users of English as a global language The second category of ideologies that were reproduced through peers’ categorization practices were those related to ideologies of English as a global language. These ideologies tend to disrupt the taken-for-granted association between English and inner circle countries (e.g., UK, 166 US, Canada…) and thus by association, between English and native speakers (De Costa, 2010). Instead, English is viewed, much as JC peers described it in this study, as the language of the mobile and multilingual elite (De Costa, 2014). Thus, learning English is associated with a desire to interact with a variety of cultural others, to travel, and to obtain elite jobs. Ideologies related to English as a global language were most clearly reproduced through peers’ predication practices, and specifically the binding of prior experience with international students to the “good” version of the expert peer category. As the analyses demonstrated, the two moral versions of the expert peer category relied primarily on those members’ experience or lack of experience with international others. Indeed, “doing being an expert peer” in these interviews involved being open and accepting of different types of English (and especially non-native English). As Samantha’s excerpt showed (Excerpt, 6.6), JC peers evaluated attributes related to acceptance and experience as characteristics bound to suitable peers with whom they enjoyed interacting. What this suggests is that experts in this context may be profitably understood as accomplished international or intercultural users of English. Importantly, membership in the good expert peer category did not exclude native speakers, as Samantha and Paulisper’s excerpts (Excerpts 6.6 and 6.4) suggest, but rather included only those native speakers with whom JC students could feel comfortable and accepted and who made a special effort to overcome the challenges or communication barriers bound to members of the novice category.  Thinking about PWU peers as experts of international or intercultural English substantially shifts the kinds of learning outcomes and experiences that one might expect in SA. It means that although students are physically in Canada, an English-dominant country, in practice, JC peers formed sustainable relationships predominantly within international and intercultural communities. While that in itself is not unusual (see Chapter 2 and 5), the fact that 167 JC peers valued and sought out participation in that community, in some cases over “authentic” Canadian communities, is not insignificant. This finding leads to important questions about the linguistic practices and values into which JC peers are socialized through participation in such communities and suggests that Canadian English native speaker norms are likely not highly relevant for these learners.  The relationship between the categorization practices observed in the interviews and the ideologies present in JC peers’ stories can be usefully mapped onto the Douglas Fir Group (2016) framework. In this model, the language ideologies identified at the macro level interact in complex ways with the construction of local expert/novice categories at the meso level.  Figure 6.1 Relationship between language ideologies and categories. Model adapted from the Douglas Fir Group (2016).  168 6.7 Conclusion In this chapter, I examined how PWU peers, JC peers, and I engaged in categorization practices related to expert and novice peers. I demonstrated how participants’ accounts of their experiences, motivations, and preferences drew on category resources when accomplishing social actions in talk. I highlighted how peers’ categorization practices often served to produce positive evaluations of JC peers as good novices. I also explained how the predicates bound to good novice, such as being open, initiating conversation, and having lower language proficiency, reproduced common assumptions about language learning in SA. Finally, I noted how predicates bound to the good peer expert category were related to experience with novices, as well as acceptance and willingness, and suggested that expertise in this context could usefully be understood as expertise in international and intercultural uses of English. In the following chapter, I build on these analyses by focusing specifically on how JC and PWU peers described their opinions and experiences related to talk-about-language.  169 Chapter 7: Rights and responsibilities in talk-about-language 7.1 Introduction In the previous chapter, I discussed PWU and JC peers’ categorization practices in interviews focusing on the predicates bound to expert and novice peer categories in participants’ reports of their interactions with and impressions of each other. In particular, I highlighted how novice category members were evaluated positively if they demonstrated “willingness” and “openness” by seeking out and initiating conversation. I also pointed to how expert peers who had experience interacting with international people (i.e., English learners) were assessed as being particularly well suited to interacting with novice peers. In this chapter, I further explore how PWU peers and I co-constructed the responsibilities of novice and expert peers in reports of their engagement in talk-about-language. I focus on the ways in which PWU peers and I oriented to norms around how talk-about-language was managed and who, the novice or expert, was responsible for initiating it. This chapter, therefore, provides preliminary answers to the second research question:  What role (if any) does talk-about-language play in interaction, how is it initiated, and how is it discursively managed?  To answer this question, I draw on the notion of deontic authority, which refers to “the capacity of an individual to determine action” in talk (Stevanovic, 2018, p. 375). Deontic authority, and the ways in which it is managed in interaction, are introduced in the first section of the chapter. In the sections that follow, I proceed with a detailed analysis of PWU peers’ descriptions of their talk-about-language practices in interviews. As the analyses will demonstrate, PWU peers’ reports attributed the responsibility of initiating talk-about-language 170 (through requests for help) to members of the novice peer category and treated unsolicited correction and “teaching” as accountable actions for members of the expert peer category.   7.2 Focus on the discursive co-construction of deontic authority As described in Chapter 3 (section, deontic stances, such as initiating a new topic, requesting an action, or giving an order, have to do with the ways in which participants’ uses of discursive resources determine a particular course of action in talk. Stevanovic and Peräkylä (Stevanovic, 2018; Stevanovic & Peräkylä, 2014) claim that, in much the same way that speakers monitor what another speaker can be assumed to know (i.e., epistemic order), another key facet of accomplishing interaction involves monitoring which speaker has the right to produce different deontic stances in a given moment (e.g., to commit to an action, to request an action, or to express a desired action). Stevanovic and Peräkylä (2012) explain that the distribution of deontic rights in a given instance of interaction is related to local understandings of participants’ deontic status – or their relative rights and authority to determine action as compared to other co-present interlocutors. In other words, participants’ treatment of each other as having the right to engage in deontic stancetaking relies on the social categories that are made locally relevant in talk as well as participants’ shared understandings of the deontic authority conventionally associated with each of those categories (Antaki, 2012; Clifton, Van De Mieroop, Sehgal, & Aneet, 2018; Landmark, Gulbrandsen, & Svennevig, 2015; Lindström & Weatherall, 2015). In this sense, various forms of deontic authority can be usefully understood as predicates bound to membership in different social categories.   For example, Landmark et al. (2015) explored the distribution of deontic rights and authority in medical consultations in Norway. They analyzed interactions in which doctors gave patients the choice between two treatment options (invasive and less/non-invasive). They explain 171 that the right to make the final treatment decision is as an important deontic right in this context that is necessarily negotiated in interaction between doctors and patients. They note that while, traditionally, doctors had been considered to have deontic authority based on their medical knowledge (i.e., in this case, the authority to decide on the treatment plan), more recent policy recommended allowing patients to have more authority over treatment decisions. In their analyses of actual doctor-patient interactions, they found that in some cases, participants were able to achieve a relatively symmetrical distribution of deontic rights, or a shallow deontic gradient. In these interactions, doctors often presented treatment options using modal verbs (e.g.,  you can) instead of imperative statements such as (e.g., you have to), which “may constrain the patient’s deontic rights to varying degrees” (p. 56). However, as the authors point out, even though doctors typically indicated their preference for one treatment over another (i.e., they effectively produced a deontic stance in which they recommended one of the treatments), patients sometimes resisted the doctor’s deontic authority. Thus, Landmark et al. (2015) note that deontic authority is not primarily achieved through one-sided claims but is achieved through others’ acceptance of deontic stances. Thus, deontic authority is best understood as a member’s matter that relies on socially and culturally available knowledge displays about the rights and responsibilities of category incumbents. One major difference between my approach to analyzing deontic authority and that of Stevanovic and Peräkylä (2014) and Landmark et al. (2015) is that rather than inferring deontic authority based on the ways in which participants treat each other’s utterances in talk, in the analyses that follow, I examine how deontic authority is explicitly constructed and discursively tied to expert and novice categories in PWU peers’ stories involving talk-about-language. In other words, in this chapter, I examine norms, or ideologies, related to the ways in which deontic 172 authority is distributed in this context rather than actual displays of deontic stancetaking. Like Landmark, Gulbrandsen, and Svennevig (2015), who focussed on the ways in which a particular deontic right was negotiated (i.e., the right to decide on treatment), in this chapter, I focus on the right to introduce a new topic of talk related to language – in other words, I examine how PWU peers and I co-constructed participants’ rights and responsibilities to initiate talk-about-language.   7.3 Deontic rights in interview questions Before proceeding to analysis of PWU peers’ accounts, it is worth considering the distribution of deontic rights that I assumed at the outset of my study. Initially, I was principally interested in understanding the role of peer interaction in JC peers’ English language learning and in whether or not PWU peers addressed language and language learning with JC peers explicitly in their talk. In other words, I intended to investigate the extent to which PWU peers initiated, controlled, or took responsibility for JC peers’ language learning. Therefore, in the interviews for this project, I explicitly asked PWU peers to describe how they “helped” JC peers with their English via the following questions (for a complete list of questions, see Appendix B): • Do you ever try to help (JC student’s name) with his/her English?  How? • What helped (JC student’s name) to improve their English, in your opinion? • Did conversations with you help (JC student’s name) improve? How?   (Questions from interview protocols for PWU peers) By referring to PWU peers as people who “help” and JC peers and people who “improve,” these questions projected the interactional space for PWU peers to produce accounts in which they described their deontic rights and responsibilities with regard to JC peers’ language learning.  Indeed, in response to these questions, I fully expected that peers would list helping strategies or stories in which they taught JC peers something about language (e.g., new expressions, grammar corrections). Instead, PWU peers’ responses tended to take the form of 173 nuanced and complex explanations in which the