Style Shifting and Social Network Development During Education Abroad Programs in Japan Dissertation Presented in Partial Fulfil

Style Shifting and Social Network Development during Education Abroad Programs in Japan

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the

Graduate School of The Ohio State University

Hiromi Tobaru

Graduate Program in East Asian Languages and Literatures

The Ohio State University

2019

Dissertation Committee

Dr. Mari Noda, Advisor

Dr. Xiaobin Jian

Dr. Mineharu Nakayama

Dr. Charles J. Quinn

Copyrighted by

Hiromi Tobaru

2019

Abstract

This dissertation explores ways that home institutions —educational institutions that send their students to affiliated universities— can provide pedagogical support to maximize the learning experience of U.S. undergraduate students during yearlong education abroad (EdA) programs in Japan. The results of this dissertation suggest several features that are key to pre-

EdA training. The most crucial is repeated experiential practice on inter-personal negotiations toward co-constructing a meaningful Third Space1 (Jian & Walker, 2017) that entails raising sensitivity to style-shifting strategies. Group, rather than individualized, format better maximizes the opportunities for negotiation. Also useful would be opportunities to hear the experience of students who have just returned from a yearlong EdA in Japan, focusing on their experience (or lack) of network building. The need for repeated practice entails that training occurs over a duration of time, such as a semester.

The first phase of this research examined difficulties in building intercultural relationships that American students and local Japanese students experience in a short-term EdA program in Japan. Data collected through interviews and observation suggest that the two groups have gaps in their expectations about speech style when interacting with each other. In particular,

1 An intercultural communicative context where norms or rules of the game do not conform entirely to either first culture or second culture, but will be dynamic and conditioned by shared iii

U.S. students struggled in shifting their speech from formal to casual. Based on these findings a pre-departure training regimen was devised for students participating in a yearlong EdA program in Japan. The training focused on style-shifting skills to develop a relationship-building persona in a Japanese college setting.

The second phase of the study investigated the outcomes and process of learning in yearlong EdA exchange programs. Of the seven undergraduate students who participated in this phase of the study, four went through the pre-EdA training, while the other three chose not to.

Two proficiency tests and a microanalysis of their discourse in terms of style shifting during proficiency interviews were used to assess the outcomes of all participants’ EdA gain. The process of their EdA learning was examined by analyzing their language use logs and interviews about social network development. The effectiveness of the EdA training was measured by comparing the findings of the two groups, i.e. the EdA training group and the non-EdA training group.

Findings suggested that the EdA training did not have a great impact on the participants’ outcome of EdA learning as measured by the proficiency tests. However, a closer examination of the learning process revealed that those participants who faced and endeavored to overcome challenges in building their social networks with locals demonstrated more EdA gain than those who did not. These students had to negotiate relationships continuously with interlocutors from the host community. Repeated negotiation experiences may have facilitated these EdA students’ learning of how to communicate effectively in Third Spaces. In the current study, such social ties are called negotiated Third Space. The findings also suggested that the structure of EdA

programs at host universities (e.g., roles and culture of dorms, school events, language programs, etc.) had the greatest impact on constructing negotiated Third Spaces.

In order to maximize EdA learning in Japan, it is crucial to support EdA students’ negotiation skills. EdA students can utilize style shifting as a communication strategy to recognize, negotiate, and co-construct expectations with locals. Such skills also help EdA students develop a relationship-building persona, which enables them to forge intercultural relationships with locals, which in turn maximizes their EdA learning in Japan. In addition, understanding each EdA program from the students’ perspectives is also crucial to maximize

EdA learning in general. Home institutions should use the returnees as a resource for those who will participate in the same EdA program.

Dedication

This dissertation is dedicated to my parents, Toshinori and Mutsuko Tobaru, my husband,

William Ma, and my daughter Miya.

Acknowledgements

I have received support and encouragement from a great number of individuals without whom this dissertation would not be possible. I first would like to express my sincerest gratitude to my advisor, Professor Mari Noda, whom I first met seven years ago at a teacher-training program.

Since then I have benefited and learned from her immense knowledge on Japanese language pedagogy, and I am very grateful to have her as my advisor, mentor, and role model. I cannot thank her enough for the guidance and support she has provided me for the last seven years.

Without her, I would never have gotten to where I am right now. I am also indebted to the members of my dissertation committee. I am grateful for a number of discussions I had with

Professor Mineharu Nakayama on various topics covering from my research to parenting. He also helped me with data collection for the first phase of the research. Professor Xiao-bin Jian who first introduced me to the concept of Third Space and provided me valuable insights on researching and teaching in a study abroad context based on his extensive experience directing study abroad programs. Professor Charles Quinn taught me Japanese language and culture from different perspectives, and I would have never learned those perspectives if I hadn’t taken his classes. My sincere thanks also go to Professor Hiroaki Kawamura for providing continuous

vii

guidance and support for studying, teaching, and parenting even after graduating from University of Findlay.

I also would like to thank the participants in the study. I owe especially Ms. Makiko Itai for helping me with data collection in the first phase of the study.

I am also thankful to Ms. Ai Terada, for her continuous support and encouragements as my supervisor as well as mama-tomo. I also would like to express my gratitude to Ms. Yuko

Kuwai and Mr. Teppei Kiyosue, who always accommodated my teaching schedule despite my limited availability. My thanks also go to my friends at DEALL for making my time at DEALL stimulating and memorable to me inside and outside of classes, especially Shun Maruyama and

Kumiko Takizawa for helping me practice my oral defense, Yeri McClain, Hannah Dahlberg-

Dodd, and Lindsey Stirek for being awesome babysitters for my daughter while being in school.

Kayo Puthawala for the valuable time we shared talking about raising daughters while being in school. I’d also like to thank Lindsey Stirek and Danielle Rymers, who proofread my dissertation drafts.

Finally, my greatest thanks go to my family, especially my parents for providing me unconditional love and support throughout my life, my husband, Will, for being patient during my dissertation writing and my daughter, Miya for being inspiration for my life.

viii

Vita

September 8, 1988……………………………………...Born – Okinawa, Japan

2012…………………………………………………….B.A. Foreign Language Studies,

University of Kitakyushu

2014………………………………….…………………M.A. Teaching English to Speakers of

Other Languages/Bilingual Education,

University of Findlay

Publications

Tobaru, H. (2019). Understanding the Difficulties in Building Intercultural Relationships from Perspectives of American Students and Japanese Students during a Short-term Study Abroad in Japan. Journal of the National Council of Less Commonly Taught Languages. 25(1), pp. 109-131. Tobaru, H. (2014). Socialization and Language Acquisition: Cases of Novice-Level Study Abroad Participants. Himeji Dokkyo University Journal of Teacher Education Research, 24, pp.209-230 Field of Study

Major Field: East Asian Languages and Literatures

Areas of specialization: Japanese language pedagogy

Table of Contents

Abstract ...... iii Dedication ...... vi Acknowledgements ...... vii Vita ...... ix Table of Contents ...... x List of Tables ...... xv List of Figures ...... xvi List of Aberrations ...... xvii Chapter 1: What We Know and What We Do Not Know about Education Abroad Learning in Japan ...... 1 1.1. Introduction ...... 1 1.2. Statement of problems ...... 3 1.2.1. Problem 1: Pedagogical support for EdA programs in Japan ...... 3 1.2.2. Problem 2: Measurement of EdA learning ...... 5 1.2.3. Problem 3: Outcome vs. process in EdA learning ...... 8 1.3. What makes intercultural communication difficult and complex? ...... 9 1.4. Conceptual framework: Third Space in the intercultural communication ...... 12 1.4.1. Co-constructing a constructive Third Space ...... 14 1.5. Purposes of the current study and definition of EdA gain ...... 15 1.6. Organization of the research and research questions ...... 16 1.7. Significance and constraints of the dissertation ...... 19

Chapter 2: What Happens during the Education Abroad Experience and How Can We Guide Students’ Learning? ...... 23 2.1. Introduction ...... 23 2.2. Style shifting as a measurement of EdA gain in Japan ...... 24 2.2.1. Definitions of forms in Japanese ...... 24 2.2.2. Style shifting by Japanese L1 speakers ...... 27 2.2.3. Style shifting by Japanese L2 speakers ...... 33 2.3. Social networks as a means of understanding the process of EdA gain ...... 38 2.3.1. Social networks as a strong predictor of EdA gain ...... 39 2.3.2. What is happening in an L2 learner’s intercultural communication during EdA programs? ...... 42 2.4. Culture in L2 learning and teaching ...... 47 2.4.1. Developing personae in C2 and L2 learning ...... 50 2.4.2. Performed Culture Approach as a framework of EdA training ...... 52 2.5. Conclusion ...... 57 Chapter 3: Understanding the Difficulties in Building Intercultural Relationships in an Education Abroad Context ...... 58 3.1. Introduction ...... 58 3.2. Methodology ...... 59 3.2.1. Setting ...... 59 3.2.2. Participants ...... 60 3.2.3. Instrument, procedure and data analysis ...... 62 3.3. Findings and discussion ...... 64 3.3.1 Shared perspectives toward the relationships ...... 64 3.3.2. Research question 1: What kinds of difficulties do local Japanese students encounter when interacting with American EdA students? ...... 67 3.3.3. Research Question 2: What kinds of linguistic and cultural difficulties do American EdA students encounter? ...... 74

3.3.4. Research Question 3: What kind of Third Space emerges when American EdA students and Japanese local students interacting with each other? ...... 78 3.4. Pedagogical Implications ...... 81 3.4.1. Pre-departure training for American EdA students ...... 81 3.4.2. Training for Japanese students ...... 87 3.4.3. Pedagogical implications for the design of EdA programs ...... 88 3.5. Conclusion of Chapter Three (The First Phase) ...... 89 Chapter 4: Methodology for the Second Phase of the Research ...... 90 4.1. Introduction ...... 90 4.2. Setting: Home institution and host institutions ...... 90 4.3. Participants ...... 94 4.4. Instruments ...... 95 4.4.1. Instrument of EdA gain ...... 95 4.4.2. Instruments for measuring social networks ...... 101 4.5. EdA Training ...... 106 4.5.1. Pre-departure training ...... 107 4.5.2. Regular Online Meetings ...... 108 4.6. Procedure ...... 109 4.7. Data Analysis ...... 110 4.7.1. Analyses of EdA gain: Style shifting ...... 110 4.7.2. Analysis of social network ...... 114 Chapter 5: Understanding the Outcomes of Education Abroad Learning in Japan ...... 119 5.1. Introduction ...... 119 5.2. The participants ...... 120 5.3. EdA gain ...... 122 5.3.1. EdA gain demonstrated by J-CAT ...... 122 5.3.2. EdA gain demonstrated with informal OPI tests ...... 125 5.3.3. EdA gain demonstrated with the examination of style shifting ...... 128 5.4. Discussion ...... 160

xii

5.4.1. What did the participants gain or not gain during EdA programs in Japan? ...... 160 5.4.2. Are there any differences in EdA gain between the participants in the EdA training group and the non-EdA training groups? If so, how? ...... 167 Chapter 6: Understanding the Processes of Education Abroad Learning in Japan ...... 170 6.1. Introduction ...... 170 6.2. Findings ...... 171 6.2.1. Cathy (High-gainer) ...... 172 6.2.2. Frank (High-gainer) ...... 177 6.2.3. Isabelle (High-gainer) ...... 181 6.2.4. Henry (High-gainer) ...... 184 6.2.5. Emma (Low-gainer) ...... 189 6.2.6. Bobby (Low-gainer) ...... 195 6.2.7. Anna (Low-gainer) ...... 200 6.3. Discussion ...... 204 6.3.1. What are the similarities and differences in the seven US students’ social networks and EdA experiences? ...... 204 6.3.2. In what way did social networking influence the participants’ EdA gain? ...... 208 6.3.3. Are there any differences in EdA social networks between the participants in the EdA training group and non-EdA training groups? If so, how? ...... 210 6.3.4. Negotiated Third Spaces during EdA experience in Japan ...... 211 Chapter 7: Maximizing Education Abroad Learning in Japan ...... 218 7.1. Summary of the objectives and the findings of the research ...... 218 7.2. Home institution training for EdA participants in Japan ...... 221 7.2.1 Pre-departure training duration and content, cast of characters: Focusing on negotiation skills and the process of EdA learning ...... 222 7.2.2. Concurrent-EdA training with the utilization of technology ...... 226 7.2.3. Post-EdA training ...... 227 7.3. Limitations and future research implications ...... 228 7.3.1. Style shifting as a means of understanding EdA gain ...... 228

xiii

7.3.2. Social networks as a means of understanding the process of EdA learning ...... 230 7.2.4. Representing diversity in EdA learning ...... 231 7.4. Conclusion ...... 232 Bibliography ...... 233 Appendix A. Phase 1: Interview Questions with EdA students (English) ...... 240 Appendix B. Phase 1: Interview Questions with local Japanese students ...... 241 Appendix C. Phase 2: Pre-EdA Interview Questions with EdA Exchange Students ...... 243 Appendix D. Phase 2: Post-EdA Interview Questions with EdA Exchange Students ...... 244 Appendix E. Study Abroad Social Interaction Questionnaire (SASIQ) ...... 245 Appendix F. Language Use Log Instructions ...... 256 Appendix G. Instructions of pre-EdA training role-play ...... 259

xiv

List of Tables

Table 1.Plain forms and masu forms in imperfective forms ...... 25 Table 2. Types of utterance ending by Japanese L1 students and L2 students' utterance (Masuda, 2010, p. 195) ...... 34 Table 3. Information of five partner universities in Japan ...... 92 Table 4. Criteria of OPI ratings (ACTFL 2012, 6) ...... 97 Table 5. Interpretation of J-CAT scores and comparison with JLPT ratings ...... 100 Table 6: Terms and acronyms used in the qualitative analysis of style shifting ...... 113 Table 7. Background information of the participants ...... 121 Table 8. EdA gain demonstrated by J-CAT ...... 123 Table 9. Participants' pre- and post-EdA OPI ratings ...... 126 Table 10. Occurrences* and percentage of utterances by categories ...... 129 Table 11. Occurrences* and percentage of types of plain form SMPs ...... 132 Table 12. Summary of EdA gain demonstrated by examining style shifting ...... 164 Table 13. J-CAT scores categorized by Training and Non-training groups ...... 167 Table 14. OPI ratings and style shift strategy development categorized by Training and Non- training groups ...... 168 Table 15. High-gainers and Low-gainers ...... 168 Table 16. Pre- and post-EdA OPI ratings ...... 207 Table 17. Summary of INoPs by gainers and low-gainers ...... 209 Table 18. Summary of INoPs by Training and Non-training groups ...... 211

List of Figures

Figure 1. Levels of language proficiency on the ACTFL Guideline (2012) ...... 7 Figure 2. Third Space in intercultural communication (Jian & Walker, 2017) ...... 12 Figure 3. Relationship scale ...... 63 Figure 4. Relationship scale with participants' responses ...... 64 Figure 5. Core and nodes in INoP ...... 115 Figure 6. Types of clusters ...... 116 Figure 7. Types of ties in INoP ...... 117 Figure 8. Example of an INoP diagram ...... 117 Figure 9. J-CAT score radar charts and OPI rating charts ...... 162 Figure 10. Cathy's INoP ...... 176 Figure 11. Frank's INoP ...... 180 Figure 12. Isabelle's INoP ...... 183 Figure 13. Henry's INoP ...... 188 Figure 14. Emma's INoP ...... 194 Figure 17. Social ties where Cathy, Frank, Isabelle, and Henry faced struggles ...... 213

xvi

List of Aberrations

The following abbreviations are used in this dissertation.

C1: base/first culture

C2 target/second culture

L1: first/native language

L2: second/foreign language

EdA: education abroad

DtSS: detached speech style

DtSS-PF: detached speech style plain form

IfSS: informal speech style

IfSS-PF: informal speech style plain form

NPF: naked plain form

AP: adjective predicate

VP: verbal predicate

NP: noun predicate

xvii

SMP: sentence matrix predicate (main clause)

NPw/oC: noun predicate without a copula

INoP: individual network of practice

xviii

Chapter 1: What We Know and What We Do Not Know about Education Abroad Learning in

Japan

1.1. Introduction

What issues do U.S. undergraduate students encounter in the process of personal relationship-building during yearlong education abroad (EdA) programs in Japan? In what ways can intervention address such issues? How much can such intervention mitigate these issues?

This dissertation aims to address these questions.

According to the Institute of International Education (2018), 332,727 American students participated in study abroad or EdA programs for academic credits during the 2016-2017 academic year and the number has tripled over the past two decades (IIE, 2018). In addition, students’ destinations for such programs have become more diverse in recent years. Although the top three traditional destination countries, namely, the United Kingdom, Italy, and Spain, have remained popular destinations for American college students, a growing number of students study abroad in less traditional venues such as China (ranked as sixth), and Japan (ranked as tenth) (IIE, 2018). This phenomenon suggests that more American students are motivated to learn languages and cultures that are less linguistically and culturally related to their own, making it an important part of foreign language education in the US. Considering the growing

number of American college students participating in programs in such less traditional destinations, it is crucial to understand what is actually happening during the programs in these less traditional destinations in order to support their learning.

The terms education abroad and study abroad are often used interchangeably to refer to enrollment in an educational program outside of a student’s home country. The two terms often are used to refer to the same or similar university-affiliated programs abroad. For example, the

Office of International Affairs at The Ohio State University (OSU) changed the name of such educational programs abroad from study abroad to education abroad in 2015 but their website indicates they still use study abroad as well.

According to the English Oxford Living Dictionary (2018), study, as a noun, is “devotion of time and attention to gaining knowledge of an academic subject, especially by means of books,” whereas education is defined as “the process of receiving or giving systematic instruction, especially at a school or university.” The former describes learning primarily from written texts, but the latter focuses on learning that takes place with pedagogical interventions. In addition, the former describes the process from the learners’ perspectives, whereas the latter includes both learners and teachers’ perspectives. Considering both OSU’s terminology change and the binary perspectives connoted by the term education, this dissertation uses education abroad (EdA) programs to also refer to those programs that are described as study abroad programs in other contexts and in previous research. EdA students in the current study are participants who seek to gain from the EdA experience. Educators are participants who create/modify the environment or intervene in students’ EdA experience.

1.2. Statement of problems

1.2.1. Problem 1: Pedagogical support for EdA programs in Japan

Students who participate in EdA programs typically expect that the presumably

“unlimited” exposure to the target culture and language provided by the programs is the best environment for culture and language learning (Dewey et al., 2014, p. 36). However, a number of studies have revealed that not all EdA students are able to take advantage of the learning environment, and that there are individual differences in the outcomes of the EdA learning (Berg et al., 2009; Isabella-Garcia, 2006 among others). Although there are a number of studies that investigated the EdA learning environment, studies that examine the effectiveness of pedagogical interventions are still limited (Davidson, 2010; Berg et al., 2009; and Cohen et al., 2005).

Considering the growing number of American college students in the EdA population, investigation of not only EdA students’ learning but also pedagogical support to enhance their

EdA learning is crucial.

Berg et al. (2009) 2 and Davidson (2010),3 whose study involved more than 1,000 participants, reported a positive impact of pedagogical intervention toward their participants’ oral proficiency gain during EdA programs. Berg et al. (2009) reported that EdA students who received pre-departure training or orientation that included cultural components outperformed

EdA students who did not receive such pedagogical support prior to their EdA programs.

2 Berg et al. (2009) study included students who studied Arabic, Chinese, French, German, Japanese, Russian and/or Spanish. 3 All of participants in the Davidson (2010) are American students who studied Russian. 3

Davidson (2010) also attributed the decease of gender as a significant predictor of oral gain4 among American students who participate EdA programs in Russia to training “in self- management strategy selection provided by ACTR5” (p. 20). However, because the focus of the two studies is more on identifying predictors of linguistic or cultural gain rather than the contents of pre-departure training, the types of cultural components and how they were taught in the pre- departure training or orientation are not specifically stated in these studies.

Contrastively, Cohen et al. (2005) provide descriptions of pedagogical interventions in their research. The authors investigated the effectiveness of pedagogical intervention toward intercultural development and the development of speech acts by U.S. college students who participated in EdA programs in French and Spanish speaking counties. Their pedagogical intervention included pre-departure training using the pedagogical material, Maximizing Study

Abroad: A Students’ Guide to Strategies for Language and Culture Learning and Use (Paige, et al., 2002), which focuses on cultural differences between the base culture of the participants (C1) and the target (host) culture (C2) regarding nonverbal communication including eye contact, personal distance, and touching behavior. Although the authors did not find significant differences between the experimental group, which had pedagogical intervention, and the control group, which did not, they found a significant difference on speech act development between the two groups regarding intercultural development (p. 64).

4 Davidson replicated the study done by Brecht et al. (1995). In Brecht et al., gender was a significant predictor of oral gain among American EdA students in Russia (see Section 1.2.3. in Chapter One). 5 The American Council of Teachers of Russian 4

However, Cohen et al.’s findings also suggested that pedagogical intervention seems to have had a bigger impact on the students who participated in EdA programs in European counties than those who participated in EdA programs in Latin America (p. 67). Some students who participated in EdA programs in non-European counties suggested that the Guide did not have enough information covering these counties, which led the authors to make an assumption that the Guide might be “biased towards the European study abroad experience” (p. 67). The findings also suggest that the contents of pedagogical intervention should vary depending on cultural differences between C1 and C2, and a better outcome may be expected if the pedagogical intervention focuses on the cultural differences between specific C1 and C2.

As the findings in Cohen et al. (2005) suggested, learning strategies for EdA students should differ depending on the cultures involved. However, the aforementioned EdA research often addresses EdA pedagogical support primarily in European countries, and there are few studies that investigate pedagogical intervention for students who participate in EdA programs in

Japan. Considering that Japan is ranked among the top ten most popular EdA destinations for

U.S. college students (IIE, 2018) as well as that a growing number of U.S. college students are participating in EdA programs in less traditional destinations, it is important to conduct research that can effectively inform the pedagogy for EdA programs in such countries.

1.2.2. Problem 2: Measurement of EdA learning

In order to measure the effectiveness of pedagogical support for EdA students in Japan, it is necessary to have reliable assessments that measure what the students learn or do not learn during the EdA experience. The American Council on the Teaching of Foreign Languages’

(ACTFL) Oral Proficiency Interview (OPI) test is a proficiency test widely used to measure U.S. college students’ oral proficiency gains during their EdA programs (Baker-Smemoe et al., 2014 etc.). The OPI proficiency test measures speaking production holistically by examining the accuracy, content, tasks, and functions of which test takers are capable of. Using this widely- used language proficiency test allows researchers to compare and contrast their findings with other research.

However, relying only on the ACTFL OPI ratings to measure EdA students’ learning may not reflect the whole picture of a student’s EdA learning, and its sensitivity to subtle language gains has been questioned in previous research (Kinginger, 2009 etc.). For example, as shown in Figure 1, the range of oral proficiency required at each major level increases as the proficiency level goes up. In other words, it requires a test taker to cover more areas of speaking skills, and as the test taker’s level goes up it takes more effort and longer time to move on to the next stage. It is expected to be more time consuming for an EdA student with Advanced-Low level at the start of the EdA program to move to the next sublevel (i.e. Advanced-Mid) than for those with Intermediate-Low to move onto the next sublevel (i.e. Intermediate-Mid). Therefore, the relying on the OPI ratings does not provide an equal assessment for EdA students who started with a different proficiency level.

In addition, the OPI criteria for Novice, Intermediate, and Advanced Levels do not consider culturally appropriate speech styles. This becomes problematic as a measurement of

EdA learning in Japan. Most US college students who participate in EdA programs in Japan are expected to be Intermediate to Advanced level L2 students. Those who are rated at the Superior

level or above would be extremely rare considering the difficulties of learning Japanese for speakers of English (cf. Foreign Service Institute, 1973).

Figure 1. Levels of language proficiency on the ACTFL Guideline (2012)

Cultural knowledge is essential for communication (Agar, 1994; Walker & Noda, 2010; inter alia), and using the target language (L2) outside of the framework of the target culture (C2) often leads to negative consequences (Walker & Noda, 2010). ACTFL (2012) claims that OPI

“assesses a speaker’s ability to use the language effectively and appropriately in real-life situations” (p. 4). However, just how a test taker manages to complete such tasks is often ignored.

Focusing on demonstrating improvement in OPI ratings might therefore encourage teachers’ and students’ tendency to think that when communicating in the L2, as long as one delivers a message in the target language, it does not matter how they are perceived by people in the C2.

This naïve way of thinking may have negative consequences and not be conductive to building successful L2-mediated personal relationships during EdA programs. Thus, relying only on OPI 7

ratings to measure the outcome of EdA learning is not sufficient for Novice and Advanced level learners. In addition, as Polanyi’s (1995) study reveals, the language proficiency measured with

OPIs may not reflect some aspect of language acquired by all EdA students (see Section 2.3.1. in

Chapter Two for more detailed discussion of Polanyi (1995)). Assessments of EdA learning should reflect the entirety of the students’ learning as much as possible.

1.2.3. Problem 3: Outcome vs. process in EdA learning

Studies that investigate EdA learning from the students’ perspective reveal that some of their learning might not have been reflected on the results of proficiency tests. For example,

Polanyi (1995) conducted follow-up research on the issue of gender as a strong predictor in

Brecht et al. (1995), and revealed that cultural differences in how gender roles were regarded between C1 and C2 affected different types of interaction that male and female students had in

Russia. Those differences in interactional experiences resulted in different types of spoken language being gained by male and female students. However, due to the nature of the language proficiency test used in Brecht et al. (1995), only the type of the language learned by the male students was considered as gain (which Polanyi called “unmarked ‘male’ language”). This eventually indicated gender as a strong predictor of oral proficiency gain in the study done by

Brecht et al. (1995). Although investigation of outcomes is important for the assessment of individual students’ EdA learning and EdA programs, focusing only on the outcomes measured through objective testing could miss important aspects of L2 learning and teaching, such as “why” the learning occurred or did not occur. Focusing on the “process” of EdA learning is important in

order to better understand the reality of EdA learning (Wilkinson, 1998; Wang, 2010) and to develop effective pedagogical support.

What makes L2 learning in EdA contexts more difficult is that there are vast differences in learning experience among students and even among the students who participate in the same

EdA programs, due to a less controlled learning environment than in the learners’ domestic programs (Noda, 2007, p. 304). Therefore, attempting to understand EdA learning only by investigating the outcomes of EdA learning measured via the proficiency tests or objective measurements may lead us conclusions that do not reflect individual EdA learning.

EdA students often encounter cultural differences in daily interaction which directly affects their construction of personal relationships with locals in the community. However, EdA students who struggle with cultural differences often do not recognize them as cultural differences (see Chapter Two). Such limitations in perception unfortunately can lead them to create a negative view regarding the C2 and its people, which, in turn, may inhibit and/or skew their EdA learning in the direction of such negative impressions. In order to provide pedagogical support that helps students overcome such negative experience and maximize their learning during EdA, obtaining data on such experiences is also necessary.

1.3. What makes intercultural communication difficult and complex?

L2 learners may not follow a process identical with their first-language (L1) socialization because they often choose to act unlike members of the C2 community for various reasons, but this is due primarily to conflicts between C1 and C2 values and norms (Kramsch, 2009; and

Siegal, 1995). Primary habitus, acquired through initial socialization at home, is more durable

than the secondary habitus learned afterward, including intercultural contexts, because during initial socialization a person usually conforms to the behaviors, norms, values, and social conventions of the community, and doing so constructs his or her history in that community

(Bourdieu & Passeron, 1977). Such histories are often challenged by the second habitus, especially in intercultural contexts (Kramsch, 2009).

Focusing on multilingual speakers’ experiences described above, Kramsch (2009) first used third place or third culture in order to “[eschew] the traditional dualities on which language education is based on L1/L2, C1/C2, NS/NNS, Us/Them, Self/Other” (p. 199). However, she later uses symbolic competence to replace those “third ~” terms, because according to Kramsch

(2009), symbolic competence emphasizes the symbolic nature of the multilingual subject, which is often ignored under the rubric of third culture/place. Kramsch discusses symbolic competence to signify multilingual perspectives of L2 speakers who are often considered as minority groups

(immigrants learning English as L2), in order for them to empower and confront norms “occulted by the monolingual policies of powerful nation-states” (p. 199).

However, when considering the process of building intercultural relationships in EdA contexts, it is crucial to focus on not only L2 learners’ perspectives, but also local people’s attitudes toward and perceptions of intercultural relationships. In addition to EdA students’ resistance to fully participating in C2 conventions, studies have revealed that locals seem to hold different expectations and norms for foreigners during their interactions (Iwasaki, 2011; Iino,

2006). Iino (2006), who investigated the norms of interaction in home stay settings in Japan, also argues that the communicative competence that EdA students are expected to have is not always the same as that which is considered as “native” or L1. For example, when an American EdA

student met her host family for the first time, she used a ritual speech act of giving a gift by saying “tumaranai mono desu ga…” which is considered conventional and appropriate in

Japanese culture (Iino, 2006). However, her host family started laughing, which suggests that they were not expecting an American to do the same thing as a Japanese L1 speaker.

Similar findings were reported in Iwasaki (2011) that some of the EdA students were explicitly told that they were not expected to use desu/masu forms6 because they were

“Americans” (p. 85), “a foreigner, or a white guy” (p. 86). In addition, she reports the participants’ surprises at the frequent use of vulgar language by young male Japanese speakers.

However, one of the participants in Iwasaki’s study discovered that he was not “allowed” to use men’s language when he tried to use the expression gomen na (sorry) to his host brother, which resulted in him receiving explicit negative feedback from the host brother.7 Because of their perceived complexity of linguistic choice in male-to-male communication, the participants in

Iwasaki’s study report that they avoided using it and used gender-neutral language. This indicates that native speakers have expectations that differ from those held by L2 learners, so that oftentimes a lack of understanding of the natives’ expectations results in self-imposed restrictions on the part of L2 learners. Outside EdA-in-Japan contexts, i.e. back home in their university program, learners may not have the opportunity to encounter such problems, but

6 Desu/masu forms are often used predominantly in a formal speech setting (see Chapter Two for more detailed descriptions of the forms). 7 This situation also suggests the complexity of the social meanings that underwrite men’s usage in Japanese. Gomen na can index friendliness/closeness as the speaker apologizes, but if that stance is at odds with the context, such as the situation above where the speaker was expected to apologize, it can be taken as rude, especially if the addressee does not believe that a close relationship is already in place. 11

outside the classroom in Japan, they do. Still, problems owing to lack of understanding have pedagogical solutions, and it the time to address this one is before EdA begins.

1.4. Conceptual framework: Third Space in the intercultural communication

Figure 2. Third Space in intercultural communication (Jian & Walker, 2017)

Such unique phenomena of, or moments in, intercultural communication during EdA can be explained with the concept of Third Space (TS) developed by Jian and Walker (2017). TS is defined as an imagined space where different cultures converge, contest or cooperate: Norms or rules of the game will not conform entirely to either C1 or C2, but will be dynamic and conditioned by shared interests and goals (Jian, 2018). TS indicates the need for co-construction between people from different cultures by recognizing cultural differences and working together toward an ideal intercultural communicative environment benefiting both sides. One difference

between the symbolic competence discussed by Kramsch and the construct TS developed by Jian and Walker is that Kramsch discusses the importance of recognizing the unique aspect of intercultural communication and L2 learning by considering that L2 learners are multilingual subjects, who are “less symbolically powerful (Zeng, 2015)” and often considered a minority group (as with immigrants learning English in the US.). Contrastively, Jian and Walker discuss the importance of recognizing the different norms and expectations in intercultural communication for L2 learners whose L1 and C1 are often considered more powerful than the

L2 or C2, as when an American college student is learning Chinese or Japanese.

From the pedagogical standpoint of teaching less commonly taught languages in the US, focusing on only an L2 speaker’s stance as a multilingual subject may result in negative consequences such as L2 speakers dictating their own norms and values to people in the C2 community. Their success cannot be accomplished without incorporating perspectives of people in the C2 into phenomena unique to intercultural communications. According to Walker and Jian, the necessity of communicating in TS increases as L2 learners takes on roles such as a student at a C2 school or a C2 employee as described at the bottom of Figure 2. In their diagram, the concept of TS does not apply to the intercultural communication situations where L2 learners take the role of a missionary, a government representative, etc., shown at the top of Figure 2. Jian and Walker point out that this is because in those roles, the speaker’s communication mode is declarative rather than conversational, speakers in these roles in such contexts do not require co- constructing and/or modifying their core message in order to communicate to the other side

(Zeng, 2015). For example, a religious missionary’s ultimate task is to propagate their belief (or

“truth”) to people. There is no room for negotiating what they believe when they communicate

with local people in C2 as a missionary on propagational matters. In other words, if people in C2 believe a different religious or “truth,” there are no negotiations or compromises to find a middle way in their beliefs. However, an American missionary can adapt and, when useful, adapt communication strategies of Chinese people to converse effectively, without compromising their message.

By contrast, international students in the US who seek degrees from universities, for example, have to follow the rules and adapt to the system of the university in order to achieve their goal, which may be more or less different from the rule or systems in their own countries.

The existence of international students in classes often results in changes on the part of professors or university administrators, e.g., adjustment of the rules of their classes, either consciously or unconsciously. This is often described as the bi- or multi-directionality of socialization (Talmy, 2008). In this intercultural communication, there is a shared goal between the international student and the professor, i.e., learning and teaching the subject successfully.

One side’s success is necessary to the other side’s success.

1.4.1. Co-constructing a constructive Third Space

The concept of TS is a helpful framework in investigating EdA learning. The goal of TS is meaningful relationships, which seek and accommodate shared goals, where L2 learner and L1 speakers cooperate to co-construct an ideal environment that benefits both parties. However, there are times when what goes on in the TS is not productive or constructive, especially in the beginning when there is a gap between what L2 learners and L1 speakers expect. For example, as we saw above, the L2 learner who followed the Japanese custom of giving a gift ended up

receiving laughter from her host family (Iino, 2006). Through this interaction, both the L2 learner and host family “edited” their co-constructed TS in a way that might not be as productive or constructive as the L2 learner wanted it to be. Because TS is not stable, but constantly changing (i.e. dynamic) as L2 learners and people in the community interact with each other on a continuous basis, successful intercultural communication can lead to co-construction of a meaningful TS that is productive for both side. However, TS can become confrontational when, for example, repeated interactions lead to the construction of negative views toward the other. As an L2 speaker successfully gains membership in a C2, the norms and values of the TS may become closer to those of the C2. However, for an adult learner, TS will never be the same as C2 due to the persistence of their pre-existing L1 socialization (Bourdieu & Passeron, 1977;

Kramsch, 2009).

1.5. Purposes of the current study and definition of EdA gain

By utilizing the concept of TS, the current study investigates American students’ EdA learning by investigating TS they construct during EdA programs in Japan in order to ascertain pedagogical implications and to maximize EdA learning in Japan. The purpose of the current study is to investigate (1) types of cultural and linguistic issues that U.S. undergraduate students encounter regarding building personal relationships during EdA programs in Japan; (2) ways in which EdA training can address the issues described in (1); and (3) the degree to which EdA training can mitigate these issues.

In the current research, EdA gain is defined as the outcome of EdA learning, (including linguistic and cultural knowledge acquired during EdA) that is measurable with a defined set of

assessment batteries. As mentioned above, not all EdA learning is measurable with objective testing, nor can any proficiency test measure L2 learners’ skills for managing communication in particular settings specific to their lives (Zeng, 2015). Furthermore, while gain indicates “an increase in wealth or resources,” outcome indicates “the way a thing turns out (Oxford

University Press, 2019),” which does not necessary indicate the constructive aspects of learning.

Therefore, in order to acknowledge these limitations in investigating and assessing the “outcome” of EdA learning as well as to highlight its constructive aspects, the term EdA gain is employed in the current study.

1.6. Organization of the research and research questions

In order to achieve the aforementioned purpose, this dissertation research consists of two phases of empirical studies. The first phase pertains to investigating TS in an EdA program in

Japan and identifying struggles that U.S. students encounter when forming intercultural relationships with locals. Experiences and struggles from the American EdA students’ perspective have been well investigated in the existing literature, but little research has addressed the same issue from the perspective of their local Japanese counterparts. Since social network construction and socialization are not one-directional, but bi- or multi-directional (Talmy, 2008), as mentioned above, Japanese people’s perceptions are essential to identify struggles and problems in the process of forming cross-cultural social networks and becoming socialized in

Japan. Three research questions guide this phase of the study, which are as follows:

1. What kinds of linguistic and cultural difficulties do local Japanese students encounter

when interacting with American EdA students?

2. What kinds of linguistic and cultural difficulties do American EdA students encounter

when interacting with Japanese local students?

3. What kind of Third Space emerges when American EdA students and Japanese local

students interact with each other?

The first phase is based on fieldwork done in a four-week-long EdA program in Kobe,

Japan during May 2017. I interviewed both the American participants of the program and the local Japanese students who volunteered to spend time with these students as language partners, in order to elicit their perceptions of their experience as well as insights related to forging intercultural social networks. The interview results were analyzed immediately and incorporated into the design of an EdA training curriculum to be used in the second phase.

Analysis of the interview data reveals that gaps in expectations about the U.S. students’ speech style between the U.S. students and the locals at the home institution. Based on the findings, a pre-departure training regimen was designed to support EdA students’ construction of a relationship-building persona. The notion of a persona in L2 learning, as used in the current study, is discussed in Chapter Two along with other relevant research literature. The EdA training, which focuses on behavioral culture8 in “peer-to-peer” interactions, consisted of a set of role-play activities and discussion on topics regarding intercultural-cultural communication. The

8 See Chapter Two for the definition 17

findings of the first phase as well as the contents of the EdA training are discussed in more detail in Chapter Three.

The second phase investigates EdA gain and the process whereby EdA gain develops

(or does not), as well as the effectiveness of the EdA training developed, based on the findings in

Chapter Three. Eight American undergraduate students, who were in yearlong EdA programs in

Japan during the 2017-2018 academic year, originally participated in the second phase of the study. Four of the participants volunteered to attend pre- and concurrent EdA training, and the other four did not. Their EdA gain was measured with three types of assessments, namely,

ACTFL’s OPI, Japanese Computerized Adapted Test (J-CAT), and examination specifically of the participants’ style shifting in the OPI data. The participants took the two proficiency tests once pre-departure and again post-return. The effectiveness of the EdA training was measured by comparing EdA gain and social network development of those who attended EdA training (EdA- training participants) and those who did not (non-EdA participants).

Data on social networks were collected through questionnaires, post-EdA interviews, and language use administrated four times during the EdA programs. Considering variables of

EdA contexts, (e.g. individual differences in interactional experiences), the current study employs Individual Network of Practice (INoP) (Zappa-Hollman & Duff, 2015) as a unit of EdA social networks analysis. INoP was designed specifically to analyze L2 speakers’ social networks during their EdA programs by focusing on an L2 speaker’s connections with people and social groups in the community. This analysis allows us to understand the EdA participants’ social network in detail. Detailed descriptions of the methodology of the second phase of study are presented in Chapter Four.

Chapters Five and Six discuss the findings from the second phase of research. Chapter

Five presents and discusses the findings of EdA gain in relation to two research questions:

1. What did students gain or not gain during EdA in Japan?

2. Are there any differences in EdA gain between the students with or without

pre-departure the EdA training group? If so, how?

Chapter Six presents the findings related to the students’ EdA experience and social network development. It also discusses the finding in relation to three research questions:

1. What are the commonalities and differences in U.S. students’ social

networks during EdA programs in Japan?

2. In what ways did their social networks influence the students’ EdA gain

discussed in Chapter Five?

3. Are there any differences in EdA social networks between the students with

or without pre-departure EdA training? If so, how?

In conclusion, Chapter Seven addresses the primary purpose of the current dissertation, i.e., pedagogical implications for American undergraduate students who participate in EdA programs in Japan. It provides pedagogical implications for how the home institution can provide the kind of support that can maximize their EdA learning. It also discusses future research implications by addressing limitations of the current research.

1.7. Significance and constraints of the dissertation

The current research is considered significant in four respects. First, it is carefully designed to address the lack of empirical research on the effectiveness of pedagogical

intervention for in EdA learning in Japan by incorporating two phases of research. The first phase investigates difficulties that American EdA students and Japanese local students encountered while trying to forge intercultural relationships. Based on the findings, EdA training was developed to support American EdA students to prepare for such difficulties during their

EdA in Japan. The second phase tests the effectiveness of the EdA training by comparing the social networks and EdA gain between four participants who went through the training prior to

EdA in Japan and another three participants who did not. In addition, by focusing on the effectiveness of culture-specific training, this research aims to provide insights on such pedagogical interventions for EdA learning.

Second, as mentioned above, in order to design an effective pre-departure training regimen to support EdA students’ social network construction in the target culture (C2), it is crucial to incorporate not only EdA students’ but also local people’s perspectives. However, very few studies have included data of local people’s perspectives regarding intercultural communication and relationships. The current study applies the concept of a Third Space, and includes data on both American EdA students and the local Japanese people to investigate difficulties in building intercultural relationships during EdA in Japan. In addition, such complementary perspectives can elucidate aspects of ignored culture, which is often difficult to recognize (discussion of ignored culture along with other aspects of culture is found in Chapter

Two). Analysis of the data informs us of not only how the curriculum of EdA training should be designed in order to maximize EdA students’ learning, but also informs us how the curricula designs of Japanese foreign language programs in the US should be designed to provide instruction for students on developing good personal relationships with Japanese people.

Third, the current research addresses the issue of using the ACTFL OPI test as the only measurement of EdA gains by employing qualitative analysis of Japanese style shifting during the OPI. Style shifting in Japanese requires not only linguistic knowledge on how to conjugate each form, but also cultural knowledge of social hierarchy, formality, and so forth in Japan. Thus, analysis of EdA students’ style shifting before and after EdA enables us to assess such knowledge gain during EdA in Japan. Comparison of the results from that analysis and the OPI ratings reveal that an OPI rating alone is inadequate in measuring gain during EdA experience.

Furthermore, in order not to limit EdA gains to oral skills measured with the OPI, the current study also employs J-CAT to assess such skills in other areas, such as vocabulary, grammar, listening and reading (see 4.4 in Chapter Four for detailed discussion of instruments used to measure EdA gain). Employing multiple assessment batteries offers a more multi-faceted picture of the students’ EdA gains in Japan. By providing and interpreting its empirical data, the current study contributes to a better understanding of the assessments of EdA gain in Japan.

Fourth, as mentioned above, in order to understand not only the outcome of EdA learning but also its process, the current study investigates US students’ social network development in relation to EdA gain. Despite the fact that strong social networks with locals have been reported as positive predictors of EdA gains, there are few studies that investigate the issue of social networking in regards to EdA programs in Japan. The current research addresses the relationships by assessing EdA gains with three different assessments and using a social network analysis (i.e. INoP) specifically designed for EdA learning contexts.

Nevertheless, the current study also includes some limitations. One limitation of the study is that the contents of the pre-departure training were based on the findings of fieldwork

research in a four-week EdA program, and the effectiveness of the training was tested with seven

American students who participated in EdA programs at five different universities in Japan.

Designing the EdA training based on one EdA population and testing the pre-departure training on a separate population may be considered a constraint of the study. For example, the dialectal differences might influence both U.S. and local Japanese students’ perspectives toward each other. However, considering the variables in yearlong EdA programs (i.e. programming, duration, interactional opportunities), it was easier to control such variables in the short-term EdA program, which contributed in developing more unified and general perspectives on building intercultural relationships between American and Japanese students.

Another limitation is that involving participants from a single home institution may affect overall findings regarding EdA students’ learning experiences. This is because previous learning experiences at the home institution may affect their perceptions of L2 and C2. To counterbalance this limitation, the current research provides detailed descriptions of the participants’ perceptions before and after their EdA in Japan including their previous learning experiences at the home institution.

Chapter 2: What Happens during the Education Abroad Experience and How Can We Guide

Students’ Learning?

2.1. Introduction

This chapter aims to provide background on the current research by presenting and discussing relevant research on learning and experiences of Education Abroad (EdA) students. It first reviews research literature that focuses on Japanese style shifting. Reviewing both Japanese native (L1) speakers and second or foreign language (L2) speakers’9 style shifting allows examination of style-shifting development as a measurement of EdA gain (see Section 1.5 in

Chapter One for the definition of EdA gain). Second, it reviews research literature that focuses on the relationship between social network development and EdA gain. By examining recent studies that have reported a strong social network with locals as a strong predictor of EdA gain, it discusses how the investigation of social networks can be used as a means of understanding the process of EdA gain. Finally, this chapter discusses the Performed Culture Approach as a

9 Although there are differences between a second language (L2) learning environment and a foreign language (FL) learning environment (Ringbom, 1980, pp. 39-41), such differences become obscure in EdA contexts (Freed, 1995, p. 4). Because the differences between the two learning environments and the two terms are not the focus of the current study, “L2 speakers/learners” will be used to refer both L2 and FL speakers/learners. 23

framework for EdA training in relation to the development of culturally appropriate personae in

L2 learning.

2.2. Style shifting as a measurement of EdA gain in Japan

Chapter One elucidated the problems of using the ACTFL10 Oral Proficiency Interview

(OPI) as the sole measurement of EdA gain. Having the speaking ability to conduct appropriate style shifting requires both the linguistic skills to produce correct forms and the cultural knowledge to judge the appropriateness of a stylistic choice in a given context. In addition, spoken data from the OPI can be used to analyze participants’ style-shifting development without collecting new data (Iwasaki, 2008, pp. 79-80). By considering these advantages, this section discusses how examining style shifting can be used as a reliable measurement of EdA gain in Japan. In order to do so, the first section presents the definitions of two style forms in

Japanese, namely, the masu forms and the plain forms.

2.2.1. Definitions of forms in Japanese

Let us first look at the linguistic and pragmatic aspects of two forms that represent different speech styles in Japanese. In the Japanese language, there are two main style forms, those that end with desu or masu forms (hereafter, masu forms) and others that have a larger variety of endings (hereafter, plain forms). Table 1 shows formal distinctions between the two in three different predicate types, namely, verbal, adjectival, and nominal predicates11.

10 American Council on Teaching of Foreign Languages 11Regarding the different types of predicates in Japanese, the current study follows the definitions by Jorden & Noda (1987). A verbal is a word whose form endings include -masu and 24

Table 1.Plain forms and masu forms in imperfective forms Type of Predicate Plain form Masu form Verbal predicate gakkoo e iku. gakkoo e ikimasu. (I go to school.) (I go to school.) Adjectival predicate ano hana wa chiisai. ano hana wa chiisai desu. (That flower is small.) (That flower is small.) Nominal predicate kore wa kaban da. kore wa kaban desu. (This is a briefcase.) (This is a briefcase)

As shown in Table 1, the English equivalents for both plain and masu forms are the same, but the social stances indexed with those two forms differ. Plain forms often predominate in casual conversation between persons whose relationship is personal and intimate, in settings that do not demand formality. Masu forms, on the other hand, identify the speaker as showing “solicitude toward, and maintaining some linguistic distance from the addressee (Jorden & Noda 1987, p.

32).” Therefore, the differences in the social stances indexed with the verbal predicates in Table

1 are differences in the stances that the speaker takes vis-à-vis his/her addressee. For example, in the same setting, gakkoo e iku may be uttered to someone with whom the speaker feels intimate, such as a friend or a family member, whereas, gakkoo e ikimasu may be said to someone with whom the speaker feels less intimate, such as a professor or a business partner. However, some settings (e.g. ritualized actions) impose roles and role-appropriate interpersonal stances, which can override the more personal side of one’s relation to her addressee, and thus constrain available choices, so the above are just a few examples of the many kinds of contexts in which those two utterances are likely to be produced.

mashita (p. 31). An adjectival is a word that includes forms ending with –i and –katta (p. 39). All other predicates are categorized as nominal, and a nominal itself does not conjugate. A nominal predicate consists of one or more nominals followed by a copula, which in some cases is obligatory but other cases it is not (p. 50). (See Chapter Four for detailed discussion of a nominal predicate without a copula). 25

While the term “plain form” is relatively consistent among most scholars, the masu form is not. Masu forms are often described as addressee honorifics “because it acknowledges distance and formality towards the addressee” (Iwasaki, 2008, p. 46), thus, some scholars who focuses on such functions of the masu form call it the polite form (Hasegawa, 2015, p. 256).12 Although its function as an addressee honorific is often emphasized, honorific treatment of one’s addressee is not the only function that masu forms are used for, as Cook (2008c) and Geyer (2008) demonstrated in their studies. Cook (2008c) also identifies various social meanings of masu forms that are referenced in casual conversations. Therefore, in this dissertation, the term “masu forms” refers to a predicate that ends with desu or masu (including their other inflected forms, such as desita or masen), acknowledging the different functions and social meanings of the forms.

As mentioned above, although they are not limited to it, sentence-final, main-predicate plain forms are predominantly used in a casual or informal speech setting, whereas masu forms are predominantly used in formal speech settings. Unlike English, there is no socially neutral predicate form for ending a full sentence in Japanese, and a speaker has to choose one at the end of each main clause, depending on the context.13 Research on style shifting by Japanese L1 speakers has examined how speakers use both masu forms and plain forms within a single conversation and thereby elucidates that choice of the two style forms does not depend solely on the formality of the speech setting or the closeness of two conversation partners. The following section discusses this phenomenon of style shifting in greater detail.

12 These forms are also named differently in Japanese textbooks used in the US. 13 Japanese L1 speaker uses an incomplete sentence as an avoidance strategy (Cook, 2006). 26

2.2.2. Style shifting by Japanese L1 speakers

Japanese L1 speakers often shift speech styles when talking to the same addressee in conversation (Ikuta, 1986; Maynard, 1993; Cook 2002, 2006, 2008a, 2008b, 2008c; Geyer 2008;

Matsumoto 2008, and Ono and Jones, eds. 2008, among others). Research suggests that factors that trigger style shifting differ depending on the context, such as a masu-form predominant discourse context or a plain-form predominant discourse context.14 Considering the formality of the OPI, which the current research uses as data to analyze participants’ style-shifting development, this section focuses on style shifting in discourse contexts where masu forms are predominantly used.

By examining style shifting by Japanese L1 speakers, researchers aim to understand the fundamental concepts of the two forms. Ikuta (1983) argues that masu forms and plain forms indicate degrees of social, attitudinal, or cohesion “distance” rather than the degrees of

“politeness” or “formalness.” When a speaker uses masu forms, it indexes the speaker’s psychological distance from the addressee, whereas when he or she uses plain forms, no such psychological distance is indexed. Ikuta points out that in the beginning of a conversation, the speaker’s choice of speech style is often based on a factor associated with social distance, such as the relative gender, age, social status, and relationship of the participants as well as place, time and purpose of the conversation. As the conversation continues, the speaker’s attitudinal distance from the interlocutor changes moment by moment, as it is indexed in the speaker’s style shifts within the conversation.

14 Makino (2002) calls a masu form predominant discourse context as “SOTO-oriented” and plain form predominant discourse context as “UCHI-oriented” context. (p. 124). 27

Maynard (1993) argues that differences in these two forms indicate a speaker’s

“addressee awareness.” She claims that when a speaker is aware of the addressee as psychologically separate from himself or herself, masu forms are used. On the other hand,

Maynard argues that speakers use plain forms to their close friends in a casual setting because they feel psychologically close to the interlocutor, which Maynard characterizes as the speaker having low awareness of the addressee (p. 178). She also points out that some naked plain forms

(NPF)15 are not deliberately addressed to the conversation partner, and such NPFs are used when the speaker focuses on the content rather than to the addressee (p. 159). Such NPFs are often found in written genres found in newspapers, academic journals or academic conference presentations, where the speakers or authors may not have specific individuals as their addressee.

Cook (2008a) disagrees with Maynard’s argument that plain forms are used when a speaker feels psychologically close to the addressee (i.e., has low awareness of the addressee) by pointing out a quarrel between strangers on the street. Cook argues that the masu and plain forms are linguistic resources to “construct various social persona and contexts” (p. 15). She considers the two forms as “affective stance markers,” which suggest the speakers’ “moods, attitudes, feelings, and dispositions as well as degrees of emotional intensity” (Cook, 2008a). For example, masu forms are used to construct a stance of self-presentation (a fairly ritualized speech act), and plain forms are used to indicate “a lack of presentational persona” (Cook, 2008, p. 15). She argues that people interpret a presentational stance differently depending on the context by shifting their speech style.

15 NPF is a plain form that does not have any affect keys such as a sentence-final particle, vowel lengthening, rising intonation, coalescence or certain voice qualities. 28

Cook (2008b) also categorizes NPFs into two styles depending on how they are used: 1) informal speech style (IfSS) and 2) detached speech style (DtSS). IfSS plain forms (IfSS-PF) are often used with affect keys, such as a sentence-final particle, vowel lengthening, rising intonation, coalescence or certain voice qualities (p. 85). On the other hand, DtSS plain forms (DtSS-PF) are typically in a NPF that is “devoid of emotion” and often used to provide “foregrounds and referential content of an utterance (p. 85).16 Cook (2002, p. 157) and other researchers often find that TV interviewers’ use DtSS-PF to evaluate or summarize the interviewee’s utterances. Below is an example of DtSS-PF.

Excerpt 1: The interviewer (A) is interviewing a few clients in a yakitori restaurant. (Cook, 2002, p. 157) 1 I17: Yakitori no miryoku wa doo iu toko desu ka. ‘Uh what is yakitori’s appeal?’ 2 C: A, yappari, yasukute oishii n de, oishii kara. ‘uh, it’s after all inexpensive and delicious, ‘cause it’s delicious’ 3→18 I: Yasukute oishii.19 ‘It’s inexpensive and delicious.’

The intention of the interviewer’s “Yasukute oishii” in line 3 is to summarize the previous comment by repeating it. It is addressed to the audience (i.e., people who watch the TV program) rather than the interviewee. “Yasukute oishii” in this context is the interviewer’s summary, but it is not directed to the interviewee.

Cook (2008a) further demonstrates these points in her study on Japanese L1 speakers’ style shifting in student-professor conversations during academic consultation sessions. During

16 Cook (2008b, p. 95) also discuss that when NPFs are used in off-task talk (i.e., ordinary conversation such as outside of happyoo routine) they index IfSS. 17 I = Interviewer; C = A female customer 18 The arrow (→) indicates the line that is primarily analyzed. 19 Bold face indicates target item analyzed 29

the academic consultations, the professors and students mainly use masu forms to present themselves as “‘professional’ professor and ‘professional’ student” (Cook 2008a, p. 18).

However, the professor sometimes switches to plain forms as “a positive polite strategy” (Cook,

2008, p. 20), which is displayed in Excerpt 2 below.

Excerpt 2: Professor Y and Student F (Cook, 2008a, 19-20) 1 P: nani ka shitsumon arimasu ka ‘Do you have any questions?’ 2 S: eeto (turning the pages of the book) ‘We:ll’ 3 → P: koko made de: [wakannakatta tokoro toka soo iu no nakatta?] (Up to this point, isn’t there anything that you didn’t understand?) 4 S: [(coughs) 5 → P: eego wa muzukashiku nakatta? ‘Wasn’t the English difficult?’ 6 S: atta- (laugh) atta to omou n desu kedo chotto ‘There were- I think there were (questions).’

The professor’s shift from masu form in line 1 to plain forms in line 3 indicates his choice to be seen as a “personal coach” rather than being a “professional professor.” In this situation, by using plain forms in lines 3, and 5 the professor tries to shorten the psychological distance between him and his student. This type of plain form is categorized as IfSS-PF. Although the professor shifts to the plain form in line 3, the student remains in masu form in line 6, which indicates a hierarchical relationship between the two. If the student also switched to plain forms in line 6, the hierarchical relationship would not have been indexed and thus maintained. To complete the construction of the hierarchical relationship, the professor’s choice alone could not be sufficient, but the student also must choose the appropriate form (i.e. maintain the use of masu forms). By choosing to or not to use plain forms, the professor and student co-construct the hierarchical relationship in an academic institutional setting.

Cook (2006) also reports that students sometimes switch to plain forms during the academic consultation sessions with professors although the frequency of switching to plain forms is much less than that of the professors. In addition, the functions of plain forms deployed by the students are different from the ones by the professors discussed above. Excerpt 3 shows a student’s use of such plain forms.

Excerpt 3: Professor (P) and Student (S) (Cook, 2006, pp. 279-280) 1 P: de sotsuron de wa ano: shuukyoo to seeji o? ‘and in your graduation thesis uh religion and politics,’ 2 S: a soo desu. ‘That’s right.’ 3 P: no kankee o mitai wake desho? ‘the relationship (of the two) you want to look at, right?’ 4 S: u::n demo nanka ano ano topikku de kaku ka doo ka: [wakaranai desu ‘Uh, but ((pause)) somewhat I don’t know if I will write on that, that topic.’ 5 P: [un srya ‘yeah, 6 P: mochiron suki na koto yatte ii n desu yo. ‘of course, you can do what you want to do.’ 7 S: hai ‘Yes’ 8 P: dakedo20 purote- ano: kurisuchan ga ‘But (in Korea) Prot- uh Christians are’ 9 S: = hai, ‘Yes’ 10 P: sanwari [sukunaku tomo sanwari tte iu koto wa Nippon no ippaasento ni mitanai ‘3 % at least 3 %. That means compared with Japanese Chris- tians which are less than 1 %.’ 11 S: [so sanwari [‘yes, 3 percent.’ 12 P: ano: kurisuchan no kazu to hon: toni kurabemono ni naranai kurai [kirisutokyoo ga ‘(in Korea) the number of Christians ’ 13→ S: [ooi [‘large’ 14 P: Shinja ga ooi. ‘(the number of) believers is much larger.’ 15 naze na no ka tte omoshiroi shitsumon da to omoimasu yo

20 It follows Cook’s (2006, p. 279) Romanization. 31

‘why that is so is an interesting question, I think.’ 16 S: ha: demo nanka yappari ironna yo --- yooin ga aru to omou kara ‘yeah, but uh as we expect, since I think there are various fa- factors,’ 17 P: un ‘uh huh’ 18 S: boku wa yappa sono uchi no ja: seeji to shuukyoo toka nanka yappa ‘among these (factors) I (will study) politics and religion sort of.’ 19 sore demo mada hiroi n desu kedo, ‘Still (that) is still broad but.’ 20 P: un ‘uh huh’ 21→ S: hitotsu ni shika: shiborenai. Hitotsu ni shibotte yaru shi ka nai. ‘I can only focus on one. There is no other way than focusing on one.’ 22 P: hai. ‘Yes’ 23→ S: de tabun shuunyu shuunyu wa: ja sono kaikyuu o daihyoo kaisyoo o arawashite iru ‘and perhaps income, income represents that social class, expresses that class.’

Cook points out that these uses of plain forms do not indicate the student’s casual stance to the professor. For example, in line 12, the professor is saying that in South Korea there are many

Christians. In line 13, the students’ utterance “ooi”, an adjective predicate in the plain form, is overlapping with the professor’s utterances in line 12, but the professor continues his turn in line

14. This “ooi” does not indicate the student’s psychological closeness to the professor, rather it is a “resource for the co-construction of the idea with the professor (p. 281).” Similarly, the use of the plain forms in lines 21 and 23 indicates the student’s conviction to himself rather than the indication of closeness with the professor. Cook also reports that students sometimes leave their utterance incomplete (i.e. as fragments) without the final masu or plain forms. By avoiding the choice between the two forms, the student remains uncommitted to a particular social relationship with the professor at that moment (Cook, 2006, p. 283).

Researchers have found various uses of these plain forms that do not indicate intimacy toward the interlocutor. Japanese L1 speakers use such plain forms for soliloquy-like utterances, such as exclamatory expressions (Okamoto, 1999, p. 62), suddenly recalling something, self- reflection, which is often expressed by repeating a word or phrase (Hudson, 2011, p. 3691). In addition, such plain forms are also used to co-construct ideas (Cook, 2006, p. 284) and to present background information (Makino, 2002, p. 134). The degree, or extent, of such style shifting in a masu form-predominant discourse context differs depending on the speakers, but Japanese L1 speakers also utilize style shifting to avoid sounding too friendly or too formal (Okamoto, 1999, p. 63).

2.2.3. Style shifting by Japanese L2 speakers

Previous research on Japanese L1 speakers’ style shifting reveals that both style shifting and fragments (i.e. incomplete sentences) as well as other features of speech21 play important roles for Japanese conversation partners in ongoing conversation, at the same time, to co- construct relationships. However, skills for style shifting are not easy to manage even for advanced L2 learners (Taguchi, 2015).

Masuda (2010, pp. 195-196) compares six Japanese L1 students (two from each a two- year junior college, a four-year college and graduate school) and six Japanese L2 students’ (five of them were undergraduates and one was a graduate student) style shifting in professor-student

21 See Japanese: The Spoken Language (Jorden et al., 1987) for detailed discussion of how these features constitute a whole dimension of interpersonal stance-taking/indexing (p. 197; pp. 226- 228). 33

conversations. Table 2 presents findings on Japanese L1 and L2 speakers’ utterances in Masuda

(2010).

Table 2. Types of utterance ending by Japanese L1 students and L2 students' utterance (Masuda, 2010, p. 195) Masu forms22 Plain forms Omission of Others Total predicates L1 students 52.5% 10.6% 3.4% 33.8% 100% (1072) (563) (114) (39) (356) L2 students 78.6% 7.5% 9.8% 4.1% 100% (1031) (810) (77) (101) (43)

The Japanese L1 students produced 52.5% masu forms and 10.6% plain forms, while the

L2 students used masu forms in 78.6% of all utterances. Although the quantitative analysis indicates that the Japanese L1 students used fewer masu forms than the Japanese L2 students, the qualitative analysis revealed that the Japanese L1 students use “Others” much more than the L2 students when they are talking about themselves or engaging in joint utterances. Masuda (2010, pp. 195-196) explained that those findings reflected Japanese L1 students’ use of fragments (i.e. incomplete sentences), who used predicates ending with “-tari, -si, -te, -toka, -mitaina, -kara, - kedo, node, -tte,” when they are talking about themselves or engaging in joint utterances.

However, the Japanese L1 students in her study used masu forms mostly when they were asking or answering questions directly to their professor, expressing acknowledgment and agreement, or giving assessment or confirmation, which are considered as high addressee-awareness situations discussed above. Masuda also found that fragments were often used for not only syntactic co- construction, but also semantic co-construction in Japanese L1 students’ utterances. In total, the

22Masuda (2010) uses teinei-tai (丁寧体: polite forms). 34

L1 students use fragments in approximately 38% of their utterances, whereas the number of the fragments is only 14% for the L2 speakers. Masuda’s comparison of style shifting between L1 speakers and L2 speakers of Japanese elucidates that the L2 Japanese speakers tended to shift style less than the L1 speakers in a masu-form-predominant context, which could result in sounding too formal or unfriendly (Okamoto 1999, 63). On the other hand, L1 speakers utilized style shifting in order not to sound too formal while still maintaining masu forms to express respect or formality to their addressees (i.e., their professors).

Although it is still limited, some researchers have investigated style-shifting development in EdA contexts. Iwasaki (2008) compared five male English L1 students’ style shifting before and after nine months of EdA experience in Japan by analyzing the students’ speech styles in pre-, and post-EdA OPI interviews. Quantitative analysis revealed that all of the students mainly used masu forms prior to EdA in Japan. However, upon their return, two students overused plain forms whereas the other three continuously used masu forms predominantly in interaction.23

Although the other three participants also increased their plain form use in main clauses, their increase rate is relatively small compared to the two participants who overused plain forms.

However, Iwasaki’s qualitative analysis suggested that all students, including the two students who displayed overuse of plain forms gained some understanding of the two speech styles, and they seemed to shift the two styles more actively in the post-EdA OPIs than the pre-EdA OPIs.

For example, when the two students who displayed overuse of plain forms asked the interviewers

23 As mentioned above, predominant use of masu forms is expected in a Japanese OPI test although non-use of masu forms does not directly affect its rating. In addition, as Iwasaki points out, the interviewer was the participants’ former Japanese language instructor, making the predominant use of masu forms appropriate. 35

questions and expressed an apology as a part of role-play, they used masu forms including self- correcting their use of plain forms, which indicates their deliberate use of masu forms. In a separate study (Iwasaki, 2011, p. 97), the author accounts for those individual differences in learning speech style (and style shifting) as following from differences in the kind and/or amount of interaction as that the participants had during their EdA experiences in Japan (see Section

2.3.2 below for more a detailed discussion of the Iwasaki’s 2011 study).

Taguchi (2015) examined EdA students’ development of the ability to use appropriate speech style including style-shifting skills,24 in an informal speech setting by using data from 20 minute-peer-to-peer friend conversations. Because all participants in her study were international students at the same Japanese university, informal language use (i.e. predominant use of plain forms) was expected in their conversation contexts (Taguchi, 2015).25 To measure the participants’ skill in style shifting, the researcher also took the role of a conversation participant who came in the middle of the conversation. Because she held a higher social status than her participants in terms of age (older) and occupation (a researcher and a professor), the participants were expected to use the formal speech style (i.e., predominant use of masu forms) toward the researcher.

The findings revealed that the total number of students’ utterances increased from 1536 in the pre-test to 1919 sentences in the post-test, which indicated linguistic development over the six-month EdA program in Japan. Similar to Iwasaki’s findings, Taguchi also found an increased

24 Taguchi calls such skills interactional competence (see Taguchi (2015) for further discussion of interactional competence). 25 Taguchi prompted her participants to have “an informal, casual conversation with a same age, close friend peer (Taguchi, 2015, p. 160)”. 36

use of plain forms and a decreased use of the masu forms. Because the majority of the conversational contexts was informal (i.e. peer conversation) except the times when the researcher joined the conversation, the increased use of plain forms indicates that the participants gained a better understanding of speech styles, at least in the informal context, over the six months.

In addition, the results of analysis of the participants’ use of affect keys26 demonstrated that the use of plain forms with affect keys increased from 19% to 23% of their total utterances.

Overall, use of rising intonation was the most frequently occurring in this category. The overall results in Taguchi (2015) showed that there was an increased use of plain forms (including plain forms followed by affect keys) and fragments in the EdA students’ utterances after six months in the EdA program in Japan.

The overall increase in plain forms was also observed in three-way conversations (i.e. the conversation among the two student participants and the researcher). Because the researcher held a higher social status than the students, the appropriate speech style for the students to use when they are talking to the researcher is that of the masu form. A quick glance at the quantitative results might suggest that, overall, the students have not developed skills for appropriate speech styles, including style shifting in a formal speech context. However, Taguchi reported that there were individual differences among the EdA students. Similar to Iwasaki’s, Taguchi’s interview data suggest individual differences resulting from different interactional experiences during the six-month EdA program in Japan. For example, one student, referred to as Student R in Taguchi

26 Affect keys are considered as important interactional resources in Japanese conversation (Taguchi, 2015). According to Cook (2008b) affect keys are “sentence-final particles, vowel lengthening, rising intonation, coalescence, and certain voice qualities (p. 85).” 37

(2015), demonstrated the highest development in the post-test. In the pre-test, she mainly used masu forms regardless of to whom she was speaking. However, in the post-test, the student used plain forms predominantly when she was talking with another student, but she maintained predominant use of masu forms during the three-way conversation. In other conversation data, she described that she learned the importance of using masu forms through a part-time working experience at a small café. She was scolded by a senior worker because she did not use masu forms when she was talking to the senior worker. This personal experience of awareness- building, helped by a Japanese co-worker, is reported to have resulted in her developing an ability to shift speech style more appropriately over the six months.

Although it is limited, the review of research literature on L2 speakers’ style shifting and development of style shifting in Japanese EdA contexts suggested that EdA learning contexts seem to facilitate the increased use of plain forms. However, utilizing style shifting as a communication strategy is not the same thing as doing it more often, and shifting was still difficult for L2 speakers in general; for example, some EdA students overused plain forms upon their return. Taguchi (2015) and Iwasaki (2011) suggest that social contacts during EdA programs may have an effect on learning style shifting. Their studies also support the importance of investigating the process of EdA gain discussed in Chapter One. In the next section, therefore, we shift focus to the process of EdA learning by first reviewing research that demonstrates a strong relationship between social network and EdA gain. We then discuss how we can utilize social networks as a means of understanding the process of EdA learning.

2.3. Social networks as a means of understanding the process of EdA gain

Sociolinguistic studies that investigated the relationship between social networks and speech style by L1 speakers have revealed that social patterns and language uses are closely related (Milory, 1987, p. 109 etc.). Studies on L2 acquisition also have suggested social interaction with locals have strong relationships with learners’ acquisitions of appropriate speech style (Siegal, 1995, p. 240) and target language pattern of phonological variation (Terry, 2016, p.

20). Those studies indicate that investigation of EdA social networks can provide insights on what types of EdA experiences facilitate or hinder EdA gain.

2.3.1. Social networks as a strong predictor of EdA gain

Baker-Smemoe et al. (2014) investigated predictors of L2 gain by 102 American students who participated in EdA programs in Mexico, Spain, France, Egypt, Russia, and China. Their findings suggested that the amount of the target language use does not have a significant impact on their language acquisition, but social networks with L1 speakers of the target language was a significant predictor. It further suggested that the quality of interaction (e.g. interaction in formal and informal settings, which may have required different style uses, or having deep conversation with close friends) might have a greater impact on EdA students’ oral proficiency gain.

Dewey et al. (2012a) investigated social networks and self-perceived language proficiency gain among 204 American college students who participated in EdA programs in

Japan. Their survey results demonstrate that the dispersion27 of social networks was a statistically significant predictor of self-perceived language proficiency gain. In other words, students who

27 Defined as the “number of social groups in which a learner participated” (Dewey, Bown and Eggett 2012, p. 111). 39

reported that they belonged to more social groups perceived more language proficiency gain than those who reported belonging to fewer social network groups.

Isabelli-Garcia (2006) investigated the relationship between extra-linguistic factors and

EdA gain, which reveals that students with high motivation developed extensive social networks with locals. Strong ties with locals provided rich interactional opportunities in the L2, which facilitates EdA gain. Similar to Polanyi’s study, Isabelli-Garcia’s study also revealed that cultural differences in gender and gender expectations often resulted in a negative view on the part of some EdA students toward the target culture. For example, one male participant named Sam in

Isabelli-Garcia’s study perceived Argentine men as “corrupt, foolish, and disrespectful towards women,” which may have influenced students with this negative view to have weaker social ties with locals.

Zappa-Hollman and Duff (2015) analyzed EdA participants’ social network development by focusing on expected returns (i.e. benefits, often referred to as expressive and instrumental returns28) that the participants obtained through investment in their social networks. They constructed an individual network of practice (INoP)29 to analyze EdA students’ social networks, a heuristic that provides rich descriptions of EdA participants’ interactions and support in different contexts. INoP incorporates a framework of language socialization that factor in the process of linguistic and social practice learning by novice or newer members of a community.

INoP also integrates the framework of social network theory (Milroy, 1987), which holds that the density of an informant’s social network is an important indicator of a certain linguistic feature

28 Examples of instrumental returns are wealth, power, and reputation. Examples of expressive returns are physical health, mental health, and life satisfaction (Lin, 2001, p. 21). 29 See Section 4.7.2 in Chapter Four. 40

(e.g. use of local dialect). In addition, INoP adapts construct of a community of practice (CoP), which aims to understand the process of community participation by newer members.

Differences, however, are the focuses of each analysis. The focus of INoP is on an EdA participant in relation to his or her multiple contexts and personal relations, whereas, the focus of

CoP is on characteristics of a community that newer members/L2 learners (are willing to) involve in. Because EdA participants’ learning of linguistic and social practices of L2 is often not limited to their social ties within a single community, the CoP analysis alone may not provide us complete data relevant for EdA gain. On the other hand, because the analysis of INoP considers an EdA student’s social ties within and beyond a social group rather than focusing on one particular community that the EdA student belongs to, it is a useful tool to understand the overall EdA student’s social-network development, which often varies greatly among individual

EdA students.

Their analysis revealed that Mexican students who participated in an EdA program in

Canada received both types of returns primarily from other Mexican students. The finding that

EdA students construct strong social ties with other EdA students of the same nationality has been well reported in the other studies, and such ties (or contacts) are often considered not as beneficial as ties with locals in terms of EdA gain because they often resulted in the use of EdA students’ L1 rather than L2. However, Zappa-Hollman and Duff (2015) revealed that such ties with other EdA students could provide not only the psychological support (expressive returns) to

EdA students but also academic support (instrumental returns).

They also reported that strong social ties often connected with an EdA student through two or more ties (i.e., a multiplex social network), and the EdA students described these ties as

most beneficial (357). By focusing on EdA students and their social ties in different social groups, Zappa-Hollman and Duff (2015) revealed the importance of constructing social ties not only with locals but also with other EdA students for both instrumental returns (i.e., academic support) and expressive returns (i.e., affective support).

In order to gain better understating of the process of building social networks and intercultural relationships, the next section focuses on research literature that investigates the difficulties in forging intercultural relationships, especially during EdA.

2.3.2. What is happening in an L2 learner’s intercultural communication during EdA

programs?

As discussed in Chapter One, aspects of the target culture that American EdA students struggled with during their stay depend on the cultural differences between first culture (C1) and second/target culture (C2).

As discussed briefly in Chapter One, Polanyi (1995), who investigated American students’

EdA learning from their perspectives reports that while the male American students often had positive experiences while interacting with female Russian speakers, the American female students experienced unpleasant interactions with male Russian L1 speakers, which often resulted in ending their friendships. To confront such unpleasant encounters with male native speakers, many of the American female students learned by themselves or were taught by female

L1 speakers to be more blunt and frank with male native speakers. Such attitudes helped the

American female students to be understood without knowing “a big vocabulary and a lot of expression.”(p. 286) However, in Brecht’s study (1995), learning such communication skills did

not result in the female students being rated at a higher oral proficiency level. By investigating

EdA learning from the students’ perspectives, Polanyi revealed what was actually happening during EdA programs in Russia, and why the female students did not improve their oral proficiency as much as their male counterparts, which led to gender being a statistically significant predictor of language gain during EdA in Russia, in Brecht’s study (1995).

By investigating the perspectives of four American EdA students in France, Wilkinson

(1998) revealed that EdA students’ discourse and expectations toward the French L1 speakers were influenced by their previous classroom experiences. In addition, their frustration with

French L1 speakers often stemmed from intercultural problems rather than linguistic misunderstandings. Failing to acknowledging the cultural differences (i.e. differences in smiling at strangers in the US and in France), the EdA students often reached the conclusion that “the

French hate Americans” or “the French are cold (hostile/unpleasant/arrogant/despicable”

(Wilkinson, 1998, p. 30). Wilkinson cautions us that there are considerably more complicated processes of EdA learning than what is suggested by outcomes measured with proficiency test scores, and we must not conclude these EdA students who do not meet a program’s descriptions are “lazy students”, based on just exit test results.

Studies that investigated the perspectives of EdA students in Japan reveal that EdA students struggle with hierarchy in groups when building interpersonal relationships with

Japanese people (Burn, 1996; Iwasaki, 2011). The EdA students in Burn’s study tried to fit into the hierarchy of the local community they belonged to during EdA, such as athletic clubs at their host institution. However, Japanese people treated them differently from other members of the community, often positioning EdA students as “foreigners” in the hierarchy (Burn, 1996, p. 160).

In addition, these studies reveal that EdA students chose not to participate fully in the customs or practices of the community even when having been invited to do so. Instead, EdA students chose to be observers rather than members of the community in such situations (Burn, 1996, p. 208).

This behavior is something that the idea of Third Space helps to account for, as discussed in

Section 1.4 in Chapter One.

Western students’ struggles with appropriate Japanese speech style also often stem from cultural differences in perception of hierarchical structures in a society. Siegal (1995) argues that the learners’ own identities and their perceptions of C2 and L2 seemed to affect their use of sociolinguistic features, specifically humble and honorific language in Japanese in the beginning of their EdA experience (p. 240). Some of the learners often did not choose to use the polite speech styles (i.e. both the honorific and humble styles) in Japanese despite their knowledge of speech style and how important is it to use such language in formal communication. For example, at the beginning of her stay in Japan, a Hungarian student in Siegal’s study perceived Japanese women’s use of the humble style as “so (too) humble, and (she doesn’t) want to be that humble”

(p. 234). However, as the same student encountered many Japanese men with higher hierarchical status also addressing their interlocutors by using the polite styles, she began to see the functions of the speech style in Japanese society from different perspectives and started utilizing them in her speech after studying in Japan for a year (p. 239).

Similar to Siegal’s study (1995), Iwasaki (2011) also reported that the learners’ experiences in interacting with locals during their EdA in Japan affect their choices of speech styles (p. 97). Although some participants acknowledged the importance of speech styles in

Japanese communication, especially in environments where hierarchical relationships were

obvious, the gaps between their social identities in C1 and the norms in C2 seems to constitute obstacles for them to use masu forms. Participants who had more diverse interactional opportunities with locals maintained the use of masu forms upon their return (Iwasaki, 2008 &

2011). Siegal and Iwasaki’s studies suggest that through being exposed to the C2 norms and carefully observing how L1 speakers use such language and what kinds of effects it has on ongoing communication, L2 learners can reconstruct their perceptions of L2 and their identities, which facilitates gains in sociolinguistic competence.

On the other hand, unlike Iwasaki and Siegal’s studies, which reported somewhat hesitant use of masu forms and polite language by participants in the beginning of the EdA program,

Burn (1996) reports that the EdA students’ language was perceived as “too formal for their

Japanese friends” (p. 226), and also mentions the inappropriate use of the masu forms or keigo, which becomes a barrier in talking with Japanese college students. Those differences are probably due to differences in their earlier instructional experience at home, which influences differences in EdA students’ use and views of the speech style. Furthermore, Burn’s study suggests that there were cultural differences in conversation topics that North American students and Japanese students considered appropriate for a first meeting. Japanese students’ questions in the initial interactions on topics considered private, such as having a boyfriend or girlfriend, also made the EdA students feel uncomfortable.

Regarding the types of relationships that EdA students and Japanese local people developed, Iino (2006) discusses power relationships between EdA students and their host families and proposes two approaches to analyze their interactional styles based on his observation (pp. 161-165). The first approach is the cultural deficiency. This is where EdA

students play the role of care-receivers and deficient participants of the C2 whereas host families act as care-providers and ideal members of the C2. For example, many of Iino’s EdA students complained, “they were treated like babies and dolls” by their host families. Some of them chose to take on the role of “a pet or gaijin ‘a foreigner’” (p. 163) because it is “the best tactic” (p.

162) to please host families, while others were frustrated by such treatment by the host families, which is another example of a Third Space (TS), but in their TSs, the EdA students in Iino’s study seem to just adapt to the host families expectations instead of negotiating or co- constructing the expectations.

The second is the two-way enrichment approach. This is where the power relationship is considered as not fixed or one-way, but dynamic and two-way. In addition to the role of information receiver, Iino finds that EdA students also played the role of providers by teaching right pronunciation of English words. According to Iino, there are three conditions that can lead to a successful two-way cultural exchange in a home stay setting (p. 164). The first condition is an interaction of cultural exchange must not change the power relationships, and it must non- competitive in nature. The second condition is fewer constrains in terms of time and energy. The last condition is that the both sides, the EdA student and host family, are interested in each other’s culture as well as language. Although Iino discusses such conditions regarding home stay contexts, these three conditions for successful cultural exchanges can be applied to other EdA contexts as well. These three components can be also applied to TS that characteristics includes productive and beneficial to both sides of interlocutors.

Among the studies mentioned so far, Burn (1996) and Iino (2006) are the only studies that include both EdA students’ and local people’s perspectives on intercultural relationships in 46

the C2. Because positive relationships cannot be built without multilateral understanding and acceptance, investigating the local people’s perspectives is crucial to understand what types of pedagogical support are needed to overcome the struggles that EdA students encounter while interacting with locals. However, research that includes both perspectives is still limited.

2.4. Culture in L2 learning and teaching

The aforementioned studies that document EdA students’ struggles indicate that the culture that EdA students often struggle with can be categorized as what Hammerly (1982) calls behavioral culture. He categorizes culture into three types, namely, achievement culture, informational culture, and behavioral culture. Achievement culture refers to an aspect of culture that represents “artistic and literary accomplishments of the society” (p. 515). Examples of achievement culture in Japanese society are Noh and tea ceremony. Informational culture refers to “the information or facts that the average educated native knows about his society,” (p. 513), such as information on political systems, history, population, industry etc. Behavioral culture refers to “the knowledge that enables a person to navigate daily life” (Walker, 2010, p. 13) with others in the culture. Examples of such culture would be first apologizing when receiving a gift in a formal setting, or using the addressee honorific-forms when talking to someone who holds higher social status in terms of age and rank.

For L2 learning and teaching whose goals are effective communication, Hammerly argues that cognitive knowledge of behavioral culture is not enough; an L2 learner must have performative knowledge of it. Furthermore, Walker (2000, p. 13) and Noda (2014) argue that from an L2 pedagogical standpoint, Hammerly’s three kinds of culture should be considered

“discourses” rather than different types of culture. In other words, what kinds of culture is being observed or regarded depends on the kind of context in which it is being engaged with. Above, I cited Noh as the dramatic art form as an example of achievement culture. However, for a Noh performer, how to walk and move his arms and hands in Noh performance is behavioral culture.

Noh performers rely on the music and chant to determine when to move this way or that, dance, appear or disappear from the stage. In this respect, the music and chant, as culture, have an informational utility for the principal Noh ‘performer’ or shite. However, even thought they share the same time and place with the performers, these different aspects of Noh culture, including the performers’ movement and the music are primarily achievement culture for the audience.

Some of the differences in behavioral culture seem obvious to the EdA students, yet it does not guarantee that they incorporate such behaviors in their L2 communication as discussed above (Burn, 1996; and Iwasaki, 2011). Others are rather difficult to recognize as cultural differences, such as in the case of American EdA students in France (Wilkinson, 1998). Walker’s categorization of different types of behavioral culture is useful to understand such EdA students’ struggles. Walker (2000) categorizes the behavioral culture into three categories, namely, revealed culture, ignored culture, and suppressed culture, depending on how native people in the target culture treat certain aspects of their own culture when they communicate with people from different cultures. Revealed culture refers to the aspects of the culture that native speakers are eager to share with people from other culture. Most of revealed culture is positive in nature, such as, people value politeness and harmony in Japanese society.

By contrast, suppressed culture refers to an aspect of culture which people in the culture are not willing to share with outsiders. The people in the culture consider most of the suppressed culture as something negative and to be ashamed of. For example, the social phenomenon of karoosi (death due to overwork) in Japanese society is considered as suppressed culture. When

Japanese people talk about their own culture to foreigners, this topic is not typically brought up in daily informal conversation. Such topics are often discussed in Japanese textbooks for advanced level students or culture courses taught in English.

The last category is ignored culture, which refers to the cultural aspects that people of the culture believe are universal or a common sense until they encounter people from a different culture doing things differently. Christensen and Warnick (2006) point out that although culture and language classes often focus on revealed culture of C2, it is ignored and suppressed culture that language students often find confusing when they are interacting with people in C2. For new and even for experienced L2 teachers who identified themselves as L1 speakers of the language, teaching ignored culture is often difficult because it is, as the term indicates, “ignored” by both

L1 and L2 speakers of the culture. Because people think it is common sense to do things in a certain way, so when a person does not do it this way, they may perceive the person as without common sense instead of considering it as a cultural difference. This is why American EdA students in Wilkinson’ study (1998) struggle with the French people “not smiling” when they met on the street. They believed that smiling at a stranger was a universal way to express friendliness, therefore, when French people did not smile back at them, they concluded that “the

French are cold (hostile/unpleasant/arrogant/despicable)” (Wilkinson, 1998, p. 30). Wilkinson’s study, among others, indicates how difficult it is to recognize ignored culture during an EdA

experience and reminds us of the danger of not recognizing people in the target culture’s behavior as cultural difference.

Investigation of EdA students’ perspectives provides crucial insights on EdA learning, especially as to what aspect of a C2 EdA students struggle with. As mentioned previously, correlation between social networks with locals and EdA gain indicates that understanding of the

C2 plays an important role for EdA gain. We also need to keep in mind that even if EdA students acknowledge cultural differences, they may still decide not to participate in some aspect of the target culture (Burn, 1996; Iwasaki, 2011), which is probably due to differences in norms and values between the first and second habitus (Bourdieu, 1977). Considering these points, we now focus on development of personae in C2, as a way to overcome such struggles.

2.4.1. Developing personae in C2 and L2 learning

Chapter One discussed why it is difficult for many foreign language learners to learn new habitus in cross-culture contexts (see Section 1.3. in Chapter One). It becomes more difficult when social differences between the C1 and the C2 are greater (Schumann, 1979). These difficulties often result from ideas such as “behaving like Japanese people (e.g. humbling oneself) does not fit my identity,” an attitude similar to the one cited above from a participant in

Siegal’s (1995) study. Because a different culture has different values, this way of thinking can inhibit learning of C2 and L2.

In order to overcome such dilemmas, the idea of a persona, defined by Walker and Noda

(2010), is practical. Walker and Noda define persona in L2 learning as “the personal information that the learner is willing to commit to the learning experience” (p. 33). They go on to state that

“the persona of a learner can vary considerably from one learning environment to another: the physics whiz may be hesitant and unsure in the language class. Not confusing the persona of the individual with the individual him- or herself is important because the persona can change rapidly in a period of language study, a change that is much faster than a change in the personality of an individual” (p. 33). They differentiate persona from identity in that the construction of persona is usually accomplished in a much shorter time than that of identity, which is really under construction until death. Identity is a compilation of personal history and develops over a much longer time span, though it is also dynamic and constantly changing. Although they discuss the importance of persona in relation to L2 learning and teaching, the concept of persona is not particular to L2 learning. For example, Karen Kelsky

(2015), an author and career consultant for academic jobs, argues that being “yourself” should be the last thing a job applicant should think about, especially for ABD30 students, in order to get a tenure-track job in American academia. She points out that graduate students are often “insecure, defensive, paranoid…” but in order to get a job, the job seeker has to develop “a professional persona” who is a “full-fledged adult who demonstrated a tightly organized research program…” even though it may not be true to “yourself” (Chapter Thirteen: Why “Yourself” is the Last

Person You Should be). She continues by suggesting that with repeated practice, a job applicant becomes more comfortable playing the professional persona, and a job applicant must “fake it till you make it” to achieve the goal. Although it is not related to L2 or C2 learning, Kelsky’s argument demonstrates the same point made by Walker and Noda: it is crucial to develop an appropriate or viable persona for a given context in order to successfully negotiate it, even when

30 All But Dissertation 51

a person does not feel comfortable playing the persona; with repeated practice, playing such a persona will become second nature.

Now, let us apply the concept of persona to the examples of the EdA students discussed in the aforementioned studies in order to understand how persona plays a role in learning of C2 and L2 during EdA experience. To the American students in Wilkinson’s study (1998), it might have been important to construct a friendly persona, and smiling at strangers on the street might have been a way to construct such a persona. However, as the study demonstrates, the same persona does not translate the same value or personal information in the French culture. Such behavior can be considered as flirty instead of friendly. Similarly, playing a polite persona expressed by using polite language (including both humble and honorific forms) may not have the same value in the participant’s C1 as it does in Japanese culture. Such personae often play an important role in getting things done without causing troubles in a Japanese business setting.

Playing an appropriate persona is necessary to achieve desirable results both in academic job searching as well as in intercultural communication during an EdA experience. If a learner cannot separate personae in L2 learning from their own identities in C1, it can become extremely difficult to have successful communication and forge positive intercultural relationships in C2.

This confusion between desirable personae in C2 and their identities in C1 often causes strife to

EdA students.

2.4.2. Performed Culture Approach as a framework of EdA training

Culture provides the framework to interpret a person’s intentions as well as the persona

(i.e. personal information) that the person constructs (Walker & Noda, 2010). Within a cultural

framework, people’s behavior and words create meanings. The meaning of performed language is heavily dependent on the context where it’s performed. When interacting with someone, we invariably refer to features that we take to be “there” in the performance context—spatial, temporal, social, affective, knowledge (including shared memories)—by pointing to them in a way that will be understood, whether with our index figure, gaze, or words like here, there, now, yesterday, I, you and so on. What such words refer to depends on where, when, by and to whom they are said. This referring to something relative to our own present position or condition is indexical. For it to work in communication, indexical reference depends on communicators’ mutual appreciation of the context in which they are interacting. Indexicality, then, is “a phenomenon associated with the situatedness of human experience in the creation of meaning”

(p. 21). The various functions of masu and plain forms discussed in Section 2.2 also are a matter of what the two forms index, or “point to”—at a minimum, a stance taken vis-à-vis one’s addressee. Indexical reference to the multiple parameters of any communicative performance context helps “anchor” the interaction for those interacting, as it confirms, maintains or adjusts the context as we proceed, and topics, referents, attitudes, feelings and information come and go.

In indexing affect, knowledge, and interpersonal stance, people rely not only on word choice and grammar, but also on how they deliver those words—prosody, voice quality, gestures, posture, eye contact, facial expression, and more. Indexical meaning is a very “local” business, and it is as complex as it is fundamental to communication and socialization. The complexity of the indexicality as well as acquisition of it is often discussed in L2 research, but the question is, how can we teach the indexicality of the Japanese language, as Japanese behavioral culture, in ways that will guide our learners toward success in intercultural communication?

By focusing on performative knowledge of behavioral culture, Walker and Noda (2000) have developed the Performed Culture Approach (PCA) as an L2 pedagogical framework for

East Asian languages. In PCA, performance is an important concept for teaching L2 as well as the assessment of L2 learning (Noda, 2007). Performance consists of five components, time, place, role, script, and audience. In other words, indexical reference in the language is, to a significant degree, reference to, or expression of, these five parameters of the interactive situation. In order to demonstrate how these five components affect overall performance, let us compare examples of several different performances in Japanese. The crucial point here is, although these are different performances, the speaker’s intention in these performances is the same, i.e., to apologize for her late arrival for an appointment.

Alison, an L2 learner of Japanese, is late for an appointment with a friend who is waiting her in front of Tokyo Station. Her way of apologizing would differ depending on how long she has kept her friend waiting. The season, falling under the component of time, would also affect her apology performance, which might be slightly shorter if it is a nice spring warm day than cold wintertime. Regarding location of the performance, Alison’s performance would likely be different whether she is meeting her friend’s at her friend house or a public location where her friend feels uncomfortable or unsafe waiting for Alison by herself. In the latter case, Alison may send an apology message via text her friend in advance of her arrival and suggest meeting somewhere her friend feels safer. After her arrival she would apologize to her friend again. Such differences may be clear in the cultural script, such as using plain forms with affect keys to indicate closeness to the friend in the former case, or using masu forms predominantly to express her seriousness in her speech act of apology in the latter case. In terms of role, if Alison’s role

changes to a subordinate and her friend’s role changes to Alison’s supervisor, not only would her speech style change noticeably but also her behavior would dramatically change. In this performance, Alison would likely bow deeply multiple times to her supervisor while apologizing. Depending on her reason for being late, she may add an explanation. However, in a business setting, an explanation with apology that is considered as under the control of the speaker is often taken as “excuse (iiwake)” in Japanese (Yao, 2017, pp. 88-89). Therefore, in such a case, Alison would not give explanation of her late arrival.

All the examples of apologies above make it seem that changes in time, location, and role affect cultural scripts. However, interpretations (or indexicality) of the same script

(e.g. sumimasen) can also differ in various performances. For example, the same script of

“sumimasen” can be interpreted as “I apologize for not turning in my assignment, I want attention (service), I want you to move over to make space for me to sit, or thank you” (Walker

& Noda 2010, p. 25) in Japanese culture depending on time, location, role and audience.

Finally, the last component, audience, refers to a person or people who do not participate in the conversation, but the conversation participants have them in their minds when they produce the discourse. Such awareness of the audience affects the speakers’ discourse in many ways. For example, the discourse of a Japanese graduate student sending an email to a fellow graduate student differs from when the email is also sent to a professor in terms of the speech style, including the choice of a first person pronoun appropriate to the context.

Audience may not be as clear as the other components but can be considered the most important component in performance (Walker, 2017). For example, Tobaru’s (2014, pp. 45-46) findings suggest that different cultures might have different degrees of acknowledging who their

“audience” is in a context that seems otherwise identical. In the case study, three American college students who participated in EdA programs in Japan all mentioned the “quietness” in

Japanese public areas was strange to them. They were explicitly told not to talk loudly when on public transportation because it bothered other passengers. One of the participants said that she thought no one would have cared about what she was talking about in the beginning of her EdA experience, but as she spent more time in Japan, she began to think that “you shouldn't be talking about how you didn't shower, the oba-san (middle age woman) standing a few rows down (at a local store) might hear you, or something. It wasn't just the group you are in, it was everyone, if I had a conversation on the bus, everybody was listening in Japan” (p. 46). The notion of other people listening to what she said clearly affected her topics of conversation in L2. Such cultural awareness is certainly important for an L2 learner to build interpersonal relationships. It helps them to avoid topics that might put their interlocutor or people around them in an awkward situation.

The complexity of indexicality in communicating in an L2 can overwhelm not only L2 learners but also L2 teachers.31 However, keeping in mind that the meaning of words depends heavily on their context of use, and that it is the culture that creates contexts (Walker, 2000, p. 9), indexicality in the L2 becomes vastly more teachable and learnable if we are aware of the five components in communicative (including L2) performance. In other words, without defining these five components clearly in C2 performance, an L2 learner cannot help but fall back on his

31 It also falls to textbook writers to decide which part of indexicality of the language, grammar, or speech style should be focused on and what order to present them in, so that they are teachable and learnable within a page limit. 56

feel for indexicality in his C1 as he tries to make sense of the L2, which often results in erroneous and negative perceptions (Wilkinson, 1998; Siegal, 1995, etc.).

2.5. Conclusion

With the aim of providing a detailed background for the current study, this chapter reviewed research literature on Japanese style shifting, the relationships between social network development and L2 learning, and the Performed Culture Approach as a framework for EdA training.

Review of such research literature and the discussion in Chapter One demonstrates the importance of incorporating these aspects into research investigating pedagogical support to maximize EdA students’ learning, however, such studies are still limited, especially for EdA contexts in Japan. To address the issues elucidated in this chapter, the current study first investigates the difficulties in building intercultural relationships by conducting fieldwork research at a four-week EdA program in Japan. In the next chapter, it presents the methodology and findings of the fieldwork research. As pedagogical implications of the fieldwork research, it also discusses EdA training developed based on the findings, which is used in the second phase of the study discussed mainly in Chapters Four, Five and Six.

Chapter 3: Understanding the Difficulties in Building Intercultural Relationships in an Education

Abroad Context

3.1. Introduction

Although many scholars call for pedagogical support for EdA students to overcome the struggles in their studies, few studies actually consider the effectiveness of such pedagogical support. Studies done by Berg et al. (2009, pp. 15-16), Cohen et al. (2005), and Davidson (2010, pp. 20-22) report that pre-departure training or orientation that included cultural components have a positive effect on EdA learning. However, the nature of “cultural components” (Berg et al., 2009, p. 15) included in these training or orientations are unclear, and such pedagogical intervention is often primarily for students who participate in EdA programs in European countries.

Thus, the current study aims to fill the gap by first investigating EdA students and local people’s struggles in building intercultural relationships in an EdA context in Japan. To investigate the difficulties in forging intercultural relationships, three research questions guided this phase (Phase One) of the current study:

1. What kinds of difficulties do local Japanese students encounter when interacting

with American EdA students during a short-term EdA program in Japan?

2. What kinds of linguistic and cultural difficulties do American EdA students

encounter during a short-term EdA program in Japan?

3. What kind of Third Space emerges when American EdA students and local Japanese

students interacting with each other in a short-term EdA program in Japan?

3.2. Methodology

As mentioned above, because this phase of the study aims to identify the difficulties in building intercultural relationships in an EdA program in Japan in order to design EdA training to address these difficulties, it focuses on the process, rather than the outcomes, of the building intercultural relationships.

3.2.1. Setting

The current study took place during a four-week EdA program at a private university in western Japan, which was sponsored by a university in the Midwestern U.S. This was a faculty- led EdA program, and all the EdA students were from the same university in the US. In order to participate in the EdA program, the students had to have completed at least the first semester of a level 3 Japanese language course at the home institution or have an equivalent level of Japanese language proficiency. The objectives of the program, as described in the 2017 program handbook, was to “deepen [the students’] understanding of the Japanese language, people, society, and culture through conducting a research project, class lectures and discussions, personal observations, and service-learning,” (p. 2). The students’ daily schedule clearly reflected this objectives, as described below.

During the program, the students had classes on Japanese culture and society in the morning, taught in English by the Japanese program resident director, who was a faculty member of the home university. In the afternoon, students worked on their own research projects and/or participated in service-learning such as participating in English conversation tables and assisting teaching English to Japanese students at the host university. During the weekends, they went on field trips to historical and cultural sites such as shrines or a sake-brewing factory. The students also stayed with Japanese families for a weekend. Except that one weekend, the EdA students stayed at a university dorm where two local Japanese resident assistants were assigned to oversee and support the EdA students throughout the program.

One of the attractive features of the program to its participants was that each EdA student was assigned a language partner who was a local Japanese student at the host university.

According to the program coordinator at the host university, these Japanese students volunteered to participate in the EdA program as a language partner. The language partners were required to meet their EdA student at least once a week, and Japanese must be the primary language of communication. Japanese language partners and EdA students were matched based on such information as personality, hobbies, academic major, and available times for meetings during the program, gathered from the students prior to the start of the program.

3.2.2. Participants

Eleven American participants of the short-term EdA program participated in the Phase

One of the study. Of the eleven, nine of them were female and two of them were male students.

In terms of ethnic backgrounds, one African American, one Asian American, one Latina

American, and eight Caucasian Americans comprised the cohort. One of the EdA students, a female Caucasian, studied abroad in Japan for nine months when she was a high school student, and another student visited Japan for two weeks in a high school field trip. For other students, this EdA program was their first time in Japan. The EdA students’ purposes of attending the programs were mainly categorized into three: 1) to prepare for an up-coming yearlong EdA program (four students); 2) to alternate to a yearlong EdA program (four students); and 3) to prepare to find a job in Japan (one student).

Because personal relationships cannot be built without mutual understanding and acceptance on each side, obtaining the local peoples’ perceptions toward intercultural relationships is essential to addressing American students’ struggles in EdA training. For this reason, I recruited to participate in Phase One of the study eleven local Japanese people, including nine students who volunteered to be language partners and two resident assistants.

Because the host institution was a women’s college, all of the Japanese participants were females.

The most common purpose for participating in the EdA program among the language partners was to gain intercultural experiences in a domestic environment, which was a much easier alternative for Japanese students who could not afford to participate in EdA programs in other countries for various reasons. Many of them also mentioned that by attending this program as a volunteer language partner, they could put down this experience on their resumes as international experience or as volunteer work when starting job searching. One Japanese student mentioned that she had participated in the same EdA program as a language partner two years before and had enjoyed it very much, so she decided to do it again. All of the language partners had part-

time jobs working at a café or restaurant, which seemed to limit their availability to meet the

EdA students on occasion.

The two resident assistants were former students of the host university. They had just graduated from the university two months before the EdA program started. They were assigned by the host program based on their previous experiences assisting and tutoring international students. They stayed at the university dorm with the EdA students throughout the program and also accompanied them on field trips.

3.2.3. Instrument, procedure and data analysis

Perspectives of both American EdA students and local Japanese students were elicited through individual on-site interviews. Semi-structured questions (see Appendices A and B) were designed to answer the three research questions mentioned at the beginning of this chapter.

When it was necessary, additional follow-up questions were also asked to gain a better understanding of each participant’s perspective on interactional communication.

The interviews were conducted on days and times based on the participants’ availability, and were conducted in a quiet room at the host university. During the interviews, the participant’s native language (i.e. English or Japanese) was used to elicit detailed accounts of their experiences and opinions regarding intercultural communication. Each interview lasted from approximately 30 minutes to 90 minutes. All of the interviews were audio-recorded for later analysis and notes were taken during the interviews.

Although the interviews were the primary data resource collected during this phase of the study, I stayed in the same dorm with the EdA students for the last ten days of the program, and

observed their interactions with the resident assistants and language partners. In addition, I talked with the program coordinator of the host university and a host family member to gain insight on the EdA students’ interactional experiences. These data collected through observation and informal talks were used to supplement data to support the interview data.

All the interviews were transcribed for data analysis. After all the interviews were transcribed, domain analysis was conducted to search for a larger unit, called domain, of the informant’s cultural knowledge by looking for vocabulary items that are similar to each other

(Spradly, 1979, p. 107). Searching for semantic relationships among words and phrases was essential to understand the informants’ perspectives because their concept or definition of

“friend,” for example, could be different from that of the researcher and among the participants

(Spradly, 1979, p. 109). In order to understand what they mean by “friend” from their perspective, follow-up questions, such as “why do you think your language partner is your friend?” were asked. Furthermore, “friend” is a vague word to describe a relationship, and the term itself cannot demonstrate how close the relationship might be. Therefore, in order to know what friend m to each participant, I used a scale similar to Figure 3 below, and asked the participants to indicate where they positioned their language partner on the scale.

Figure 3. Relationship scale

The white part on the left indicates that they do not consider their language partner to be a friend. The darker part on the right indicates that they consider their language partner to be a

friend, and 10 on the further right of the scale indicates that they consider their language partner to be a best friend. While there are some disadvantages to using such a scale, (e.g., the space between 0 to 5 and 5 to 10 cannot possibly be equidistant), it nonetheless provided a uniform means to talk about the participants’ perspectives on forging intercultural relationships.

3.3. Findings and discussion

Domain analysis revealed two domains for the first research question, and one for the second and third research questions respectively. Before discussing these findings, this section first presents the findings regarding the relationship scale discussed above in order to understand how Japanese and American students saw their relationships.

3.3.1 Shared perspectives toward the relationships

Figure 4 shows the responses to the questions about the level of friendship the 20 student participants (eleven American EdA participants and nine language partners) felt they had with their respective language partners.

Figure 4. Relationship scale with participants' responses

White circles represent the eleven American EdA students’ perspectives and the circles with grids represent the nine local Japanese students’ perspectives.32 Although there are a few outliers, all the other responses fell somewhere between 0 and 5 on figure 4. All except one EdA students’ answers indicate they considered their language partners as friends. In addition, all except one

Japanese student said that they considered their American partners to be friends, as represented by the positions of the circles with grids on the scale. This indicates that there was not much difference between how the local Japanese students and the American students saw their relationships. This suggests both the Japanese and American EdA students considered their friendships to be somewhat weak and at the initial stage. Nevertheless, it is clear that there was a shared expectation that both American and Japanese students interacted with each other as friends.

Before moving on to other findings, let us discuss the outliers on the scale. A white circle on the left part of the scale (i.e. acquaintance) in the Figure 4 represents a response from an

American student’s response, which indicates she did not see her relationship with her Japanese language partner as that of friends. When I asked the student how she saw her language partner, she responded:

I look at her and think […] maybe she sees me as a younger sister or something like that, she does take care of me which is very nice, she is very responsible person [...] She just takes care of me, and that’s it.

This account clearly indicates that this EdA student perceived there was an unequal power- relationship between her and her Japanese language partner. This type of relationship is similar

32 Because the two resident assistants were not language partners, I did not ask them the questions. Therefore, the circles with grids in Figure 4 represents responses from the nine local Japanese students. 65

to what Iino (2006) refers to as culture deficiency approach where EdA students play a role of

“care-receivers” and “deficient participants” of C2 (161-163). Although she used the term

“younger sister,” she used it to emphasize her language partner’s responsibility to take care of her rather than to suggest their relationship is as close as that of a sibling relationship. Her overall account clearly indicates that her relationship with the language partner was not as close as she wanted it to be. She saw her language partner as more of someone who took care of her rather than someone with whom she felt intimate.

Similar to the American EdA student above, one Japanese student showed some hesitancy with her answer and pointed at the zero on the scale. When I asked her reasons, she responded that although she enjoyed spending time with her language partner, their relationship was not yet as close as what she would expect in a relationship with a friend because she could not talk about topics freely with her American language partner. For this student, a friend meant someone she could talk about anything, and this might be due to the limited time they shared together, her language partner’s limited language proficiency, or both.

On the other hand, one American EdA student considered her language partner to be a very close friend, but it seemed that her definition of friend was broader than the other participants. For example, when I asked the reasons, she mentioned that because her language partner helped her, she considered her language partner as her good friend. However, further discussion of her interaction with her language partner revealed that they only met twice during the program. Both of these meetings were coordinated by the program. More specifically, their first meeting was the welcome party held by the host university, and the second one was the formal farewell party hosted by the host university at the end of the program. When this student

introduced me to her language partner at the formal farewell party, their interaction was very limited, such as a casual greeting in Japanese. The American student told me in Japanese, “[name of her Japanese language partner]-san wa totemo subarashii paatonaa desu! (X-san was a very good partner!)” and the Japanese language partner was just smiling but did not say anything. I exchanged a couple words with the Japanese language partner, but the American student went to talk to other American students. Her accounts of her interaction with her language partner and my observation suggest that this student’s definition of friend may have been broader than the others. Unfortunately, I did not receive consent from her Japanese language partner, so the data from the Japanese student’s perspectives on their relationship was not available.

As mentioned earlier, these responses are considered as outliers in this context. The majority of the responses indicate that there were shared perspectives toward their relationships

(i.e., friends) between the Japanese students and American EdA students. However, despite the shared perspectives, other findings reveal gaps in the perspectives and how they had behaved and positioned each other. I argue that these gaps resulted in struggles experienced by the local

Japanese students and American EdA students. We discuss these findings in relation to the three research questions below.

3.3.2. Research question 1: What kinds of difficulties do local Japanese students

encounter when interacting with American EdA students?

Two domains were revealed regarding the research question: 1) American students’ reactive behaviors; 2) and gaps between Japanese students’ images of American people and the actual EdA students they met. This section discusses these two findings respectively.

3.3.2.1 American students’ reactive behaviors

Regarding the first domain, when the Japanese participants were asked about their struggles in interacting with the American EdA students, most of them described the American

EdA students as “ukemi sugiru (too passive)” “ukemi na tokoro ga aru (they are sometimes passive)” in term of initiating meeting up or daily communication. In the excerpt below, a

Japanese local student and I (i.e., interviewer) are talking about her daily communication with her American language partner.

Excerpt 4: Discussing daily communication with a Japanese student Interviewer 連絡があったことも含めて、どちらから？ Renraku ga atta koto mo fukumete dotira kara? (Who initiated meeting up and contacting via phone?) Japanese student 私からの方が多かったですね。 Watasi kara no hoo ga ookatta desu ne. (I did more (than my language partner).) Interviewer (アメリカ人学生の名前)の方から進んで連絡とったことは？ [Name of her language partner] no hoo kara susunde renraku totta koto wa? (Were there times that [name of her language partner] contacted you?) Japanese student それはなかった… Sore wa nakatta… (That, it didn’t happen.)

In the excerpt, the Japanese student mentions that she had initiated all of the communication with her language partner (i.e. an American EdA student). Other Japanese students also made similar comments. The interviews with both Japanese and American EdA students revealed that they met up three to five times during the four-week program on the average, but except one initiation by an American student, all the meet ups were initiated by the Japanese students. The Japanese students also mentioned that the American students gave unclear answers when they asked what the American students wanted to do, or where they wanted to go. This struggle was also shared 68

by the two residents who spent most of their time with the EdA students during the program. One resident assistant said:

なんか結構、京都に行くことは、事前にスケジュールはされてましたよね？でも結構みんな、行くってだけになっていて、[…]33下調べというか、ソオゆうのが少なくって、[…]「どこか行きたいところあるの？」って聞いてみても、なんか結構「わからない」「なんでもいい」っていう子がすごく多いんですよ。[…]これは私の勝手なイメージなんですけど、アメリカ人は「イエス」と「ノー」で答えるとかってあるじゃないですか。と思っていたので、あ「行く、いかない、どっち？」できいてみたら、「わからない」「あれ、ちがう… 」と思って、多々悩むことがありましたね。 Nanka kekkoo, kyooto ni iku toka wa, jizen ni sukejuuru wa saretemasita yone? Demo kekkoo minna, iku tte dake ni natteite […] sitasirabe to iu ka, soo yuu no ga sukunakutte, […]‘doko ka ikitai toka tte aru no?’ tte kiite mitemo, nan ka kekkoo ‘wakaranai’ ‘nandemo’ tte iu ko ga sugoku ooi n desu yo. […] kore wa watasi no katte na imeeji nan n desu kedo, amerika-jin wa ‘yes’ to ‘no’ de kotaeru toka tte aru ja nai desu ka, to omotte ita node, ‘iku, ikanai, dotti’ de kiite mitara, ‘wakaranai,’ ‘are, tigau…’ to omotte, tata nayamu koto ga arimasita ne. (Well, pretty much, the plan that they were going to Kyoto was previously scheduled, right? But everyone pretty much had an attitude of ‘just going there’ […] It seemed that (the EA students who did) research on (where they wanted to go or what they wanted to do in Kyoto) beforehand were very few. […] Even when I asked ‘do you have anywhere in mind you wanna go?’ they often answered like ‘I don’t know’ or ‘anywhere.’ The answers like that I got a lot. […] This might be only me thinking like this but, isn’t it that American people usually answer questions with ‘yes’ or ‘no’? Because of that, I asked a question like, ‘are you going or not, which choice?’ But, when I got an answer like ‘I don’t know,’ I thought ‘what, it is different (from what I expected).’ There were times that I was confused (because of that).)”

In this account, the resident assistant was describing her struggles with the American EdA students’ reactive behaviors that she encountered during a field trip to Kyoto. During the Kyoto field trip, free time was scheduled after vising some historical places as a group. The EdA students were expected to utilize the time to explore the Kyoto area based on their own interests.

However, as it was mentioned in the excerpt above, most of the EdA students did not have

33 “[…]” indicates ellipsis of some contents, and “…” in dictates a pause in the utterance. 69

particular plans for the free time, and relied on the resident assistants to decide what to do during the free time for them. Another resident assistant mentioned similar struggles when the American students said ‘tuite iku (we will follow you)’ when she asked where they wanted to visit in similar situations. In addition, the account above indicates that this resident assistant had a stereotype of American people being decisive, but the actual American EdA students she interacted with were somewhat opposite of what she imagined, which made her more confused about the American EdA students’ reactive behavior. This kind of image gap is discussed further in a later section.

When I asked the American students why they did not take initiatives in interacting with the Japanese language partners, most of the students mentioned that they were too busy with assignments, but one student indicated her fear of being rude. She said “because I don’t know the area well enough…so if I suggest something that Japanese people know it’s bad, Japanese people don’t say ‘no’ even if they know it’s bad.” The American student’s feeling of being “foreign to the area,” and their image of Japanese people, being polite by not expressing their true feelings seems to have resulted in the American students’ reactive behavior toward interacting with their language partners.

By applying Iino’s two approaches of power relationships discussed in Chapter Two, the

American students’ reactive behaviors clearly indicates that they positioned themselves as “care- receivers” and “deficient participants” of the C2 (Iino, 2006, pp. 161-163). Considering the

Japanese students’ role as “language partners34,” one might assume that EdA students positioning

34 Originally the term was “conversation partners,” but over the years, the term was changed to “language partners” by the host university. 70

themselves as care-receivers is natural because the term “language partners” indicates someone who provides helps for EdA students’ learning of the language as well as the culture35.

Data from the observation also suggests the unequal power relationships between the

American EdA students and the Japanese students. For example, the EdA students held a pizza party at the dorm and invited their language partners to express their appreciation for their helps during the program. In this situation, the expected role of the EdA students was hosts and the language partners were guests. However, most of the EdA students did not serve food or drinks to their language partners, or try to sit next to their language partners. On the other hand, the

Japanese students found their American partners, and sat next to their American partner, and asked what the EdA students wanted to eat or drink.36 Similar passive behaviors by American

EdA students were observed on the last day of the program, when both the American EdA students and the resident assistants were cleaning the common area. The EdA students stopped cleaning the room much earlier and went to downstairs while the resident assistants were still cleaning the common area. This also suggests that they were acting more as guests or care- receivers that did not have the responsibility to clean the common area thoroughly, even though the two resident assistants, who they might have seen as care-provider, were still cleaning.

The findings above suggest the gap between what both American and Japanese students expected in the relationships (i.e. friends) and how the EdA students positioned themselves (i.e., care-receivers) and the Japanese students (i.e., care provider). For the Japanese students who also

35 The definition of “language partner” and the roles of “language partners” are different depending on students, teachers, programs etc. 36 This might be also due to the cultural differences between what hosts and guests are supposed to do in such a context. 71

saw their EdA students as friends, such a gap resulted in struggles in interacting with the EdA students. However, as Iino points out in his study (2006, p. 163), being a care-receiver requires the role of a care-provider who “perceive themselves as being needed” by the care-receiver. It is possible that those Japanese students might have expected some degrees of assisting the EdA students due to the fact that they volunteered to be language partners whose tasks included helping the EdA students. At the same time, they also expected to become friends with the

American EdA students. The gaps between the expected relationship (peer-to-peer) and the actual relationships (care-receiver and care-provider) may have resulted in the Japanese students’ struggles with the American EdA students’ reactive behaviors.

3.3.2.2. Japanese students’ prior mental images of American people

The second domain also relates to the first domain. The interview data suggested that there was a gap between the Japanese students’ images of American people and the American

EdA students with whom they actually interacted during the program. The Japanese students had images of American people being “furendorii de sekkyoku-teki (friendly and out-going),”

“kekkoo rafu (pretty casual),” “ akarui hito ga ooi (probably a lot of cheerful people),” “jibun no isi ga tsuyo-soo (strong willed: different from Japanese people).”

When the same Japanese students described the American EdA students with whom they interacted during the program, their descriptions were different from their images of American people. Most of them expressed positive surprise in terms of the American students’ language and behaviors, such as comments like “nihon no koto ga suki de, nihon no shuukan ni somatte iru na…(they like Japan and they are accustomed to Japanese customs).” “sugoi ki ga kiku na to

omotta… omotte ita ijoo ni nihongo ga shaberete te sugoi na to motta (I thought (she) was very considerate. Her Japanese was better than I expected, and it was amazing.)” and so forth. On the other hand, as the excerpt below shows, many of them also described the American students’ language and behavior as “too formal,” which made them feel like they couldn't “become friends.”

友達になりたい感じできている。[…]同じ年代の子がくるし、その時に『[本人の名前]さんは』と来られると、一気にまず壁ができちゃうんです。『そうゆう感じなのか、パートナーは。じゃあ、私もパートナーは、敬語で話さないといけないのか』 “Tomodati ni naritai kanji de kite iru […] onaji nendai no ko ga kuru si, sono toki ni ‘[name of the student]-san wa’ to korareru to, ikki ni mazu kabe ga dekichau n desu. ‘sooyuu kanji na no ka, paatonaa wa. Jaa, watashi to paatonaa wa, keigo de hanasanai to ikenai no ka’ (I became a language partner because I wanted be friends with them […] In addition, the American students are around the same age (as me). But when the American students start addressing me as ‘[name of the student]-san’ that immediately creates a barrier. That makes me think, ‘oh, our relationship is like this. So we have to speak in polite language.’

This Japanese student also mentioned that even being addressed by her first name with honorific suffix –san was too much for her. What she expected was an informal and friendly relationship, but her American partner’s polite language use set the tone of a formal relationship although she wanted to be a friend with her American language partner. Similar to this student, the most of the Japanese students indicated that they did not expect the American students to use masu forms.

My personal discussion with the program coordinator at the host university revealed that some of the EdA students were using the humble form, “X to moosimasu (I am (humbly) called

X),” in the first meeting. She mentioned that her image of Americans were “casual and friendly,” therefore, when hearing the American EdA students’ such language use, it was a surprise to the

program coordinator as well. The accounts from the Japanese side indicate that they did not expect the American EdA students to use masu forms, but in the actual interaction, not only did the EdA students use masu forms (i.e., addressee honorifics) but also the humble form in self- introductions, which was unexpected and many felt psychological “distance” because of that.

More findings related to the Japanese students’ perspectives on this topic are discussed in relation to the third research question in the section 3.3.3.

Because the host institution was located in the Kansai region, and many of the Japanese student participants spoke Kansai dialect, this might have also caused the Japanese student participants to feel psychological distance from the American students. This is because dialectal features are more salient in the casual speech style. Therefore, that may have contributed to the

EdA students not comprehending the language partners37 when the Japanese students used the casual speech style. In addition, the EdA students’ inability to use plain forms freely and the fact that EdA students kept using masu forms predominantly might have also resulted in the Japanese students’ perceptions toward the EdA students as too formal, and categorizing them as “cannot become friends with.”

3.3.3. Research Question 2: What kinds of linguistic and cultural difficulties do American

EdA students encounter?

37 American EdA students expressed their difficulty in understanding the Japanese students’ Kansai dialect, and Japanese students expressed their difficulty in not using Kansai dialect when talking with the American EdA students. 74

American students’ struggles also related to the findings we discussed so far. The

American EdA students, indeed, struggled with using the casual speech style in Japanese. The quotes I present below demonstrate the American students’ such struggles.

I was almost kinda afraid of language partners just because my Japanese is really weak. […] The direct-style38 is really difficult for me because I always try to keep it respectful but it’s like, the RA39s have been talking…40 in the direct-style. […] The direct-style41 is really hard for me to do, and I think that probably ummm… ummm, factors into why I have hard time speaking to the language partners, really. […] I just always have the urge to use the distal-style. Because, I… coming to Japan, that’s what I thought, I wanted to do was, I wanted to treat everyone, like with the masu form. […] I wanted to use that with everyone. (I: Why is that?)42 Just coz, I’m a guest in the country, and ummm, I just feel like, … umm, it definitely is weird being here just because, this customer service and everything is so different from America. They all treat (you) with keigo so strong. So, I don’t know, I feel like I just wanna be like that back because I’m NOT43 a native Japanese, and like, I feel like I’m a guest here. Where… what is it… ‘when in Rome, do the Romans do’ sort of. (I: But, then you and your partners are not business partners… it’s more like friends?), I KNOW! That, and….No, I WANT to speak the direct-style, to her (i.e., her language partner), but it’s just hard for me to do the transition, because I am used to doing the distal-style to everyone else. […] it’s just easier for me to say i-masu, comes out so much easier. […] I tried to speak the direct-style.

As the quotation indicates, the student used to think that all Japanese always used masu forms, and this was why she wanted to use masu forms with every Japanese with whom she interacted during the program. However, as she interacted with her language partner and the resident assistants, she realized that they seemed to hold different expectations when interacting with her. The quotation also revealed this student’s prior misunderstanding regarding how

38 Plain forms 39 Resident Assistant 40 This indicates a pause 41 masu forms 42 An utterance by the interviewer. 43 The student puts emphasis on the word 75

Japanese people interact with each other (i.e., always speaking in masu forms). At the same time, it also indicates that she considered herself as a guest, therefore, masu forms were the safe choice.

When I pointed out that her language partner was not a business partner, she exclaimed that she knew it and expressed her strong desired to use the plain forms with her language partner, but it was difficult for her because she had been used to using the masu forms due to previous language socialization, probably at the home institution. This resulted in her fear that she mentioned in the very beginning of the quote of to her incapability of using the expected language style (i.e. the casual speech style).

The quotation below also expresses a similar struggle by another student. This student mentioned that he wanted to use the casual speech style (i.e. predominant use of plain forms) when he was interacting with language partners, but using the casual speech style was more difficult than using distal style (i.e. masu forms), so often times he unconsciously ended his sentences with masu forms.

There are plenty of words that are…just uhh, easier to say, and for a certain that…, it’s gonna sound right in the distal-style […] The direct-style is harder than it should be. I’m getting used to now, because that’s what we have been using most exclusively. […] In class (at the home institution), we mostly [inaudible] in the distal-style, so a lot of words that I... words and phrases and staff almost always in distal-style, so it just comes out that way easiest.

The quotation above also indicates different norms and expectations regarding speech style between L2 learning at the home institution and during the EdA program. A similar reasoning was mentioned in the second quote, but he seemed to understand the importance of using plain forms with his language partners and RAs from the beginning of the program, and wanted to use them, but he has gotten used to using the masu forms.

These two excerpts elucidate two aspects of the American EdA students’ struggles with using the casual speech style. First, as the student in the first quote indicates, there is a misunderstanding that all Japanese people are “respectful” because they always used masu forms in any context. Her quote also indicates that she has constructed this image by using masu forms with everyone at her home institution. Second, as both excerpts suggest, there are gaps in expected speech style between the home institution and the EdA context they encountered.

Both of them along with the other EdA students, however, mentioned that through being exposed to plain form predominant discourse contexts during the EdA programs, they became better at using the plain forms. In addition, through interacting with local Japanese students, the

American EdA student in the first quote realized her assumption of using polite language with everyone to be respectful was wrong.

The findings of EdA students struggling with using the casual speech style are also reported in Burn’s (1996) study. However, in terms of using humble/honorific forms, other studies reported opposite perspectives, namely, that EdA students struggled with using Japanese polite/female language (Iino, 2006; Iwasaki, 2011; Siegal, 1995). These previous studies report

EdA students’ negative perceptions toward using humble/honorific forms. In other words, the participants in these studies did not want to use humble/honorific forms. However, the findings of the current study suggest opposite struggles: students struggle with using the casual-speech style compared to masu forms, including humble and honorific forms.

One possible reason for such differences in findings in the current study and other studies is time differences in collecting data. The current study collected the data when the EdA students had been there for three weeks. On the other hand, other studies collected data when the EdA

students had been in Japan more than three months, or even years. The differences in the amounts of time being socialized into the Japanese society as “foreigners” might have affected the EdA students’ understanding and attitude toward polite language.

Another possibility is that, as the students mentioned, it is probably due to the Japanese language program in their home institution. It emphasizes the importance of appropriate use of the language, both informal and formal including humble and honorific language, but there may have been more learning experiences for the formal speech style than informal settings. In addition to masu forms which have much easier conjugations compared to plain forms, most frequent natural interaction (i.e., not role-play) in Japanese at the home institution might have been conversation with their language instructors, in which a clear hierarchical relationship between instructor and the students. Because of such instructional language socialization at the home institution, the EdA students in the current study did not have a negative attitude toward using the polite language in the beginning of their EdA experiences. However, because of that, they struggled with using the casual speech style in the EdA context.

3.3.4. Research Question 3: What kind of Third Space emerges when American EdA

students and Japanese local students interacting with each other?

As defined in Chapter Two, Third Space in the current study refers to intercultural communication contexts where norms and expectations of such contexts do not entirely conform to the base culture (C1) alone or the target culture (C2) alone. As discussed above, the findings suggest a gap between the Japanese students’ prior images of American people and actual

American EdA students they met. Because of the gap, many of the Japanese students described

the American students as “too formal” and expressed their surprise when the American EdA students used masu forms in the beginning of the program. The Japanese students probably struggled with the American EdA students’ reactive behaviors as well.

However, interestingly, the Japanese students mentioned that they do use masu forms when meeting new Japanese students for the first time. For example, on the first day of school or class, they would start using masu forms and keep using them until they feel comfortable enough to use the casual speech style.

Interviewer 日本人だったら、最初はデス・マス、アメリカ人だったらタメ口の方が接しやすい… Nihon-jin datta ra, saisho wa desu/masu, amerika-jin datta ra tame-guchi no hoo ga sesshi-yasui…(I: You mean, if a person is Japanese, then you use desu/masu in the beginning, but if s/he is American, it’s easier for you to interact with him/her in the casual speech style) Japanese student 特に学生同士だと、そうやと思います。 Toku ni gakusee doosi da to, soo ya to omoimasu. (I think so, especially if s/he is a student).

This student told me that there are different expectations when she interacts with

American students, especially when both sides are students. The next account from a different student also indicates that she held different expectations toward American students because she recognizes a different cultural background.

初対面は、デス・マス…よりも日本人同士ならデス・マスでっていう形があると思うんですけど、やっぱり、違うっていう認識はあると思うんで、そしたら、パーって言っちゃった方が、受け入れやすいかもしれない。パートナーとかだったら、バーって来た方がやりやすい。 Shotaimen wa, desu/masu yori mo… nihon-jin dooshi nara desu/masu de tte iu katati ga aru to omou n desu kedo, yappari chigau tte iu ninsiki wa aru to omou n de, sositara, pa tte icchatta hoo ga ukeire-yasui kamosirenai… paatonaa toka datta ra, baa tte kita hoo ga yariyasui. (When meeting for the first time, rather than desu/masu…., if they are both Japanese, I think ‘desu/masu’ form is the norm, but I think there is a recognition of (cultural) differences, so it is probably 79

easier to accept Americans right off the bat (indicating casualty rather than formality), so if it is a language partner, then it’s even easier.)

The local Japanese students mentioned that the use of masu forms in peer-to-peer self- introduction was the norm in Japanese culture. However, their accounts also indicate that their expectations toward the American EdA students were different from the expectations or norms when they communicate with people from the same cultural background.

On the other hand, use of humble and honorific forms, such as “X to moosimasu (I am humbly called X)” is not the norm in Japanese culture in the context where two Japanese college students meet for the first time. Use of humble and honorific forms rarely occurs in conversation among college students, and it is considered as too formal or overly polite. The American EdA students’ use of humble or honorific language in a self-introduction is clearly not C1 (i.e.

American culture) negative transfer because American college students do not usually use humble or polite expressions in English when introducing themselves to other students at an

American university. Then, why did the American EdA students use such humble and polite expressions? As mentioned in the previous section, it may have been due to their misunderstanding of Japanese communication and their assumption that all Japanese people communicate in that way. Some of them might also think that using humble/honorific language makes them sound more advanced speakers. The findings show that American EdA students and local Japanese students were interacting in Third Space, characteristics of which do not entirely conform to the American culture and Japanese culture.

These findings also suggest that such gaps between American EdA students’ behaviors and language use and the Japanese students’ expectations became obstacles to build personal

relationships. This type of Third Space is not productive to construct a good personal relationship, where both parties work together to achieve shared interests. To address this issue elucidated in the current study, the next section discusses pre-departure training for both

American and Japanese students.

3.4. Pedagogical Implications

Although this phase of the research was originally aimed at providing insights on pedagogical support for U.S. college students who participate in EdA programs in Japan, the findings suggested that such support is also needed for local Japanese students to co-construct a productive TS. Therefore, based on the findings of the current research, this section presents pedagogical implications for American EdA students, local Japanese students, and EdA programs.

3.4.1. Pre-departure training for American EdA students

Pre-departure training for American EdA students should address the American EdA students’ struggle with presenting a “relationship-building persona” that is appropriate when meeting Japanese college students for the first time. Because EdA students often encounter situations where they have to introduce themselves to others at the beginning of EdA in Japan, it is important for them to be able to exhibit a persona through self-introduction and following conversation that makes their interlocutor feel comfortable and want to develop a personal relationship with them. The performance of self-introduction is especially important for a short- term EdA program where there is a limited amount of time to fill “gaps” and “psychological

distance” compared to long-term EdA programs, where EdA students and local Japanese students may have more time to fill such gaps.

A suggested activity in pre-departure training is performance of self-introduction, which includes the tasks of 1) introducing oneself; 2) finding common hobbies with the interlocutor; and 3) initiating the next meeting with him/her based on the conversation. If students have already learned self-introduction in a Japanese language course at the home institution, the instructor first gives them a clear context for the performance and has them perform without any instructions.

After the performance, the instructor and other students give explicit feedback on the students’ speech style including non-use of honorific and humble forms, use of first-person pronouns and behavioral aspects such as facial expressions, tone, as well as gesture. When giving feedback on speech style, it is also important to draw the students’ attention to style shifting. As discussed in Chapter Two, research reports that style shifting is a salient and important aspect of Japanese conversation (see Section 2.2 in Chapter Two for details). By utilizing style shifting, American students can avoid sounding too formal while at the same time, also express their initial respect to the interlocutor with whom they are interacting for the first time. As EdA students continuously interact with locals, their relationships change gradually rather than suddenly. Being able to style shift to express their psychological stance moment by moment is a crucial communication strategy in Japanese. In addition, by carefully observing how their interlocutor uses their speech style, they understand what kinds of expectations that the interlocutor holds in terms of their relationship. Practicing appropriate style shifting based on how they want to present themselves to their interlocutors enables students not only to be active

users of style shifting but also to take initiative for building personal relationships with local

Japanese people.

As mentioned above, the current study reveals that it was easier for the EdA students to use masu forms than plain forms for two reasons. First, conjugation of verbal masu forms is simpler than plain forms, and because of that, masu forms are introduced before plain forms in most of Japanese language programs in the US. Second, formal instruction often focuses on use of polite language including masu forms, because it is considered “safe” to use the masu forms and be polite (or too polite) instead of using plain forms, which can be considered as “rude” in certain contexts. In addition, in the foreign language-learning environment, the natural interaction with L1 speakers (or near L1 speakers) that students have most is interaction with their teachers. In such interactions, there is a clear hieratical relationship between the teacher and the student, and the students are expected to use masu forms predominantly.

In Japanese society, there is a gap between the language used by gakusee (students) and that used by shakai-jin (a full-fledged member of the society). When new graduates (shinsotsu) start their career at a company, they often have to go through initial training (shinnyuu-shain kenshuu) where they will learn the appropriate use of language in the business setting. Because the primary purpose of higher education is for students to acquire knowledge and skills to be successful in their career, formal language use is more emphasized than informal language in

Japanese language programs in the US. However, because of the difference in language used by gakusee and shakai-jin, the American students in the current study struggled with developing a relationship-building persona in the EdA context.

What L2 students need to understand is that an appropriate speech style changes depending on the situation, and overuse of masu forms in a certain context can be an obstacle in building interpersonal relationships rather than being “safe.” As discussed in Chapter Two, the indexicality of masu and plain forms heavily depend on the context. In order to make the indexicality of the forms learnable, establishing the five components (i.e., time, place, role, script, and audience) of performance is essential when practicing self-introduction, including style shifting within the self-introduction. By making these five components clear, Japanese L2 speakers can use the skills for style shifting as a powerful communication strategy to recognize, negotiate, and co-construct expectations with locals. The ability to style shift can be used to conduct communication successfully and to project a “relationship-building persona” toward the interlocutor, which consequently facilitates building close intercultural relationships.

Explicit feedback on using personal pronouns is also important especially for male learners. Male learners, who have socialized themselves in a formal instruction setting where instruction focuses on language use in a formal setting such as business meetings or academic conferences, are often accustomed to using “watashi” as their default first person pronoun.

However, the same first person pronoun used by a male speaker might index different social meanings in different contexts, such as sounding too polite or too soft. Although students are often aware of such social meanings of “watashi” and know that other first person pronouns, such as “boku,”44 are more appropriate in a casual setting, without actually practicing it, they cannot develop a procedural knowledge of appropriate first-person pronoun use in casual

44 Another male personal pronoun, “ore,” can be introduced in this activity, but as previous studies revealed that L2 speakers often face difficulties when using such male language (Iwasaki, 2011, p. 97). 84

communication. Furthermore, as the findings of the current study reveal, Japanese local students often struggled with the American EdA students’ reactive behaviors toward interaction.

Practicing how to initiate a meeting up in a friendly and culturally appropriate manner can enable

EdA students to be more active and engaged in interactional experiences with local Japanese people.

Presentation of a “relationship-building persona” is not only constructed based on scripts but in behavioral aspects as well, which was pointed out by one of the Japanese students below.

敬語を使うとしても改まった感じじゃなくて、もう少し軽い感じで「あるじゃないですか」, (produced with a low voice)’じゃなくて「あるじゃないですか」 (produced with a higher voice, indicating cheerful attitude),’ 違うじゃないですか。やっぱ使うとしても初対面でもビジネスじゃないじゃないですか。だからそうゆう友達になるくらいの勢いで…45 Keigo o tsukau to shite mo aratamatta kanji ja nakute, moo sukoshi karui kanji de ‘aru ja nai desu ka (produced with a low voice)’ ja nakute ‘aru ja nai desu ka (produced with a higher voice, indicating cheerful attitude),’ tigau ja nai desu ka. Yappa tsukau to shite mo, shotaimen de mo bijinesu ja nai ja nai desu ka. Da kara soo yuu, tomodachi ni naru kurai no ikioi de…(Even if you use the polite language, it’s not formal like ‘isn’t it? (With a low voice),’ but with a more light- hearted feeling, but like ‘isn’t it? (With a higher voice, indicating cheerful attitude),’ they are different, right? After all, even when using it, even when meeting for the first time, it is different from business settings. So in situations like that, (American EdA students) should approach it with the sprit of making friends.)

This student expressed her view of the EdA students with whom she interacted during the program, and commented that the behavioral aspects (such as gesture and voice tone) of the EdA students were too formal, which resulted in creating a “(psychological) wall.” The crucial point here is, even using the masu form, friendliness could be expressed. It is important for American

EdA students to learn not only linguistic codes or scripts in Japanese conversation, but also

45 This transcription is based on the actual speech data, and there is not typographical error in the excerpt. 85

behavioral aspects of a relationship-building persona, such as tone, facial expression, and gesture when meeting a Japanese student for the first time, as well as daily communication after the first day of meeting. This type of communication skill cannot be practiced without actually performing the task. In order to practice such skills, performance of self-introduction and style shifting by focusing on behavioral culture including language, gesture, facial expressions and other important communication aspects through actual practice is a crucial for successful L2 learning in an EdA context (Walker & Noda, 2000).

Another suggested activity includes observation and discussion of how Japanese young people communicate for the first time meeting. For such activities, use of non-pedagogical videos taken from Japanese TV dramas or movies is recommended. The potential discussion topics would be 1) what type of the speech style are used in a given the context; 2) what kinds of changes in the context lead to changes in the speech style; and 3) what kinds of changes in the speech style as well as expectations occurs when foreign students are interacting with Japanese students? After discussion, an activity of creating a script of meeting a Japanese student for the first time would be effective for the students to apply what they have observed and discussed, and practice effective communication strategy with Japanese college students.

Because EdA students often have to figure out and resolve the communication problems by themselves, pre-departure training must prepare the EdA students for such an environment.

Thus, instead of providing a script and practicing it, having them going through the process of creating their own culturally and linguistically appropriate script will cultivate self-monitoring skills in actual communication.

3.4.2. Training for Japanese students

As mentioned in Chapter Two as well as in this chapter, it is important that the both sides understand the other side’s of culture and cooperate in order to co-construct a productive Third

Space in intercultural communication and relationship building. The findings of the current study suggest that one of the causes of the Japanese students’ struggles is the gap between their prior mental images of American people and the actual American EdA students they met. Because they recognized cultural differences between Japan and the US, consciously or unconsciously, they did not expect the American EdA students to follow Japanese norms (i.e. use of desu/masu forms in the first meeting). Therefore, training for local Japanese students should focus on adjusting their prior images of American people if there are gaps between their prior images and

American EdA students. This can be done by holding discussion with former language partners at the host institution and sharing their prior experiences and struggles with EdA students. Such insights would help the Japanese language students to prepare to adjust their images of American

EdA students easily if necessary during the program, instead of feeling a “psychological distance.”

It is also practical to discuss the difficulty of using plain forms for EdA students in terms of the conjugation as well as the EdA students’ prior language socialization at home. This could be done by pointing out some English verbs that have irregular forms for conjugation, or the difficulty of using colloquial English. By doing so, Japanese students can also see American students’ struggles as very similar to their own struggles in learning English, which may help them to be more open to the American students’ struggle to use the casual speech style in

Japanese.

Because most Japanese students in the current study participated in the EdA program as language partners to gain intercultural experiences, such training can facilitate their intercultural learning and development of personal relationships with EdA students and provide a sense of accomplishment toward their objectives. Such learning experiences can also help the Japanese students feel less like “care-providers” to EdA students and overcome the struggles that the

Japanese participants in the current study had.

3.4.3. Pedagogical implications for the design of EdA programs

In order for American EdA students not to be perceived as “reactive” or to position themselves as care-receivers all that time, an extra curriculum for the EdA program should be designed in such a way that EdA students take initiative in their own learning. For example, instead of having “free time” after field trips and expecting the EdA students to naturally have productive time with their language partners and/or resident assistants, which often cases did not happen based on the current study’s findings, it is more effective to call such time “community exploration time” and instruct them to find places they want to visit and figure out how to communicate effectively with their language partners prior to the program. Such program interventions are also reported as effective in the Dewey et al. (2014) study on L2 language use.

The authors report that program design, rather than such factors as the social environment, personality traits of participants, or the language proficiency background of the students, is a significant predictor of EdA students’ L2 language use. Such program intervention is necessary to maximize students’ EdA learning.

3.5. Conclusion of Chapter Three (The First Phase)

The primary purpose of the first phase of the study was to develop pre-departure training based on empirical research. In order to do so, it was essential to investigate what was difficult in building intercultural relationships by obtaining perspectives of both American EdA students and

Japanese local students in a four-week EdA program. The findings suggest that there were gaps between the expectations about speech style on the part of the American EdA students and of the local Japanese students. In addition, these expectations were different from the norms of C1 or

C2, indicating the emergence of the Third Space within which both sides were operating. The study also revealed American EdA students’ struggles in developing a relationship-building persona, and because of these struggles, the Japanese students perceived the American EdA students as being “too formal” and “difficult to become friends with.” The findings indicate the need for pre-departure training that focuses on speech style shifting as a strategy for American

EdA students to build an appropriate relationship-building persona in order to recognize, negotiate, and co-construct expectations with locals and manage a productive Third Space.

Based on the findings discussed in this chapter, the following chapters focus on the second phase of the study, which primarily aims to assess the effectiveness of the training. The next chapter, Chapter Four, presents and discusses the methodology of the second phase of the study.

Chapter 4: Methodology for the Second Phase of the Research

4.1. Introduction

Review of research literature in Chapter Two reveals that the assessments of education abroad (EdA) learning must reflect students’ EdA learning. In addition, it also reveals that investigation of the process of EdA learning is crucial to understand what is actually going on during EdA experience in Japan. The current chapter presents the methodology for the phase two of the research, which aims to investigate the outcomes and process of U.S. students’ EdA development and the effectiveness of the EdA training designed based on the findings of the first phase of the research discussed in Chapter Three.

4.2. Setting: Home institution and host institutions

The current study involved undergraduate students from a large public research university in the Midwest US. The Japanese language program at the university offers various types of modern Japanese language courses including intensive and individualized courses. It consists of seventeen courses at the undergraduate level and six courses at the graduate level.

The university offers two types of EdA programs for Japan: faculty-led short summer

EdA programs, and semester- or yearlong46 exchange programs. The Japanese program requires that students complete at least level three Japanese language courses or have an equivalent level of language and cultural knowledge and skills before participating in any of the EdA programs for Japan, with the exception of one yearlong exchange program, which only requires Level 2 or equivalent. This is because students who complete level three Japanese language courses are expected to have mastered both language and cultural skills to navigate their own learning in

Japan. Every year, ten to twelve students participate in EdA exchange programs at affiliated

Japanese universities for a semester or a full academic year.

Applying for a yearlong EdA exchange program in Japan at the university is a long process starting from around November of the year previous to the intended EdA participation to the summer preceding the start of the EdA in autumn. Depending on the partner school, students receive the paperwork necessary for their visa application as early as May or as late as August.

The student application for the exchange program in Japan consists of a statement of purpose, a recommendation letter, and a current transcript. After the initial review of the applications, the students then have individual interviews with the program director of the Japanese program and the EdA program director from the Office of International Affairs (OIA). A number of factors are discussed during the in-depth interview, including but not limited to applicants’ language proficiency, goals, location preferences, future aspirations, financial needs, and readiness for managing their lives independently in a new culture. With a recommendation from the home institution, the students then submit their applications to their recommended host schools. Many

46 Yearlong refers to an academic year (i.e., approximately nine months). 91

of them also apply for scholarships and financial aid. Because all EdA students have to go through the rigorous application process, we consider that they are all highly motivated EdA participants.

Seven students attending five of the partner universities in Japan agreed to participate in this study.47 Table 3 presents the information about these five partner universities. As we will discuss in Chapters Five and Six, overall findings suggest that the design of the EdA programs play a crucial role in EdA gain. Furthermore, the analyses of findings suggest that the EdA program at University CFI, where three students went, provided the best environment in terms of developing strong social ties that benefited in their EdA gain. The detailed information on how the design of the EdA program influenced the participants’ EdA gain will be discussed in

Chapter Six. This section provides general information of the five universities in the current study.

Table 3. Information of five partner universities in Japan Host Type of University Student Population universities Undergraduate Graduate International students students students University CFI Liberal arts college 2,780 230 300 University A National university 11,100 6340 1850 University B Private research university 39,470 5,950 4,800 University E Private university 9,800 210 250 University H National university 10,850 5,550 1,950

47 The alphabet letters of the host universities represent the first letter of the pseudonym assigned to each participant who attended the university. Originally eight students agreed to participate in this phase of the study, and University B in Table 3 was to be named University BD. However, because one student (pseudonym Duke) did not meet all the requirements for the research, this university was named University B. Pseudonyms of the participants are presented in Chapters Five and Six. 92

Three students went to University CFI, a four-year liberal arts college, located in the

Kanto region. The university hosts approximately 2,780 undergraduate, 230 graduate, and 300 international students. According to the host institution’s website, the university has one of the most well-known programs in the world for teaching Japanese to foreign speakers.

One student went to University A. It is one of the major national universities and is located in one of the five largest cities in the Northern part of Japan. The university website indicates that approximately 11,100 undergraduate and 6,340 graduate students are enrolled as well as 1,850 international students coming from 67 countries. University A offers for its international students an English-only track program that includes Japanese language, history, and culture. In addition, international students may take courses offered in Japanese for the non- international students, if they pass the schools’ language placement test.

One student went to University B, a private research university and the largest university among the universities in the current study in terms of the size of the student population with approximately 39,470 undergraduate, 5,950 graduate, and 4,800 international students. The

University is also located in the Kanto region, which includes Tokyo and several other prefectures. International students in the exchange program take language courses according to their levels, as well as such courses as Japanese culture, Asian culture, society, Japanese politics and economics, all taught in English on. Students do not have ready access to regular courses taught in Japanese.

One student went to University E, a private university located in a city in the Chubu region, in the central area of Honshu. There were approximately 9,800 undergraduate, 210 graduate, and 250 international students at the university at the time of the study. In addition to

multi-level Japanese language classes, participants in the EdA program can also attend courses in

Japanese and Asian culture, politics, economy, society, and many other disciplines taught in

English.

One student went to University H, a national university that is considered to be one of the top universities in Japan. It is located in one of the largest cities in the Tohoku region, north of

Tokyo, and is home to approximately 10,850 undergraduate, 5,550 graduate, and 1,950 international students. In addition to the Japanese language courses, international students can take courses in Japanese and Asian culture, politics, economy, society, religion, literature, and history, which are taught in English. General education courses for local students are also available for international students if their language proficiency is high enough.

4.3. Participants

The data from seven EdA students constitute the core data for this research. Ten undergraduate students at the home university participated in EdA programs in Japan during the

2017-2018 academic year. I contacted all of the ten undergraduate students by email for the recruitment. Eight out of the ten students originally expressed their interest to participate in the study. After detailed explanations of the procedure of the study, all eight students agreed to participate in the current study. However, as mentioned above, some data from one student was not available, therefore, I excluded his data from the analyses of the current research.

The seven students chose whether they participated in EdA training group or in non-EdA training group. Four of them opted to participate in the EdA training group. The EdA students in this group were given access to EdA training and all four attended it. The content of the

training was designed based on findings of the previous research discussed in Chapter Three but not all activities were conducted in the actual training due to time constraints. They also received periodic online meetings with the researcher throughout their time in Japan. The contents of the EdA training will be discussed further in detail in Section 4.5. in this chapter.

The other three students agreed to participate in the study but opted not to participate in the EdA training (i.e., non-EdA training group). Chapter Five and Six provide detailed information about the participants in the current study along with the related findings.

4.4. Instruments

In this section, we first discuss the instruments used to collect data on the outcomes of

EdA learning, which we defined as EdA gain in Chapter One. Second, we discuss instruments to collect data on the process of EdA learning by investigating social network development during

EdA programs in Japan.

4.4.1. Instrument of EdA gain

In order to measure participants’ EdA gain quantitatively, two language proficiency tests were employed. One of the proficiency tests was the American Council on the Teaching of

Foreign Languages’ (ACTFL) Oral Proficiency Interview (OPI) test. Because OPI test is a proficiency test to measure primarily test takers’ oral skills, the current study also employed another language proficiency test, Japanese Computerized Adapted Test (J-CAT) to measure other skills, such as listening and writing skills. The following sections describe the two proficiency tests in detail. The students’ utterances during the OPI were also examined to gauge

students’ style-shifting development. This examination was intended to shed light on skills on cultural aspects that may not be demonstrated in OPI ratings, though the OPI rater may consider such skills as one of the multiple indicators of proficiency.

4.4.1.1. Informal Oral Proficiency Interview (OPI) test

Oral Proficiency Interview (OPI) test is a language proficiency test to measure speaking skills and is a one-to-one interview in a target language (L2). The interviewer elicits various speaking tasks, such as answering a word-level question (e.g., “what is your hobby?”)48 to complicated tasks of stating, defending opinions on social issues, as well as discussion of hypothetical situations. As mentioned in Chapter Two, the ACTFL OPI rating scale consists of five major levels, which are Novice, Intermediate, Advanced, Superior, and Distinguished levels.

Because it is extremely rare for undergraduate students whose native language (L1) is English to be rated at the Superior or Distinguished level in the Japanese OPI test, this section focuses on the discussion on the Novice to Advanced levels. Table 4 presents the descriptions of outcomes that are used as criteria of the OPI rating.

Communication at the Novice-level is limited to everyday topics that are “highly predictable (ACTFL, 2012, p. 14).” A Novice-level speaker’s production often consists of memorized vocabulary and phrases, and they often have difficulties applying these memorized words or phrases to create a sentence in a new situation. Hence, their utterance often consists of a word or a phrase.

48 Although a sentence-level response, “my hobby is running” is also possible as a response to the question, a test taker also can answer this question with one word, “running” to engage in a natural conversation. 96

By contrast, the major difference between Novice and Intermediate-level speakers is that an Intermediate-level speaker is able to create with language when dealing with topics that are related to his or her daily life. In other words, they can talk about familiar topics by combining and recombining previously learned words and phrases, and they are able to produce utterances at a sentence-level by applying what they know in limited contexts. Intermediate-level speakers are often described as survivors because learners at this level can survive with their language skills when they face contexts that are familiar to them in the target culture (C2).

Table 4. Criteria of OPI ratings (ACTFL 2012, 6) Proficiency Global Tasks and Context/Content Accuracy Text Type Level* Functions Advanced Narrate and Most informal and Understood Paragraphs describe in major some formal without difficultly time frames and settings. Topics of by speakers deal effectively personal and unaccustomed to with an general interest. dealing with non- unanticipated native speakers. complication. Intermediate Create with Some informal Understood, with Discrete language, initiate, settings and a some repetition, by sentences maintain, and bring limited number of speakers to a close simple transactional accustomed to conversations by situations. dealing with non- asking and Predictable, native speakers. responding to familiar topics simple questions. related to daily activities. Novice Communicate Most common May be difficult to Individual minimally with informal settings. understand, even words and formulaic and rote Most common for speakers phrases utterances, lists, aspects of daily accustomed to and phrases. life. dealing with non- native speakers.

An Advanced-level speaker is able to communicate beyond the survival level, meaning they are able to talk about topics related to their own community or current issues. They do so with paragraph-length utterances. Because a speaker at this level can narrate and describe topics that are personal as well as general, such as community-related topics, they can be described as narrators. A jump in rating from Intermediate to Advanced levels, in general, is considered much greater than that from Novice to Intermediate-level because a speaker has to master much more linguistic skills to manage communication at the Advanced level.

The major levels are divided into three sub-levels, which are Low, Mid, and High depending on how well a test taker sustains the level of performance. For example, as mentioned above, in order to be rated as an Intermediate-level speaker, a test taker has to create with language for all intermediate-level tasks. A test taker who sustains sentence-level utterances with creating the language for all intermediate tasks shows a few difficulties such as frequent hesitancy and inaccuracy, but they do not affect negatively to the interlocutor’s interpretation,49 the test taker is likely to be rated as Intermediate-Low. A speaker who is comfortable communicating in all Intermediate-level tasks but shows breakdowns in performing most of the

Advanced-level tasks, such as they cannot respond with paragraph-length utterances or cannot describe a current issue, he or she is likely rated as Intermediate-Mid. If a speaker can perform all the Intermediate-level tasks with confidence and most Advanced-level tasks but shows breakdowns in one or two Advanced-level tasks, the speaker is likely to be rated as an

Intermediate-High speaker. On the other hand, if a test taker shows breakdowns in his or her

49 Intermediate speakers are understood by people who are “accustomed to dealing with non- native speakers” (ACTFL, 2012, p6). 98

performance in an intermediate-level task, meaning they cannot sustain a sentence-level response by creating with language, the test taker is likely rated as a Novice level speaker. If the test taker is able to sustain most of the intermediate-level tasks, but not all of them, then the speaker is likely rated as Novice-High.

I completed a four-day OPI Workshop hosted by ACTFL in June 2015 and August 2017.

For this data collection, I administrated the OPI tests with the participants. Each participant had

OPI interviews twice, before and after their EdA programs in Japan. The OPI ratings refer to a range of proficiency levels. In other words, when a student receives the same rating before and after EdA doesn’t necessarily mean there was no gain in proficiency. It means that the level of gain is too small to change the OPI rating.

Because I was the only tester who conducted and rated the participants’ OPI interviews of the current study, and the ratings in the current study were not official ratings given by ACTFL, the

OPI ratings are considered as informal.

4.3.1.2. Japanese-Computerized Adaptive Test

Japanese Computerized Adaptive Test (J-CAT) is a computer-operated Japanese proficiency test developed by researchers at the University of Tsukuba in Japan (see http://www.j-cat.org/ for the official website of the test). This test was employed to assess participants’ EdA gain in vocabulary, grammar, listening, and reading skills. The on-line test was highly accessible in that it could be taken at any time and, at the time of data collection for this research, was free of charge. The ease of accessibility was a crucial advantage Japanese

Language Proficiency Test (JLPT), a widely used Japanese language proficiency test both in

Japan and in the US (see https://www.jlpt.jp/e/). JLPT is offered only twice a year in Japan and once a year in the US and there are four distinctive tests for different levels that test-takers need to select to take, N1 through N4, were passing the N1 level is the highest level measured by

JLPT. In contrast, J-CAT provides a discrete score without having to take multiple levels of test.

In addition, all EdA applicants to Japan were required to take the J-CAT. Because a J-CAT score sheet includes an interpretation of a test taker’s score with simple descriptions of what they are expected to be able to do with their language skills in addition to a comparison with the JLPT score (see http://www.j-cat.org/html/ja/pages/interpret.html), we can refer to the sheet for the interpretation of J-CAT score regarding what the test taker is likely able to do or not do.50

As the name indicates, a learner takes J-CAT on the computer. The format is multiple- choice, and it consists of four sections including listening, vocabulary, grammar, and reading.

The number and difficulty of the questions vary depending on how well the students answer (see http://www.j-cat.org/html/ja/pages/interpret.html). How long the test takes also differs depending on the proficiency level of test-takers, and it ranged from forty-five to ninety minutes. The maximum score for each section is 100, and the highest score a test taker can receive is 400.

Table 5 describes the interpretation of a J-CAT score and the corresponding JLPT test levels.

Table 5. Interpretation of J-CAT scores and comparison with JLPT ratings J-CAT Proficiency Level JLPT -100 Basic Level N4 100-150 Pre-Intermediate Level N3 150-200 Intermediate 200-250 Intermediate-High Level N2 250-300 Pre-Advanced Level N1 300-350 Advanced 350- Near Native

50 It also provides estimates of JLPT levels based on the J-CAT score. 100

According to the J-CAT official website (2012), Basic learners can exchange basic ideas, such as understanding simple self-introductions by teachers and friends in a classroom. Intermediate learners can manage daily communication, and generally, understand conversations on familiar everyday topics. Advanced learners can manage academic and professional communication, such as understanding the general contents of TV programs covering familiar everyday topics and contents regarding their work or specialization (see http://www.j- cat.org/html/ja/images/about/cando.pdf for more detailed descriptions). Compared to the detailed descriptions of each OPI proficiency levels in by ACTFL, these interpretations in J-CAT are somewhat brief and ambiguous. Nonetheless, the discrete scores that this test offers seemed to complement well the OPI ratings that covered a range of proficiency levels in assessing the proficiency gain of EdA participants.

4.3.1.3. Examination of style shifting

In addition to the informal OPI and J-CAT discussed above, EdA gain was measured by analyzing participants’ development on style shifting in Japanese. Pre- and post-EdA OPIs were used as the data of the style-shifting analysis. After the OPIs, all the interviews were audio- recorded and transcribed for data analysis (see detailed description for data analysis of style shifting in Section 4.7.1 below).

4.4.2. Instruments for measuring social networks

The review of research literature on EdA learning in Chapter Two also revealed a strong relationship between EdA gain and social network construction during EdA experience (Baker-

101

Smemoe et al., 2014; Isabelli-Garcia, 2006). Thus, the current study examines participants’ social network development during EdA programs in Japan to gain a better understanding of the process of EdA learning.

For social network analysis, the current study chose three instruments, namely, questionnaire at the end of the EdA program, interviews with participants shortly before their departure and again after their return to their home institution, and participants’ logs on the social network throughout the duration of the EdA programs in Japan. Collecting multifaceted data improves the validity of the data (Milory, 1987). Observation of an informant’s interaction with people within his or her social network and interview with people in the informant’s social network (Milroy, 1987) are also common ways to collect social network data. However, these instruments were not employed in the current study because it involved participants who attended five different host universities in different parts of Japan, and it made observation and interview with all people in the participants’ social networks extremely difficult.

4.4.2.1. Study Abroad Social Interaction Questionnaire (SASIQ)

Study Abroad Social Interaction Questionnaire (SASIQ), developed by Dewey et al

(2012a, b.) was used to ascertain EdA students’ social networks (see Appendix E). The questionnaire consisted of fifteen questions. The first five questions asked questions regarding people in the participants’ social network, and how often the participants interacted with them during EdA. It also asked how they met and how close the participant felt with the person.

The second part consisted of six open-ended questions asking for the participants’ thoughts and opinions on successful and unsuccessful experiences in building personal

102

relationships. These were based on their own experiences during EdA programs in Japan. The third part of the questionnaire asked the participants to categorize the people in their social networks into certain groups, such as groups of host family, club, and volunteer groups, etc.

Analysis of this question enabled us to observe the tendencies of what social networks during undergraduate EdA programs in Japan consisted of and how they were organized. The eight students completed this questionnaire one to two months after their EdA experiences.

4.4.2.2. Pre- and post-EdA interviews in English

I also interviewed the participants regarding their experiences in interacting with people and constructing social networks during EdA programs before and after their EdA experiences.

The interviews consisted of semi-structured questions and additional questions depending on how participants responded to each question. I asked questions related to SASIQ to gain a better understanding of their social networks. In order to elicit detailed accounts the participants’ first language, English was used.

The pre- EdA interview was conducted primarily in order to understand the participants’ background and to assess their needs for pre-EdA training. The interview consisted of nine semi-structured questions, including questions regarding their initial motivation for studying

Japanese, their personal history of Japanese study, EdA destinations, type of accommodation in

Japan, goals, expectations of EdA experience and so forth (see Appendix C for the pre-EdA interview questions). The interviews lasted from approximately twenty to eighty minutes.

By contrast, the post-EdA interviews aimed to gain more detailed accounts of the participants’ social networks during their EdA experiences as well as their learning experiences.

103

The interviews consisted of twelve semi-structured questions (see Appendix D for the post-EdA interview questions) regarding their overall impressions of EdA experience, struggles in interacting with locals, strategists used to overcome these struggles and their opinions on how pre-departure EdA training can support U.S. students who participate in EdA programs in Japan.

The length of each interview varied, ranging from approximately forty to ninety minutes.

4.4.2.3. Education Abroad Language Use Log

In addition to pre- and post-EdA interviews, Education Abroad Language Use Log (EdA-

LUL) was employed to record students’ perspectives on their language use. This type of data was used as supplementary data to confirm participants’ responses in the questionnaire and the interviews. Regarding instruments for collecting such data, Language Contact Profile (LCP)51 developed by Freed, Dewey, Segalowithz, and Halter (2004) was widely used among the research that aimed to obtain data on EdA students’ both L1 and L2 use. It consists of the pre- and posttest, and the pretest is used to collect data on students’ background information on L2 learning, whereas the posttest version of the LCP is used to collect the students’ actual use of L1 and L2 administrated after their EdA programs. However, collecting this type of data after EdA programs based on the participants’ memory on L1 and L2 use may result in low reliability. In addition, reporting usage of each activity separately (i.e., activities related to speaking, listening, reading, and writing) might not reflect actual L2 use because oftentimes, more than one such skills are used at the same time (i.e. listening and speaking skills are used when in an activity of

51 The LCP consists of two parts. The first part pertains to collecting data on participants’ background information as well as L2 learning history and usage prior to the EdA experience, and the second part is to collect data on the participants’ L2 use during the EdA experience. 104

talking with a friend). We should not assume that all speaking activities are interactive and reading and writing activities are non-interactive. Some speaking activities are non-interactive

(e.g., practicing dialogue in L2 or singing an L2 song and, some reading and writing activities are interactive (e.g., receiving and sending emails, or using online-chat with a friend).

Considering these elements of language use during EdA experiences, the current study created a new instrument rather than employing the existing instrument for language use during

EdA programs. Participants’ language use was collected via a phone application, which was developed with online software called AppSheet (https://www.appsheet.com/). In this way, participants could download the app on their smartphone or tablet that they carried around with them and log their language use as much as possible instead of collecting such data at the end of the program.

Information collected with the app includes types of language (such as L1, L2 or others), time (when and how long), types of language use (i.e., interactive or non-interactive), with whom, descriptions of the language use activity, and their thoughts and feelings during the activity.

Regarding the type of language, participants could choose English, Japanese, or Other (i.e., a language other than Japanese or English) when only one language was used for the activity.

When more than one language were used in one language activity, they could choose “J = E”

(Japanese and English were used at the same time around the equal amount), “J > E” (Japanese was used more than English), or “E > J” (English was used more than Japanese).

105

Regarding types of language use, participants first decide if it was interactive or non- interactive language use52 and also chose what types of language skills (i.e. speaking, listening, reading, and writing) were used in the activity (multiple answers were possible). When the language use was interactive, participants were also asked to input information about their communication partners. Such information includes their relationships with the conversation partners (e.g., friend, professor, and members of a club) and their names.

The participants were asked to log their language use, both in English and Japanese for a week during the time periods of October and December 2017, and February and April 2018.

Data was automatically recorded on excel sheets.

4.5. EdA Training

One of the purposes of the current study was to investigate the effectiveness of the data- driven EdA training. Four of the seven study participants agreed to receive EdA training. The

EdA training was developed based on the data collected in the phase one research and focused on developing style-shifting skills. The EdA training consisted of two parts: pre-departure training and regular online meetings during EdA in Japan. The following sections explain the contents of each part.

52 Participants received Instructions (see Appendix F) on how to use the language app before logging their language use. The distinctions of interactive and non-interactive language use were also discussed in the instructions. 106

4.5.1. Pre-departure training

The training was originally intended to be administrated in person at the campus of the participants’ home institution. However, because the training was conducted at the beginning of

August 2017, all EdA participants stayed in their hometowns, and it was difficult for them to commute to the campus. Therefore, the training was conducted through Skype and done individually.

The findings of the first phase of the research discussed in Chapter Three suggest that

American undergraduate students struggled with developing a relationship-building persona that was appropriate in the EdA contexts primarily due to the different speech styles they were socialized in at home where they were accustomed to using formal speech style (i.e., a masu- form-predominant discourse context) and in the EdA context where they were expected to use casual speech style (i.e., a plain-form-predominant discourse context) most of the time.

Therefore, the training was designed to support the development of a relationship-building persona appropriate in undergraduate EdA programs in Japan. Performed Culture Approach

(PCA) was used for the framework of the training. The pre-departure training consisted of a role- play and discussion of the speech style in the role-play. The setting was at a host university, and it was their first time to visit a school club they were interested in participating. They saw a

Japanese student when they entered the clubroom, and their primary task was to communicate with the Japanese student. The researcher played the role of the Japanese student who they wished to befriend. The detailed tasks included 1) introducing oneself; 2) finding common hobbies with the interlocutor, and; 3) the initiating next meeting up with him/her based on the conversation. The prompt was shown in both Japanese and English (see Appendix G).

107

After the role-play, the participant and the researcher discussed their performance focusing on their speech style. Discussion questions included whether or not they were aware of their speech style while performing the tasks, their reasons for choosing a certain speech style including the use of first-person pronouns. Behavioral aspects, such as tone of voice were also focused on, but given that it was done via Skype, I felt it was more difficult to pay attention to such aspects and give feedback than doing so in person such as in a classroom. In addition, the feedback was also given via texts so that the participant could review it later on. Repeated practice is a crucial element of developing automaticity in performance. However, due to time constraints, the participants in the EdA training groups were asked to practice the performance by themselves after the Skype meeting.

4.5.2. Regular Online Meetings

Regular online meetings were conducted only with EdA-training participants three times while they were in Japan. During the meeting, the participants and I discussed their experiences including struggles and problems in Japan. As a prompt of the discussion, several discussion questions were sent to the participants prior to the meeting (see Appendix H). These discussions were aimed for the participants to see their struggles from the point of cultural differences. These meetings were used both as an extension of the pre-departure EdA training as well as a means to collect data on each participant’s progress in networking. Regarding the language, both English and Japanese were used for these meetings.

108

4.6. Procedure

In order to recruit participants for the current research, I sent emails to ten undergraduate students who participated in yearlong EdA exchange programs in Japan during the

2017-2018 academic year. As mentioned above, eight of them expressed their interests and had orientation where I explained the procedure of the second phase of the study. When a participant could not come to the campus, individual online orientation via Skype was held to explain the same contents as was in the orientation. During the orientation, I set up another individual Skype meeting with each participant. The date and time were chosen based on each student’s availability.

During each individual Skype meeting, I collected results of the pre-EdA J-CAT from the participants. I conducted an informal OPI test online via Skype with each participant. One to two weeks after the individual Skype meeting, I implemented pre-EdA training with the EdA training participants (the experimental group). The contents of the training are described in

Section 4.5.

During the EdA program, I sent emails to the participants as a reminder to log their language use for a week. The emails included a link so that they could log their language use online. The emails were sent three times during the time periods: October 2017, February, and

June 2018. In addition, in October, February, and May, I also sent emails to the EdA training participants to set up a regular meeting with each of them. The contents of the meetings during

EdA programs were described above.

After the participants returned from Japan, I set up dates for individual meetings with them. The individual meeting was held in a classroom at the students’ home institution. Before

109

the individual meeting, the participants took a Study Abroad Social Interaction Questionnaire.

During the meeting, I conducted a post-EdA Informal OPI test with each student. After the OPI test, I conducted a post-EdA interview in English.

4.7. Data Analysis

This section discusses the analysis of EdA gain that is measured through examination of style shifting and the analysis of social network development during EdA programs in Japan.

4.7.1. Analyses of EdA gain: Style shifting

After transcribing the audio recording of OPI tests, both quantitative and qualitative analyses were conducted. In the quantitative analysis, I followed a part of Iwasaki’s (2008) quantitative analysis discussed in Chapter Two. All utterances were divided into predicates (i.e., clauses). Among the predicates, a predicate that functioned as a main clause was categorized as a sentence matrix predicate (SMP). All SMPs were divided into plain forms, masu forms, fragments, and forfeited. We have already discussed the definitions of masu and plain forms in

Chapter Two so that we will discuss the definitions of fragments and forfeited here.

SMPs that are categorized as fragments or forfeited are both incomplete sentences, but they are separated because they have different effects on on-going communication. A SMP that is categorized as a fragment is a natural incomplete sentence, which does not affect the communication negatively. The use of fragments is common in Japanese spoken communication.

The excerpt below is an example of an utterance categorized as a fragment.

Example 1 1 Interviewer Fuun, jaa, anoo, ima gakusei de, isogashii to omou n desu kedo, yasumi 110

no hi wa, nani o shite imasu ka. (Well, I think you are busy because you are a student right now, but what do you do for day off.) 2 Interviewee Yasumi no hi? (Day off?) 3 Interviewer Yasumi no hi. (Day off.) 4 Interviewee → Anoo, Sapporo de? (Well, in Sapporo?) 5 Interviewer A, ieie, koko de. Ima, yasumi no hi wa, nani o shite imasu ka. (Oh, no, no, here. Now, what do you do when you have day off?)

The SMP in line 4 is missing a nominal predicate nani o shite iru to iu koto desu ka (is it that what I am doing?) or a copula, desu ka (is it?), but what the speaker wants to say is still understandable from the context, and it sounds natural. The use of fragments allows speakers to communicate effectively, and it can be considered as a communication strategy. On the other hand, SMPs that are categorized as forfeited are incomplete sentences resulting from a speaker’s lack of linguistic ability to produce a whole sentence. In other words, this type of incomplete sentences is particular to non-native speakers of Japanese, which is demonstrated below.

Example 2 1 Interviewee Ee to, uun, ma, kono jugyoo wa, etto, “Minna no Nihong”o no kyookasho o tsukai masu kara. (Well, hmmmm, well, this class, we use the textbook called Minna no Nihongo, so…) 2 Interviewee → Etto, chigau no , a… order…. JSL… Dakara…. (Well, it’s different….. um, order….. JSL…. So…)

The interviewee’s utterance in line 2 clearly shows the breakdown due to the lack of skills to construct a full sentence. This speaker ends her utterance with da kara (so), and her main point in line 2 is missing. Such incomplete sentences are separated from fragments and categorized as forfeited in this study.

111

All SMP plain forms were further categorized into four different groups, based on types of predicates. The four types of SMP plain forms are nominal predicate (NP) plain forms, verbal predicate (VP) plain forms, Adjectival Predicate (AP) plain forms, and other plain forms (see

Section 2.1. in Chapter Two for the definitions of the NP, VP, and AP predicates). Plain forms that are categorized as other includes un (a casual (plain) equivalent of hai) with a falling intonation, and un? with a rising intonation, and un maa.

Definitions of nominal, verbal, and adjectival are based on the definitions of Japanese: the Spoken Language (JSL) (Jorden & Noda, 1987), which was used in the Japanese language courses at the home institution where the participants in the current study studied Japanese prior to their EdA experiences in Japan. These definitions used in JSL are based on the forms

(conjugation) rather than meaning or functions. Using the definitions that the participants in the current study were familiar with allowed us to understand the participants’ thinking process in

EdA gain, in particular, linguistic procedural knowledge.

Although the omission of a copula in both masu and plain forms in SMPs is considered as inappropriate in prescriptive grammar, a copula in an NP in informal speech is often dropped in the spoken language (Iwasaki, 2008, p. 50). Therefore, an NP without copula (NPw/oC) was counted as a plain form SMP.

For the qualitative analysis, I followed a part of Cook’s categorization of plain forms focusing on the functions of plain forms discussed in Section 2.2.2 in Chapter Two (Cook, 2008b, p. 85). Table 6 represents acronyms and terms used in the qualitative analysis of style shifting.

112

Table 6: Terms and acronyms used in the qualitative analysis of style shifting Acronyms Terms IfSS informal speech style DtSS detached speech style IfSS-PF informal-speech-style plain form DtSS-PF detached-speech-style plain form NPF naked plain form SFP sentence final particle

According to Cook, plain forms are categorized into informal speech style (IfSS) and detached speech style (DtSS). IfSS plain forms (hereafter IfSS-PFs) are often produced with affect keys and index an “off-stage, relaxed and uninhibited” stance by the speaker. DtSS- plain forms (hereafter DtSS-PFs) are naked plain forms (NPF) that are “devoid of emotion” and often used to provide “foregrounds and referential content of an utterance (p. 85)”. Such DtSS-PFs are used in the repetition of prior utterances and soliloquy-like utterances (Cook, 2008b, p. 87).

However, such DtSS-PFs, especially for soliloquy-like utterances, are not limited to NPFs, but also plain forms with affect keys (e.g., kinoo datta ka na). Therefore, I extended the definition of

DtSS-PFs in the current study. In the current study, DtSS-PFs are not limited to NPFs, but it also includes plain forms with affect keys that index speech to oneself as well as utterances that indicate low “addressee awareness” (Maynard, 1993). In other words, DtSS-PFs in this study are plain forms that do not indicate intimacy toward the interlocutor. Such DtSS-PFs are often used for soliloquy-like utterances, exclamatory expressions (Okamoto, 1999, p. 62), suddenly recalling something, or self-reflection, and are often expressed by repeating a word or phrase

(Hudson, 2011, p. 3691). Furthermore, plain forms that are used to co-construct ideas or to provide background information are also categorized as DtSS-PFs in the current study. When a speaker uses DtSS-PFs, they tend to focus on the contents rather than the relationship with the

113

conversation partner. From this point, this study considers a speaker’s low-awareness of the addressee as the fundamental concept of DtSS-PFs. Japanese speakers use various linguistic recourses to express their state of mind, such as tone and volume of voice and use of sentence- final particles (SFPs) (i.e., kinoo datta kke53).

The judgment of IfSS-PFs and DtSS-PFs are based on types of functions (i.e., soliloquy like utterance, utterance for co-constructing ideas) and affect keys, such as intonations (i.e. rising intonation or falling intonation), tone, volume and speed of utterance. For example, if a speaker says “seikatsu” with rising intonation and indicates that the speakers wanted to request a clarification of its meaning, it is marked as an IfSS-PF. On the other hand, a word repeated with a falling intonation and a soft voice, is categorized as a DtSS-PF because it is normally interpreted that the speaker is trying to figure out or recalling the meaning of the word by him or herself instead of asking a question to the interlocutor.

4.7.2. Analysis of social network

Individual network of practice (INoP) developed by Zappa-Holloman and Duff (2015) is used for the analysis of the participants’ social networks during EdA in Japan. The focus of INoP is on an EdA participant’s social networks in different contexts and with various people. Because

EdA participants’ learning of linguistic and social practices of L2 and the target culture (C2) is not limited to the social ties within a community (e.g. attending a Japanese class, club activity,

53 When kinoo datta kke (was it yesterday?) is uttered without SFP kke, it (i.e., kinoo datta) is interpreted as a statement rather than a soliloquy of wonder. When it is uttered with a rising intonation with or without the SFP kke, it is interpreted as a question to the addressee. Thus, analysis of affect keys is crucial for the judgment of IfSS- or DtSS-PFs. 114

volunteer group), using INoP as a unit of analysis enables us to understand EdA students’ involvement in different social groups and their entire social network construction.

A diagram of INoP was constructed based on the data collected through Study Abroad

Social Interaction Questionnaire (SASIQ), Language Use Log, and post-EdA English interviews.

A diagram of INoP consists of a core, nodes, clusters and, ties. A core represents a participant whose social network is examined. Nodes are people with whom the participant (i.e., core) developed personal relationships. Based on the participant’s responses on the SASIQ and post-

EdA interviews, all nodes up to twenty were first determined. If more than twenty nodes were identified, then the participant was asked to choose the top twenty nodes that they had the strongest ties with. In order to examine social ties with Japanese people and non-Japanese people, all nodes are categorized into Japanese and non-Japanese (e.g., other international students).

Figure 5 shows core, which is presented with a striped oval (1), a Japanese node with ovals with the dotted diamond pattern (2-a), and a non-Japanese node with the white blank ovals (2-b).

Figure 5. Core and nodes in INoP

After all nodes were determined, these nodes were categorized into different social groups based on the characteristic of the groups (e.g., university, dance club, etc.). These groups are called clusters. The original INoP (Zappa-Hollman & Duff, 2015) was developed to analyze

EdA students’ social network development in an academic setting (i.e., a host university).

Therefore, EdA students’ social networks analyzed in Zappa-Hollman and Duff (2015) only included social ties within the host university. Because some EdA students also develop social

115

ties in contexts outside of host universities, social network analysis in the current study includes these social groups that are not limited to the university setting.

In order to understand the process of social network development, examining the relationships of clusters (social groups) is also important. Therefore, the current study expanded the analysis of clusters by Zappa-Hollman and Duff (2015). All clusters were organized based on whether it was within another cluster or not. If a cluster is categorized as under another cluster, then it is called a sub-cluster. For example, a cluster “dorm” is categorized as a sub-cluster of

“(host) university.” This categorization enables us to examine the relationships among the clusters. A main cluster is represented with a rectangle with triple lines, and a sub cluster is a rectangle with a single line. A sub cluster is connected with a line to indicate where it belongs.

Figure 6 represents a main and two sub-clusters.

Figure 6. Types of clusters

In Figure 6, University is a main cluster of Dorm and Roommates. Although Dorm can be considered as a main cluster of Roommates, this study does not differentiate ranks among sub- clusters. Differentiating main clusters and sub-clusters enable us to identify the contexts where

EdA students are likely to form social networks.

Finally, each node was connected with the core (i.e. a participant in the current study) with a tie. Eight different ties were used to represent a different degree of the strength of the relationships between the core and the nodes, which are represented in Figure 7 below. The left- most tie is the strongest tie, and the right-most tie represents the weakest tie.

116

Figure 7. Types of ties in INoP

Incorporating all these elements, an INoP was constructed. Figure 8 below represents an example of INoP. This INoP represents social networks developed by Alan (core) represented with a grid circle.

Figure 8. Example of an INoP diagram

Alan’s INoP includes two main clusters, University, and Part-time job, and all four sub-clusters are connected to University. In this INoP diagram, nine nodes were identified, and three out of the nine nodes were Japanese. The strong tie that Alan developed within the INoP is with Brian with whom Alan interacted in both Aikido club and Japanese class. The weaker ties that Alan constructed are with Taro and Kelly who were co-workers at Alan’s part-time job. Because the

INoP focuses on a participant’s (i.e., core) social network development instead of how the social

117

network within a community was constructed (i.e., the focus of community of practice), it provides information on the locations where the EdA student construct the social ties (e.g., mainly in the host university in this INoP), and with whom, (e.g., mostly with non-Japanese although various social status of Japanese nodes are observed) and how strong each social tie is.

By using the analysis discussed in this chapter, the following chapters, Chapter Five and

Six present and discuss findings revealed through these analyses. In Chapter Five, the findings related to EdA gain are presented and discuses along with the two research questions, and the findings related to the social networks as well as diagrams of INoPs are presented and discussed in Chapter Six.

118

Chapter 5: Understanding the Outcomes of Education Abroad Learning in Japan

5.1. Introduction

The previous chapter outlined the methodology used to investigate both the outcomes and the process of EdA learning. This chapter presents and discusses the findings regarding the seven education abroad (EdA) participants’ outcomes of EdA learning (hereafter EdA gain) in Japan.

As defined in Chapter One, EdA gain refers to second or foreign language (L2) learners’ linguistic and cultural learning during their EdA experiences that is measurable with some type of assessment, such as proficiency tests. We also discussed the limitation that not all EdA learning is reflected in EdA gain. In order to compensate for this limitation as much as possible, we employed two proficiency tests; informal OPI54 and J-CAT.55 In addition, the participants’ style shifting development was also examined to understand their EdA gain from different angles.

For the purpose of understanding the participants’ EdA gain and the effectiveness of the

EdA training toward EdA gain, this chapter investigates two research questions as follows:

1. What did the participants gain or not gain during EdA programs in Japan?

2. Are there any differences in EdA gain between the participants in the EdA

training group and the non-EdA training groups? If so, how?

54 American Council on Teaching Foreign Languages’ Oral Proficiency Interview 55 Japanese Computerized Adaptive Test 119

This chapter first presents the findings related to the participants’ background information revealed through their pre-EdA English interviews. It then presents and discusses the results of pre- and post-EdA OPI ratings, J-CAT scores, and examining style-shifting development. Finally, it discusses the findings in relation to the research questions mentioned above.

5.2. The participants

As mentioned in Chapter Four, one participant was removed from the current data because he did not submit his pre- or post-EdA J-CAT scores. Therefore, this chapter and the next chapter discuss the findings related to the seven participants who completed yearlong EdA programs in Japan. One of the seven students (i.e., Frank) also did not submit his post-EdA J-

CAT score, but he met all the other requirements of the study. Therefore, his data, except his post-J-CAT score, was also included for analysis in this chapter. All seven participants took informal OPI tests and J-CAT before and after EdA in Japan. Table 7 presents the background information of the seven participants in the current study. As we will discuss in more detail later,

Cathy, Frank, Isabelle, and Henry are high-gainers whereas Anna, Bobby, and Emma are low- gainers.

120

Table 7. Background information of the participants Name56 Type of University for Japanese Course Previous Experience in EdA Completed before EdA Japan Cathy Liberal arts college Level 2 0 Frank Liberal arts college Level 3 4 weeks for EdA Isabelle Liberal arts college Level 3 2 weeks for traveling Henry National university Level 5 0 Emma Private university Level 4 9 months + 4 weeks for EdA Bobby Private research Level 3 4 weeks for EdA university Anna National university Level 3 4 weeks for EdA

Cathy, Frank, and Isabelle attended the same liberal arts college during their EdA experiences. Their host, University CFI seemed to provide EdA students with the best environment for fostering the development of strong social ties that benefited their EdA gain.

These findings will be discussed in Chapter 6. The other four participants attended four different host universities, either private or public universities. Those universities seemed to be less structured than University CFI in terms of participants’ demonstration of EdA gain with the assessments used in the current study.

In terms of Japanese course completion prior to EdA programs, Cathy completed level- two Japanese language courses, while Frank, Isabelle, Bobby, and Anna completed level-three

Japanese language courses prior to their EdA experiences. Henry and Emma completed levels- four and five courses, respectively, before their EdA experiences.

The yearlong EdA exchange programs were the first visiting and living experience in

Japan for Cathy and Henry. Frank, Bobby, and Anna attended a four-week EdA program in

Japan prior to their yearlong EdA exchange programs, and Isabelle visited Japan for two weeks

56 The participants are given pseudonyms. 121

for travel. Among the seven participants, Emma had the most experience of living in Japan. She did a homestay for nine months as a high school exchange student, and she also attended the same four-week EdA program with Frank, Bobby, and Anna prior to her yearlong EdA exchange program this time. In the next section, we discuss the results of EdA gain.

5.3. EdA gain

This section first presents and discusses EdA gain demonstrated by comparing the results of pre- and post-EdA J-CAT scores, and the ratings of pre- and post-EdA OPIs. It then discusses

EdA gain measured by examining the participants’ style shifting in OPI interview data. The following sections discuss the findings in the order of Cathy, Frank, Isabelle, and Henry, who was also a high-gainer, but attended a different university, Emma and Bobby, who were low- gainers and whose EdA gain seemed to be limited, and lastly, Anna, who demonstrated the least gain among the seven participants.

5.3.1. EdA gain demonstrated by J-CAT

As mentioned above, all participants except for Frank took J-CAT before and after EdA in Japan. Among the six participants, who took post-EdA J-CAT, all except Anna demonstrated improvement in the total scores and proficiency levels. As we will discuss later, J-CAT seemed to be more sensitive to the participants’ EdA gain than OPI tests. Table 8 shows the seven participants’ pre- and post-EdA Japanese language proficiency measured through J-CAT conducted before and after EdA programs in Japan. The texts and numbers within the rectangles indicate the participants’ non-gain. Numbers with shading indicate the area of the highest gain

122

within the J-CAT score sub-scores, i.e., among listening, vocabulary, grammar, and reading, for that individual. Numbers in parentheses indicate improvement by points.

Prior to their EdA programs, the lowest score among them was Bobby’s at 108, and the highest was 180 by Henry. PL in the table represents Proficiency Level based on the total score provided in the official J-CAT website, which ranged from Pre-Intermediate level (PI in the table) 57 to Intermediate level (I in the table) in the pre-EdA J-CAT scores. Cathy, Bobby, and

Anna, pre-EdA J-CAT scores are interpreted as the Pre-Intermediate level and Frank, Isabelle,

Henry, and Emma’s and pre-EdA scores are interpreted as Intermediate level.

Table 8. EdA gain demonstrated by J-CAT Name J-CAT* T** PL L V G R Pre > Post Pre > Post Pre > Post Pre > Post Pre > Post Pre > Post Cathy 127 > 222 PI > IH 52 > 71 15 > 49 26 > 45 34 > 57 (95) (19) (34) (19) (23) Frank 153 > N/A I > N/A 45 > N/A 40 > N/A 26 > N/A 42 > N/A Isabelle 160 > 219 I > IH 44 > 59 41 > 54 36 > 45 39 > 61 (59) (15) (15) (9) (22) Henry 180 > 217 I > IH 49 > 64 45 > 48 34 > 51 52 > 54 (37) (15) (3) (17) (2) Emma 158 > 211 I > IH 48 > 58 34 > 48 45 > 47 31 > 58 (53) (10) (14) (2) (27) Bobby 108 > 151 PI > I 29 > 49 11 > 34 41 > 26 27 > 42 (43) (20) (23) (-17) (15) Anna 144 > 126 PI > PI 29 >19 15 > 34 70 > 40 30 > 33 (-18) *** (- 10) (19) (- 30) (3) * The symbol “>” in this and other pre-/post- comparison tables refers to the change from pre- to post- measurements, rather than to its usual reference to one entity being “greater than” the other. ** T = total, PL = proficiency level, L = listening, V = vocabulary, G = grammar, R = reading. ***Numbers in the parentheses indicates gain by points

57 Intermediate-level learners are capable of conducting daily conversation (http://www.j- cat.org/html/ja/pages/interpret.html) 123

The participants’ proficiency level ranged from Pre-Intermediate to Intermediate-High

(IH) level in the post-EdA J-CAT, and all except Anna demonstrated improvement in the total scores and the proficiency levels. For example, Bobby improved from Pre-Intermediate to

Intermediate, and Isabelle, Emma, and Henry improved from Intermediate to Intermediate-High.

Cathy demonstrated the largest gain in proficiency level (i.e., from Pre-Intermediate to

Intermediate-High) and in scores (i.e., from 127 to 223). The smallest score gain among the five participants who demonstrated improvement (i.e., Cathy, Isabelle, Henry, Emma, and Bobby,) is

Henry whose score improved from 180 to 217, but even within the smallest increment, his proficiency level improved from Intermediate to Intermediate-High. Anna was the only one whose score declined from 144 to 126, but her proficiency level remained the same, i.e., Pre-

Intermediate.

Regarding the gain in different skills, all participants demonstrated the gain in vocabulary

(V in the table) and reading (R in the table), but the degrees of the score gains in each section depends on each participant. For example, Anna, Bobby, and Cathy, whose scores were interpreted as Pre-Intermediate in the pre-EdA J-CATs, gained most in the vocabulary section, and Emma and Isabelle, who were Intermediate in the pre-EdA J-CATs, improved their scores most in reading. Henry, who was also Intermediate-level but had the highest score in the pre-

EdA J-CAT, improved the score most in the grammar (G in the table) section. However, demonstrating the gain in the grammar section seems to be the most difficult for other participants. For example, Cathy, Emma, and Isabelle demonstrated gains in the grammar section, but their gains were the smallest among the gains in the other sections. Anna and Bobby’s scores

124

in the grammar section in the post-EdA J-CATs dropped by thirty points for Anna and by fifteen points for Bobby. Anna’s listening score also dropped by ten points.

As mentioned above, Anna’s pre- and post-EdA J-CAT total score indicates regression on her language proficiency. It might have been partially due to the reliability of the test score.

Anna described that her pre-EdA J-CAT score might not have been accurate and might have included “a few lucky guesses.” She also explained that she felt better when taking the test after her EdA experience in Japan than before and was “much less stressed” in trying to figure out answers. Therefore, although the scores indicate regression in Anna’s language proficiency, it is difficult to conclude that her Japanese language proficiency in receptive skills actually regressed.

In Section 5.4, we further discuss the findings of the J-CAT scores in relation to comparison with the other two measurements during the EdA programs. As mentioned earlier, the J-CAT scores reflect mostly the participants’ receptive skills. The results of J-CAT scores do not demonstrate what they can do in terms of productive skills. Therefore, the next section focuses on the seven participants’ EdA gain in productive skills, specifically oral proficiency measured with the informal OPI tests.

5.3.2. EdA gain demonstrated with informal OPI tests

The overall OPI ratings indicate that four participants (i.e., Cathy, Frank, Isabelle, and

Henry) demonstrated improvement in oral proficiency, but the other three participants (i.e., Anna,

Bobby, and Emma) did not. Among the participants who demonstrated improvement in the post-

EdA OPIs, Cathy, Frank, and Isabelle made improvement by one sub-level (Low to Mid, and

Mid to High) and Henry made improvement by one major-level (Intermediate to Advanced).

125

Table 9 demonstrates the seven participants’ pre-EdA OPI ratings on the left half and post-EdA OPI ratings on the right half. It also includes proficiency-levels based on pre- and post-

EdA J-CAT scores for the purpose of comparison. Pre-EdA OPI ratings ranged from Novice-

High (Anna’s) to Intermediate-High (Henry and Emma’s). As mentioned in Chapter Four, a

Novice-level speaker’s communication is often limited to “most common informal settings” with memorized words or phrases, whereas an Intermediate-level speaker can communicate “in some informal settings and a limited number of transactional situations” by combining and recombining what they know. The pre-EdA OPI ratings suggest most participants were able to communicate in predictable informal settings at the beginning of their EdA programs, but some of them (Cathy, Isabelle, and Bobby) might not have been confident or comfortable in some informal settings and might have struggled in such contexts.

Table 9. Participants' pre- and post-EdA OPI ratings Name Pre-EdA OPI Post-EdA OPI Pre-EdA J-CAT Post-EdA J-CAT Cathy Intermediate-Low Intermediate-Mid Pre-Intermediate Intermediate-High Frank Intermediate-Mid Intermediate-High Intermediate N/A Isabelle Intermediate-Low Intermediate-Mid Intermediate Intermediate-High Henry Intermediate-High Advanced-Low Intermediate Intermediate-High Emma Intermediate-High Intermediate-High Intermediate Intermediate-High Bobby Intermediate-Low Intermediate-Low Pre-Intermediate Intermediate Anna Novice-High Novice-High Pre-Intermediate Pre-Intermediate

Upon their returns, the range of proficiency level widened, which ranged from Novice-

High (Anna’s) to Advanced-Low (Henry’s). In the post-EdA OPI ratings, most of them were now able to converse at the Intermediate-level comfortably. Furthermore, Henry, who was the only to demonstrated gain at a major level, demonstrated that he could handle most informal and

126

some formal settings as well as discussion of personal and general interests by utilizing narration and description skills.

In addition to Anna whose post-EdA J-CAT score did not indicate her improvement, the

OPI ratings did not demonstrate a significant gain in Bobby and Emma’s oral skills after the yearlong EdA programs in Japan. The differences in the participants’ EdA gain between J-CAT and OPI suggests two points. First, considering the two proficiency tests measured different skills of Japanese proficiency (i.e., J-CAT measures receptive skills in vocabulary, listening, reading, and grammar, and OPI measures oral productive skills), the differences may suggest that demonstrating improvement in productive skills were more difficult than that of receptive skills.

Second, the differences may suggest that J-CAT may be more sensitive to the participants’ gain than OPI, especially for participants who started their EdA programs with a higher proficiency level. As the diagram of a reverse triangle in ACTFL OPI Guideline (ACFTL, 2012), it is more difficult and takes longer to move up to the next level as the proficiency level goes up. Therefore,

Emma, who demonstrated the highest proficiency in pre-EdA OPI along with Henry, might have improved her oral skill but her improvement was not great enough to demonstrate the improvement in the OPI rating.

So far, we have discussed EdA gain or non-gain measured with J-CAT and OPI tests.

However, these proficiency tests provided a limited understanding of EdA gain in terms of the participants’ communication skills in a culturally appropriate manner, which is a crucial aspect of building intercultural relationships. For this reason, the current study also examined the participants’ style shifting development. The following sections present and discuss the findings from the analysis of style shifting development.

127

5.3.3. EdA gain demonstrated with the examination of style shifting

The utterances from OPI interviews were used as data for the analysis of style shifting development. The OPI audio data was transcribed, and all utterances were divided by predicates.

Each predicate is categorized into sentence matrix-final predicate (SMP) or subordinate predicate

(see Chapter Four for more detailed categorizations of the two predicate types). The next two sections focus on the findings from the quantitative and qualitative analyses.

5.3.3.1. Findings from quantitative analysis

Overall findings suggest that most participants either produced more sentences or longer and more structurally complex sentences upon their return. In addition, all of the participants used masu forms predominantly before and after EdA programs in Japan, but all, except Bobby, increased the use of plain forms (PF) in their overall utterances.

Table 10 shows the total numbers and proportions of utterances by categories. It includes numbers and proportions of predicates (P in the table), total numbers of SMPs, masu form SMPs, plain form SMPs, fragments (frg. in the table), and forfeited (frf. in the table). SMPs that are categorized as “fragment” includes omission of a predicate that was understood in previous utterances, such as “Sapporo de (shimashita),” which is a natural omission. On the other hand,

SMPs that are categorized as “forfeited” results from a speaker’s lack of language proficiency. In other words, such incomplete sentences are considered as mistakes (see Section 4.7.1 in Chapter

Four for an example of forfeited SMP). Numbers in the brackets indicate proportions, which were calculated by dividing the number in each SMP type by the total number of SMPs.

128

Table 10. Occurrences* and percentage of utterances by categories Name Pre-EdA OPIs Post-EdA OPIs # of # of # of # of # of # of # of # of # of # of # of # P SMP masu plain frg. frf. P SM masu plain frg. of forms forms P forms forms frf. Cathy 133 103 84 14 5 0 236 156 122 28 6 0 [%] [81.6]** [13.6] [5] [78.2] [17.9] [4] Frank 235 144 131 10 3 0 293 175 130 32 9 4 [91] [7] [2] [74.3] [18.9] [5.1] [2] Isabelle 120 103 95 7 1 0 175 71 54 10 7 0 [92.2] [7] [1] [76.0] [12.7] [10] Henry 142 114 109 2 3 0 195 117 87 17 14 0 [95.6] [2] [2] [75] [14.5] [12] [0] Emma 216 118 109 6 3 0 217 104 72 19 10 3 [92.3] [5] [3] [69.2] [18.3] [9.6] [3] Bobby 140 108 83 21 2 2 109 87 68 13 6 0 [76.8] [19.4] [2] [2] [78.1] [15] [7] Anna 72 63 55 6 0 2 88 82 64 12 5 1 [87.3] [10.0] [0] [3] [78.0] [14.6] [6] [1] * Because occurrences of the same utterance were no more than three, multiple iterations of the same utterance were counted as individual occurrences. ** The numbers in brackets ([]) indicates the percentages.

Four participants, Cathy, Frank, Henry, and Anna increased the numbers of SMPs in the post-EdA OPIs. In other words, they produced more sentences upon their return. In addition,

Isabelle and Emma produced fewer SMPs in post-EdA OPIs than pre-EdA OPIs, but they increased the total number of predicates, which indicates they produced longer and more structurally complex sentences upon their return.

All of the participants used masu forms predominantly before and after EdA programs in

Japan. However, the proportions of the plain forms have become larger in the post-EdA OPIs than the pre-OPIs for all participants, except Bobby. The findings align with the findings in

Iwasaki (2010) and Marriott (1995) that L2 learners increased the use of plain forms after EdA programs in Japan. Masuda’s (2010) study also suggests that there are correlations between the use of plain forms and the experience of living in Japan. 129

In the pre-EdA OPIs, the participants who were rated as Intermediate-Mid or higher (i.e.,

Frank, Emma, and Henry) tended to produce less plain forms than the participants with lower

OPI ratings (i.e., Cathy, Isabelle, Bobby, and Anna). Upon their return, however, there seems to be no relation with the use of plain forms and the proficiency level, and most of the participants increased the use of plain forms. This overall phenomenon of increased use of plain forms among the current study indicates that L2 learners learn to use plain forms more frequently from yearlong EdA experience, regardless of their proficiency levels. This further suggests EdA contexts can provide more interactional opportunities for L2 learners to observe and practice plain forms than a domestic learning environment.

The findings of the predominant use of masu forms before and after their EdA experiences suggest the participants’ cultural knowledge of the appropriate speech style in a formal interview setting even after a yearlong EdA program. This finding does not align with the findings in Marriott (1995) in which all the high school participants overused plain forms after

EdA programs in Japan. This might have resulted from the differences in interactional opportunities between high school EdA students and university EdA participants: EdA learning environments for high school students may have been more informal than that of college EdA settings. By contrast, Iwasaki (2008) who investigated four college students’ speech styles before and after EdA programs in Japan found individual differences. Two of the participants in

Iwasaki’s study overused plain forms, but other participants still maintained the predominant use of masu forms after their EdA programs. Because the participants with higher OPI ratings tend to use masu forms predominantly in her study, Iwasaki (2008, p. 50) assumed that the use of

130

appropriate styles and shifts correlates with proficiency levels measured with OPIs.58 However, considering that all participants in the current study whose pre- and post-EdA OPI ratings varied from Novice-High to Intermediate-Low, the findings in the current study do not support this assumption. In fact, similar to the current study, the four participants in the Iwasaki’s study were all from the same home institution in the US. This suggests, in addition to the differences in each participant’s learning experiences during EdA in Japan, previous learning experience on speech style might have also played an important role in appropriate speech style acquisition during

EdA programs in Japan.

Furthermore, the examination of the participants’ speech style in an informal role-play activity59 revealed that all participants except Anna were able to use plain forms predominantly in the post-EdA OPIs.60 Considering that only Cathy, Emma and Frank were able to use plain forms predominantly in the informal role-play setting in the pre-EdA OPIs, these findings suggest gain in the participants’ linguistic skills to produce Japanese plain forms and in the cultural knowledge of the appropriate speech style in informal settings.

In terms of the use of fragments, all participants, except Cathy, slightly increased the use of this category at SMPs in the post-EdA OPIs. Considering that L1 speakers tend to use fragments more than L2 speakers (Masuda, 2010, p. 195), the increased use of fragments by the

58 In Iwasaki’s (2008, p. 50) study, participants with higher OPI ratings used appropriate styles and shifts compared with the participants with lower OPI ratings. 59 In the role-play, the participants were asked to talk with their friend on the phone. 60 However, there was different degree of IfSS, such as use of formal first pronoun “watashi” by a male speaker as well as sentence final particles. See Section 5.2.3.2 in this chapter for detailed discussion of each participant’s speech style development. 131

current participants suggests their speech style becomes more or less similar/closer to that of L1 speakers.

For further analysis of plain forms, I categorized all plain-form SMPs into nominal predicate (NP), verbal predicate (VP), adjectival predicate (AP) and others (see Section 2.2.2 in

Chapter Two for the detailed discussion of the definition on each predicate). As mentioned in the methodology chapter (see Section 4.7.1 in Chapter Four), the sentences that end with a nominal without a copula are categorized as an NP (hereafter NPw/oC). Table 11 shows the numbers and proportions based on this type of analysis.

Table 11. Occurrences* and percentage of types of plain form SMPs Name Pre-EdA OPI Post-EdA OPI # of # of # of # of # of # of # of # of # of # of PF NP VP AP Oth PF NP VP AP Oth. Cathy 14 9 3 1 1 28 23 4 1 0 Frank 10 6 0 4 0 32 16 11 2 3 Isabelle 7 3 2 2 0 10 7 1 2 0 Henry 2 1 1 0 0 17 11 6 0 0 Emma 6 6 0 0 0 19 13 4 2 0 Bobby 21 16 4 1 0 13 9 1 1 2 Anna 6 4 1 1 0 12 12 0 0 0 Total 66 45 11 9 1 131 91 27 8 5 [%] [68.0] [16.6] [13.6] [1.5] [69.5] [20.6] [6.1] [3.8] * Because occurrences of the same utterance were no more than three, multiple iterations of the same utterance were counted as individual occurrences.

In the pre-EdA OPIs, approximately 70% of PF-SMPs are NPs, and VPs and APs are accounted for approximately 15% of all PF-SMPs respectively. In the post-EdA OPIs, the proportions of NPs, VPs and Other became greater by 1.5% for NPs, 4% for VPs, and 2.3 % for

Other.

132

Regarding NPs, most NPs are NPs without a copula (NPw/oC). Only three out of the 49

NPs in the pre-EdA OPIs and three out of ninety in the post-EdA OPIs are NPs with a copula

(NPwC), such as “nan da kke” (what was it?), “hon-dana da kke” (was it (called) hon-dana

(bookshelf)?), and “sono hito wa nan datta ka.” (what was the person…) The percentage of

NPw/oC may also include omissions of VPs or APs. In Japanese, it is possible to state new information within the same topic without mentioning the whole topic if the topic is clear to the conversation partner. The excerpt below demonstrates this point.

Excerpt 5 (Cathy, Post-EdA OPI, rated Intermediate-Mid) 1 Interviewer Donna mono desu ka, sono joohoo no fesutibaru to iu no wa. (What kinds of things is it, that thing called an “informational festival”?) 2 Cathy Maa, hakimarimasu, wa, ookii, heya de, takusan kurabu ga suteeji de, kurabu ni tsuite hanashi mashita. (Well, it starts, at a big, room, many clubs talked about clubs on the stage.) 3 Interviewer Dare ga hanashita n desu ka (Who did the talking?) 4 Cathy → kurabu no menbaa. (Club members) 5 Interviewer Aa, hai hai. (Ah, I see, I see.)

In line 1, the interviewer asks what Cathy meant by joohoo no fesutibaru. In line 2, Cathy explains that many clubs talked about their own clubs on the stage. In line 3, the interviewer asks who did the talk, and Cathy answers with a NPw/oC, kurabu no menbaa (the club members) in line 4. This could be kurabu no menbaa desu (it was club members) or kurabu no menbaa ga hanashimashita (club members talked about it). In the latter case, it would be categorized as a

VP. However, it is difficult to know from the context which type of predicate Cathy intended

133

when she uttered kurabu no menbaa61. Therefore, for a case like this, I categorized the predicate as NPw/oC.

As I motioned above, stating only the topic and omitting the rest of the predicate mostly occurs in informal conversation in Japanese. However, I argue that the participants’ frequent use of NPw/oC in a formal speech context stems from a lack of linguistic abilities to end a sentence with desu forms rather than that they were using NPw/oC as informal speech (which can be considered as a lack of cultural knowledge) for three reasons. First, research on first language acquisition has reported that function words often develop later than content words (Tomacello,

2002; and Pinker, 1982). A copula carries grammatical information, such as aspect (perfective or imperfective) rather than semantic information.62

Second, conjugations of VP and AP plain forms are structurally more complex than that of an NP (i.e., as long as a speaker knows how to conjugate the copula, he or she only needs to conjugate the copula in all NPs). Much less use of VP and AP plain forms in the participants’ utterances may also support this assumption.

Third, some participants did not use a copula (da or datta) when it was obligatory, which also supports a lack of their linguistic abilities. The following excerpt demonstrates this point.

61 Yoshimura & Nakayama (2018) report that Japanese L2 speakers whose L1 is English tend to replace the verbal predicate with the copula desu in their responses when the focus of a question is a subject of the predicate (e.g., Q: Dare ga oshiemashita ka. R: Tanaka-sensee desu). On the other hand, when the focus of the question is an object, the L2 speakers tend to state the whole sentence (e.g. Q: Nani o benkyoo shimashita ka. A: Nihongo o benkyoo shimashita.). The response strategy may reflect the EdA gain (M. Nakayama, personal communication). However, because the current study examined EdA students’ style shifting development, I did not investigate the relationship between the questions and responses regarding the omission of predicates. 62 However, in some cases, use or non-use of da also influences the speech style. For example, X da yo/ne sounds more blunt than X yo/ne (Jorden & Noda, 1988). 134

Excerpt 6 (Bobby, Post-EdA OPI, rated Intermediate-Low) 1 Interviewer A, osake mo ii yo. Donna osake katte kuru? (Um, yeah, alcohol is also good. What kinds of alcohol do you buy?) 2 Bobby → aa, anoo, okane wa tyotto* kedo, (ummm, well, money is a little, but,) 3 Bobby ii osake o katte ru. (I am buying good alcohol.)

In line 2, Bobby does not use copula da before kedo. Because kedo is a clause particle and it follows a predicate, a copula is obligatory when it is used with an NP even in the spoken language. Therefore, an utterance like line 2 in Excerpt 6 is considered as a grammatical error and suggests that the speaker does not have a good command of using a copula in his utterance.

Such errors were also observed in other participants’ utterances (i.e., ichi-nen-kan gurai *(da) soo desu.) in pre- and post-OPIs.

5.3.3.2. Findings from qualitative analysis: Focus on each participant

In order to gain a better understanding of style shifting development specifically in relation to cultural knowledge, this section presents findings from qualitative analysis. In the qualitative analysis, all plain form SMPs are classified into informal speech style plain forms

(hereafter IfSS-PFs) or detached speech style plain forms (DtSS-PFs) based on the categorization described in Chapter Four. Plain form SMPs considered as IfSS-PFs, when they indicated intimacy toward the addressee, and all other plain forms, such as soliloquy-like utterances or utterances focusing on contents rather than the relationships with the addressee including co- constructing ideas, providing background or additional information, were categorized as DtSS-

PFs (see Section 4.7.1. in Chapter Four).

135

5.3.3.2.1. Cathy

Cathy’s EdA gain in oral proficiency was apparent in OPI ratings: Her oral skill was rated as Intermediate-Low in the pre-EdA OPI, but it was Intermediate-Mid in the post-EdA. Her improvement in oral skill was also apparent from the quantitative analysis where she produced nearly twice as many predicates in the post-EdA OPI as the pre-EdA OPI. She also increased her use of plain forms slightly in the post-EdA OPI.

Qualitative analysis also suggests her gain in style shifting. Cathy used DtSS-PFs only for soliloquy-like utterances, and all of them were plain forms without any affect keys (i.e., sentence-final particles or rising intonations), such as arugorizumu (trying to pronounce the word algorithm with Japanese pronunciation following the interviewer’s model), booru produced with a slow and soft voice (for digesting the word by repeating the word after the interviewer told that

“a ball” in Japanese is also called booru), and muzukashii (a conviction to herself). In the post-

EdA OPI, she used DtSS-PFs more often and for more various functions. Below are some examples of Cathy’s DtSS-PFs use in the post-EdA OPI.

Excerpt 7 (Bobby, Pre-EdA OPI, rated Intermediate-Low) 1 Interviewer A, soo desu ka. Donna arubaito o shite imashita ka. (Oh, is that so. What kinds of part-time job did you do?) 2 Cathy Sore wa, kookoo no tekunorojii-bu de, tetsudai shimashita. (That was, I helped the department of technology at a high school.) 3 Interviewer Tekunorojii-bu. Donna tetsudai o shita n desu ka. (The department of technology. What kinds of help did you do?) 4 Cathy → A, kono natsu wa, takusan….. namae wasurechatta. (Oh, this summer, many…. I forgot the name…) 5 Cathy → Chiisai konpyuutaa nan desu kedo. (It was a small computer, but,) 6 Cathy → Kuronbukku! (Chromebook!) 7 Cathy → Hai. (Yes.)

136

8 Cathy Etto, sono kookoo wa, kyuuhyaku no kuronbukku o kai, katte, watashi ga, sono kuronbukku wo hiraite, settoappu-shite, soshite, aa, biru ni tsurete ikimashita. (Well, the high school bu-, bought 900 Chromebooks, I opened them, set them up, and brought to a building.)

In Excerpt 7, Cathy is describing a part-time job that she did during summer vacation. Up until line 4, she uses masu forms to answer the interviewer’s questions regarding the job. At the end of line 4, she pauses a few seconds after takusan (many) and then utters wasuretyatta (I forgot the name [of the computer].), a VP-PF, with a soft voice. Her soft voice suggests it is a low- addressee-awareness situation (Maynard、 1993, p. 175) and is considered as a DtSS-PF for soliloquy-like utterance rather than she intends to deliver the message to her interlocutor.

In line 5, she is describing a characteristic of the computer saying chiisai-konpyuutaa nan desu kedo (it was a small computer but), in which she uses the masu form. Her use of the masu form here indicates that she now has high addressee-awareness, and it is clear that the message is uttered to the interviewer. In line 6, she suddenly recalls the name of the computer and says,

“kuronbukku!,” an NPw/oC, which also indicates her low addressee-awareness situation

(Maynard 1993), therefore, it is categorized as a DtSS-PF. In lines 7 and 8, she shifts back to masu forms, which indicates that she is now talking to the interviewer. In the post-EdA OPI,

Cathy used more variety of linguistic resources to mark her utterance as DtSS-PFs for soliloquy, such as using multiple SFP kana (e.g., kinoo datta kana) than the pre-EdA OPI. She also incorporated DtSS-PF to provide background information in her utterances.

Cathy was able to use IfSS-PFs in a casual setting both before and after her EdA experience in Japan. Among the three participants who were rated as Intermediate-Low (i.e.,

137

Bobby, Cathy, and Isabelle) in the pre-EdA OPI, Cathy was the only one who could sustain the use of IfSS-PFs in a casual setting, which suggests even among the same proficiency level learners, their abilities in speech style and style shifting vary, and the developments of the overall oral skill and the style-shifting skill do not correlate to each other.

Overall, the increased use of plain forms in Cathy’s post-EdA OPI suggests that her speech became more approachable while maintaining the predominant use of masu forms. In addition, she used more DtSS-PFs with more variety of linguistic recourses (i.e., using SFPs) for more variety of functions (soliloquy-like utterances and for co-construction), which demonstrates her EdA gain in oral proficiency.

5.3.3.2.2. Frank

Frank’s OPI ratings demonstrate his EdA gain. His pre-EdA OPI was Intermediate-Mid and his post-EdA rating was Intermediate-High. Quantitative analysis revealed that he increased his use of plain forms much more than anyone else. In the pre-EdA OPI, he used plain forms 10 times in SMPs, which accounted for 7% of the SMPs. In the post- EdA OPI, he used plain forms

36 times in SMPs, which accounted for 20.6% of the SMPs.

Qualitative analysis revealed that he was able to use IfSS-PFs in a casual setting before and after EdA in Japan. It also revealed that he used DtSS-PFs for soliloquy-like utterances and for co-construction. Excerpt 8 shows his use of the plain form for co-constructing an idea with the interviewer.

Excerpt 8 (Frank, Pre-EdA OPI, rated Intermediate-Mid) 1 Interviewer Moo hitotsu yonde iru hon ga aru tte, osshatte imashita ne. Nan deshita kke. Geemu, geemu, (One more book, you said you have read, what was it? Game, game…) 138

2 Frank → geemu, obu, toroonzu. (Game of Thrones) 3 Interviewer a, sore sore, sono hon wa, zenbu yomimasita ka. (Oh, that one, that one, did you finish reading it?) 4 Frank un, zenbu yomimashita kedo, anoo, mada tsugi no hon o matte imasu. (yeah, I read them all, but well, I am still waiting for the next book.)

In line 1, the interviewer is about to ask a question regarding the book Frank mentioned before, but she cannot recall the title of the book. She asks Frank to say the title one more time by saying

“nan deshita kke” (what was it?) and stated the part that she remembers, “geemu, geemu.” (game, game,) In line 2, Frank picks up the word “geemu” and completes the whole title for the interviewer by saying “geemu obu toroonzu,” (Game of Thrones) which is an NPw/oC. This use of plain forms for co-construction of an idea is often observed in Japanese L1 speakers’ conversation (Cook, 2006; Ikuta, 2008, etc.). In line 4, although he responds to the next question sono hon wa yomimashita ka (did you read the book?) in IfSS un (yes) first, he shifts to the masu form when explaining that he is still waiting for the next book to come out. Although un can be considered an inappropriate IfSS here, by shifting to the masu form right after, he is able to minimize the negative effect of this speech.

In the post-EdA OPI, he used DtSS-PFs for soliloquy-like utterances and for utterances in which he focused on the content (Makino, 2002). Excerpt 16 shows his use of a DtSS-PF for an utterance where he focused on the content.

Excerpt 9 (Frank, Post-EdA, rated Intermediate-High) 1 Frank Umi ni anoo, modotta kedo, sono ato, hito, hito, anoo, fune ni iru hito wa, ma, a, anoo, kurokute ikimashita. (To the ocean, well, they returned but, the people, the people, the people on the ship, well, started losing their mind (with a wrong pronunciation)) 2 Frank → U? A, anoo, kuruu? (Um, w, well, lose their mind?)

139

3 Interviewer Kuruma? (A car?) 4 Frank → Kuruu. (Lose their minds.) 5 Interviewer A, kuruu. (Oh, lose their minds.) 6 Frank → Kuru, hai, kurutte kimashita. (lose, yes, they started losing their mind.) 7 Frank Anoo, anoo, hitori hitori, kurutte kimashita. (Well, one by one, they started lost their minds.)

In the previous utterances and in line 1, Frank has been explaining the plot of the book that he recently read. In line 1, he is not able to conjugate the verb kuruu (lost one’s mind) to the gerund form, kurutte, and said kurokute kimashita. After saying it aloud, he seems to have noticed his mistake and produces the plain form of the verb kuruu with a rising intonation, which indicates his uncertainty with the word conjugation, and his wish to confirm it with the interviewer. This plain form is considered as an IfSS-PF. In line 3, the interviewer does not hear the word correctly and asks if what Frank said is kuruma (a car). Frank repeats the word kuruu without saying yes or no to the interviewer’s question. Japanese L1 speakers often use this type of plain form (i.e., when focusing on the content,), which Cook (2008, p. 85) categorized as DtSS-PFs.63 Frank’s use of DtSS-PF in line 3 indicates that he focused on the content rather than the relationship with his interlocutor in order to continue his unfinished story. After confirming the interviewer understood what he meant to say, Frank continues the narrative of the story in masu forms in line

6, which indicates he again acknowledges the formal relationship with the interviewer as well as

63 In Cook (2006), she calls such plain forms as impersonal plain forms, but in Cook (2008) she mentioned that she changed the terminology to DtSS plain form. 140

the formal discourse context that he is in. This series of style shifting demonstrated that Frank was able to shift his style appropriately in order to deliver his messages effectively.

In addition, he demonstrated the appropriate use of SFP kke in hon-dana da kke (was it call hon-dana?) for recalling the word hon-dana (bookshelf) in the post-EdA OPI. Such SFP use was absent in the pre-EdA OPI, demonstrated in Excerpt 10 below.

Excerpt 10 (Frank, Post-EdA OPI, rated Intermediate-High) 1 Frank Anoo, fuun, suu, aa, sore de, anoo, sono mono wa, fantajii to iu mono desu kedo, kono hon wa, seiji, seiji no kankee ga yoku takusan arimasu. (Well, hmmmm, shhh, ummm, then, well, that thing is called fantasy, but this book has politics, things, well, that relate to politics.) 2 Frank → Seiji, (Politics) 3 Frank → Un, soo, anoo, (Yeah, right, well.) 4 Interviewer Seiji wa anoo, amerika no seiji to iu koto desu ka. (Politics, well, is it American politics?) Frank Uuun, ieie, igirisu no mae no seiji mitai desu. (Hmmmm, no, no, it was like, previous British politics.)

In line 2, Frank repeats seiji (politics) one more time after saying seiji kankee ga yoku, takusan arimasu (this book has a lot of politics related topics). By doing so, he checks whether the word is correct or not, and then he confirms it is correct in line 3 by saying un, soo (yes, that’s right).

Utterances in lines 2 and 3 are produced in a softer voice than the previous utterances, therefore, they are categorized as soliloquy-like utterances. In line 5, he shifts to masu forms when he responds to the interviewer’s question, which also suggests that the previous utterances are directed to him.

141

Similar DtSS-PF use was observed in the post-EdA OPI, but this time, he used SFP kke for recalling to clearly mark the utterance was a self-directed, soliloquy-like utterance (Okazaki,

2015, p. 153), as shown in Excerpt 11.

Excerpt 11 (Frank, Post-EdA OPI, rated Intermediate-High) 1 Frank Anoo, nan ka, hon-dana ga arimasu. (Well, you know, there is a bookshelf.) 2 Frank → Hon-dana, (Bookshelf,) 3 Frank Nan to iu (What is it call…) 4 Frank → Anoo, hon-dana… (Well, a bookshelf,) 5 Frank Hon-dana? (Bookshelf?) 6 Frank → Hon-dana da kke. (Was it bookshelf?) 7 Frank → ga, arimasu. (There is.)

In the previous utterances and line 1, he is describing his room in masu forms. Lines 1 to 7 are all

Frank’s utterances, in which he is trying to figure out if the word hon-dana is the right word, and he is doing so by repeating the word a few times. In line 3, Frank utters nan to iu, which is an

IfSS-PF, although he might have intended it as a soliloquy-like utterance.64 The same use of nan to iu was also observed several times in Emma’s post-EdA OPI, which will be discussed more extensively in her section. Likewise, a rising intonation after hon-dana makes the utterance as

IfSS interrogative. In line 6, he again repeated Hon-dana this time with kke SFP, which marks the utterance clearly as a soliloquy-like utterance for recalling (Cook, 2006; Masuda, 2010 etc.) and also functioned as an indicator that all of his previous utterances (line 2 to line 6) are

64 Japanese L1 speaker would likely say nan te itta kke or nan daroo in the same context (Okazaki, 2015, p. 153). 142

intended to be soliloquy-like utterances. Although starting a predicate with phrase particle ga in line 7 can be considered grammatically incorrect, it is clear that hon-dana is the subject of the predicate and sounds natural in this context. By using masu form in line 7, Frank demonstrates his cultural knowledge to communicate appropriately in the context.

Although Frank was able to use DtSS-PFs for recalling without SFP prior to his yearlong

EdA program in Japan as shown in Excerpt 8, he started utilizing SFP to mark DtSS-PFs more clearly. In addition to the increased use of the plain forms in the post-EdA OPI without sounding too formal, Frank’s use of kke in the post-EdA OPI demonstrated his EdA gain in the use of

DtSS-PFs.

5.3.3.2.3. Isabelle

Isabelle’s EdA gain was also apparent in her OPI ratings. Her speech was Intermediate-

Low in the pre-EdA OPI, but it was Intermediate-Mid in the post-EdA OPI. Quantitative analysis of her OPI data shows that she decreased the number of SMPs but increased the number of predicates, which suggests that she produced longer and more complex sentences in the post-

EdA OPI. Although qualitative analysis indicates improvement in her linguistic and cultural knowledge, it also suggests that her gain in DtSS-PF use was limited.

In the pre-EdA OPI, she used IfSS-PFs inappropriately when making a request to the interviewer in pre-EdA OPI, however, such use was not observed in post-EdA OPI. Excerpt 12 below shows her inappropriate use of IfSS-PF in the pre-EdA OPI.

Excerpt 12 (Isabelle, Pre-EdA OPI, rated Intermediate-Low) 1 Interviewer Dare ga imasu ka. (Who is there (in your family)?) 2 Isabelle → Anoo, chotto, matte. 143

(Umm, wait a little bit,) 3 Isabelle → anoo, chotto kikoenikui, kikoenikukatta… (Well, it is little difficult to hear, it was little difficult to hear) 4 Interviewer A, izabera-san no kazoku ni tsuite oshiete kudasai. (Oh, please tell me about your family.)

In line 2, she asks the interviewer to wait for a second by using a verbal gerund form, matte, which is considered an informal request form. She then provides the reason for her request in a plain form kikoenikukatta (it was difficult to hear) in line 3. Isabelle clearly has the interviewer in her mind when making a request and giving a reason for the request. This case is considered a high addressee-awareness situation instead of low address-awareness situation (Maynard, 1993, p. 178). Therefore, Isabelle’s use of IfSS-PFs in the two speech acts in this context is considered culturally inappropriate.

However, upon her return, she demonstrated not only her cultural knowledge to recognize her inappropriate speech style but also her linguistic procedural knowledge to correct herself naturally, which is demonstrated in the excerpt below.

Excerpt 13 (Isabelle, Post-EdA OPI, rated Intermediate-Mid) 1 Isabelle tabun… sono onna no ko wa, takusan, chuugaku no ni-nen-see mitai na, anoo, keiken ga aru kedo, (Maybe... that girl has many experiences in doing things, well, that eight- graders would likely do, but) 2 Isabelle → sore wa chotto, muzukashii yone. That is a little difficult SFP That is little bit, difficult, you know. 3 Isabelle → anoo, amerika-jin ja nakere ba, chotto muzukashii to omoimasu. (Well, I think it is difficult (to understand it) if you are not an American.)

In the previous utterances, Isabelle was describing what she means by “koukou-sei mitai na koto

(things that high school students are likely to do).” Her utterance in line 1 suggests that she is

144

having difficulties performing this task. In line 2, she expresses the difficulty by using an adjectival plain form, muzukashii, with multiple SFPs, yone. SFP yo, ne, and yone require some changes in the interlocutor’s cognition, and it may add an “obtrusive” tone, and many Japanese

L1 speakers often avoid using it when talking to someone who holds higher social status (Izuhara

2003). The AP with SFP yone in line 2 is clearly considered as an inappropriate IfSS-PF.

However, in line 3, she corrects herself by adding a more detailed explanation of why it is difficult to understand with the masu form (omoimasu). This shift indicates her linguistic and cultural knowledge of the appropriate speech style in the contexts.

Regarding her IfSS-PF use in a casual setting in the pre-EdA OPI, Isabelle was not able to use IfSS-PFs in a role-play setting where she met a Japanese college student for the first time.

For the introduction, she used not only addressee honorific masu forms for all her sentence endings even after her interlocutor switched to plain forms, but she also used a humble form,

Izabere to mooshimasu (lit. I am (humbly) called Isabelle), which is, as discussed in Chapter

Three, extremely uncommon in conversation among university students. However, during the role-play in the post-EdA OPI, she demonstrated that she was able to use IfSS-PFs predominantly including a couple of self-corrections, which are shown below.

Excerpt 14 (Isabelle, Post-EdA, rated IM) 1 Isabelle → Chotto takaku naru to omoima, omou kedo, (I thin, I think it will become a little bit expensive) 2 Isabelle watashi okane aru kara, (I have money, so) 3 Isabelle tabun Yuki-chan ni agete, (Maybe, I can give it to you, Yuki-chan,) 4 Isabelle → Anoo, sono kuruma o naosu koto ga dekiru. (Well, you can repair that car.)

145

In the role-play, her task is to describe a minor car accident and offer a solution, and the excerpt above shows the parts where she offers a solution. In line 1, she says, “I think it will be a little bit expensive but,” by first saying omoima, which is an incomplete masu form, omoimasu ((I) think).

But before she finishes the form, she seems to realize that it is not the appropriate speech style in this context, and self-corrects to “omou kedo (I think but),” the plain form equivalent of the masu form. There was another occasion when she self-corrected her use of a masu form, kakarimasu to the plain form, kakaru during the role-play.

As mentioned in the beginning of this section, Isabelle’s gain in DtSS-PF seems to be limited. For example, all of her DtSS-PFs were soliloquy-like utterances with repetitions of the interviewer (i.e., “kankoo…”) or herself (e.g., “kunai…” and “tanoshikatta…”) in both the pre- and post-EdA OPIs, which is considered a somewhat reactive use of DtSS-PFs rather than using them actively for co-construction or providing background information. Isabelle’s data did not demonstrate much change in her DtSS-PFs use before and after EdA in Japan.

As discussed above, Isabelle managed to self-correct her own speech style (i.e., IfSS to formal speech style, and formal speech style to IfSS) in the post-EdA OPI, which was not observed during her pre-EdA OPI. This suggests that she can now self-monitor her own speech style in order to interact effectively in a given context. Although her DtSS-PF seems not to have changed much, her increased use of plain forms, non-use of inappropriate IfSS-PFs in the post-

EdA OPI, and use of IfSS-PFs appropriately in a casual setting upon her return also indicate her

EdA gain in style shifting.

146

5.3.3.2.4. Henry

Henry’s pre- and post-OPI ratings demonstrated his EdA gain. His OPI rating improved from Intermediate-High before EdA in Japan to Advanced-Low upon his return. Moving up from the Intermediate to the Advanced-level in Japanese OPI is not an easy task, even after spending a year in Japan. Therefore, his change at a major level in the OPI ratings suggests that Henry’s improvement in oral proficiency was significant.

Quantitative analysis of his style shifting also demonstrated his EdA gain. His masu form use was accounted for 94.8% of all his SMPs in pre-EdA OPI. Although he still used masu form predominantly (74.4%) upon his return, he increased his use of plain forms (14.5%) and others

(12%). Considering the Japanese L1 speakers in Masuda’s (2010) study used fewer masu forms

(52.5%) than the L2 speakers (78.6%), and they used other forms (37.2%) more than the L2 speakers (13.9%), Henry’s increased use of other forms as well as the decreased use of masu forms in SPMs suggests his style shifting became closer to that of Japanese L1 speakers.

Qualitative analysis revealed that most of his PF use was for IfSS-PFs, Henry used a

DtSS-PF only for co-constructing ideas in both pre- and post-EdA OPIs, as shown in Excerpt 15.

Excerpt 15 (Henry, Pre-EdA OPI, rated Intermediate-High) 1 Interviewer sakki hima na toki wa, taitee, hon o yomu to oshatte imashita kedo, sumimasen, nan to osshatte imashita kke, hon o yondari… (Well, you said, you usually read books when you have free time, but, hmm, excuse me, what did you say, you do things like reading books and…) 2 Henry >> anoo, ongaki* o kiitari suru. (Well, listening to music.) 3 Interviewer aa, naruhodo. (Oh, I see.) 4 Henry ee. (Yes.)

147

In the excerpt, the interviewer is about to ask questions regarding Henry’s hobbies. Prior to the topic, Henry was describing his tasks at a part-time job in masu forms. In line 1, the interviewer tries to recall Henry’s hobbies, which he mentioned earlier. She leaves her sentence incomplete, hon o yonda ri… (things like reading books and…) to signal Henry to provide the missing information. In line 2, Henry finishes the rest of the sentence with plain form ongaku o kiitari suru. Similar use of plain forms was often found in L1 speakers’ speech for co-construction, and it does not indicate IfSS (Cook, 2006; and Masuda, 2010). He then shifts back to the formal style ee (yes) in line 4 as a response to the interviewer’s utterance in line 3, which indicates he now puts emphasis on the formal relationship with the interviewer.

In the post-EdA OPIs, Henry also used a DtSS-PF for co-construction once but most of his plain form use was considered as IfSS. However, considering Henry still uses masu forms predominantly in SMPs and non-use of IfSS plain forms for speech acts called for in expressions of deference, such as making a request, suggests Henry’s cultural knowledge on the appropriate speech style. The excerpt below shows Henry’s use and non-use of IfSS-PFs.

Excerpt 16 (Henry, Post-EdA OPI, rated Advanced-Low) 1 Interviewer Jaa, kondo itte mimasu. Toohoku, demo watashi iku kikai ga amari nai node nee. (Well then, I will try going there next time. Tohoku, but I don’t have many opportunities to visit there so… you know.) Henry → itte kudasai. (Please go.) 2 Henry → Yuumee na, nihon no daiji na tokoro. (Popular, and important places in Japan, ) 3 Henry Miyajima, Amanohashidate, Matsushima. (Miyajima, Amanohashidate, Matsushima) 4 Interviewer Jaa, hoka ni doko ka itta koto arimasu ka. ikimashita, nihon de. (Well, other than that, did you visit anywhere? Did you visit anywhere in Japan?) 5 Henry Etto, ma, Sendai kara, Tokyoo, Yokohama, Takayama to Kanazawa, Kyoto

148

to Hiroshima to Nagasaki to, (Well, um, starting from Sendai, Tokyo, Yokohama, Takamatsu, and Kanazawa, and Kyooto, Hiroshima, Nagasaki, and) 6 Interviewer Sugoi desu ne. (Wow) 7 Henry → Beppu ni ikimashita. (I went to Beppu.)

In line 1, after Henry talks about his trip to Matsushima, the interviewer mentions that she would like to go there sometime, but she does not have opportunities to visit the Tohoku region, where

Matsushima is located. Henry then requests the interviewer to go there in line 2. He performs this speech act with the masu form. Because his speech act here is considered as making a request, which indicates high awareness of the addressee, it would have sounded rude if he had used the

IfSS request form (i.e., itte). In lines 2 and 3, he switches to the plain forms, where he is describing that Matsushima is a popular and important location in Japan, along with Miyajima and Amanohashidate. His use of plain forms in lines 2 and 3 does not sound rude, which is probably due to the fact that he is stating facts (Makino, 2002, p. 127). He again shifts his style back to the masu form in line 7 when he responds to the interviewer’s question, which is considered as an appropriate shift because answering a question often puts a speaker in the situation where the speaker has a high awareness of the addressee.

Examination of style shifting in pre- and post-EdA OPIs demonstrated that Henry was able to use DtSS-PFs appropriately and actively before and after EdA experience in Japan.

Because his IfSS-PFs in a casual setting was not available, we do not know if his appropriate use of IfSS-PFs resulted from EdA gain or not. Overall, Henry still maintained the masu forms predominantly, which is the appropriate speech style in the context while using PFs more often

149

in his speech style, which suggests his EdA gain (i.e., a communication strategy to avoid sounding too formal).

5.3.3.2.5. Emma

Unlike the previous four participants, Emma’s OPI ratings did not demonstrate her EdA gain. Both her pre- and post OPIs were rated Intermediate-High. This might be due to the nature of the OPI criteria that indicated she had relatively high oral proficiency prior to her EdA program in Japan, and in order for her to demonstrate EdA gain in an OPI rating, she might have to spend more time and effort than other participants with lower OPI ratings.

However, the quantitative analysis of her style shifting suggests her EdA gain. She produced the same amount of predicates in the pre- and post-EdA OPIs, but fewer SMPs in the post-EdA OPI, which indicates each of her SMP became longer and structurally more complex.

She also used more plain forms in the post-EdA OPI (19.4%) than the pre-EdA OPI (5%), which indicates that she shifted between the two styles more actively in the post-EdA OPI.

Findings from the qualitative analysis suggest gain in her style shifting skills, but it also suggests some regression like change. Regarding her gain, she used DtSS-PFs for a larger variety of functions in the post-EdA OPI than the pre-EdA OPI. For example, her DtSS-PF use was limited to soliloquy-like utterances for recalling (i.e., nan da kke) and to buy time by repeating the interviewer’s utterance (i.e., ki ni naru nyuusu). By contrast, she used DtSS-PFs to provide additional information and for self-correction (another function of soliloquy-like utterances) in the post-EdA OPI in addition to DtSS-PFs for recalling and buying time to hold the floor, as shown in Excerpt 17 below.

150

Excerpt 17 (Emma, Post-EdA OPI, rated Intermediate-High) 1 Interviewer A, jaa, omocha no koojoo de hataraite iru n desu ka. (Oh, so, he works at a toy factory?) 2 Emma → Iya, tada no ree desu keredo mo. (No, it was just an example,) 3 Emma soko wa, sono sooda no kikai (There, that soda machine,) 4 Emma nan ka, yoku resutoran de, sooda no kikai, (You know, there are often soda machines in restaurants) 5 Emma anoo, nomi-mono o katte, (Well, buy drinks and,) 6 Interviewer sooda no kikai (Soda machine), 7 Emma sono naka de bakku ga atte, sore ga (Inside the machine, there are bags, and they…) 8 Interviewer bakku. (Bags.) 9 Emma → toomee na bakku. (Clear bags) 10 Interviewer aa. (Ahhh) 11 Emma → sono, toomee na bakku no naka ni sooda ga haitte iru n desu ne. (That, it’s that soda is inside the clear bags.)

In the excerpt, Emma is describing her brother’s occupation. In lines 3 to 5, she is trying to explain sooda no kikai (soda machine) in a restaurant. In line 6, the interviewer repeats sooda no kikai, which indicates she is trying to follow what Emma has been saying. In line 7, Emma is trying to describe the bags inside the machine, but the interviewer interrupts her by saying bakku

(bag) in order to ask for more clarification. In line 9, she says toomee na bakku (clear bags) with the plain form (i.e., NPw/oC). This plain form use is very similar to what we will discuss in

Bobby’s DtSS-PF use (amefuto no purei) in the way that it also provides additional information to the previously mentioned word bakku (see Section 5.3.3.2.6), which indicates Emma is focusing on the content rather than the relationship with the interviewer. If this toomee no bakku

151

had been uttered in masu form (i.e., toomee no bakku desu), it would have sounded somewhat unnatural probably due to desu putting an emphasis on the relationship with the interviewer rather than the content.

As previously stated, there was one change that can be considered as regression of her speech style ability, instead of improvement. Emma used the DtSS-PF, nan da kke (what was it?) three times in the pre-EdA OPI when she was trying to recall information, which is considered culturally appropriate even in the formal interview setting. However, after returning from Japan, she used nan to/te iu (what is it called?) three times for recalling. Nan to/te iu is considered as an

IfSS interrogative in Japanese. As mentioned above, the same utterance was also used by L2 speakers in Okazaki’s (2010) study as well as Iwasaki’s study (2008). Because the OPI setting is considered as a formal interview setting, the multiple uses of the IfSS-PFs can be considered culturally inappropriate. The question is, why did Emma65, who was able to use the appropriate

DtSS-PF before her EdA experience in Japan, start using the culturally inappropriate IfSS-PF repeatedly after returning from EdA in Japan? There are two potential factors that may explain this change. The first one is negative L1 transfer. In English, what is it called? is often used as a soliloquy-like utterance for recalling information. In this case, the change is indeed considered as regression instead of improvement because she started relying on an English direct translation that does not fully function the same way as the Japanese one does. The second is that it is an incomplete utterance of nan te iu n da kke, which L1 speakers often use for recalling. If this is the case, she, as well as Frank, might have heard Japanese L1 speakers use the phrase often

65 Frank also used nan te iu in the post-EdA, but it was only once, and he did not use nan dakke in the pre-EdA OPI. Based on the findings, it is difficult to say that Frank’s use of the utterance in the post-EdA OPI is regression of his style shifting ability. 152

enough to start incorporating it in their speech, but they might have lacked the linguistic skills to produce the correct version of the phrase.

It is worth mentioning that Emma completed level four Japanese courses in the summer before her EdA program in Japan. During that time, she learned that the phrase nan da kke was used for recalling. Therefore, this might have accounted, at least partially, for her frequent use of nan da kke in the pre-EdA OPI which was conducted one month after her completion of the level four Japanese course.

Emma’s style shifting development also suggests that she learned not to sound too formal

(Okamoto, 1999) by using IfSS-PFs more but still maintained appropriate speech style overall.

She was able to use IfSS-PFs appropriately before and after her EdA experience in Japan, which suggests that she was able to communicate effectively in a casual setting even before this EdA experience. Her DtSS-PT development was somewhat complicated: She used DtSS-PFs more actively in the post-EdA OPI. However, her appropriate DtSS-PF (i.e., nan da kke) for soliloquy in the pre-EdA OPI seems to switch to the inappropriate IfSS-PF (i.e., nan to/te iu) after this EdA experience, which is considered as regression of her style-shifting ability.

5.3.3.2.6. Bobby

Similar to Emma, Bobby’s OPI ratings did not show improvement in his oral skill. In addition, Bobby was the only one who decreased the use of plain forms in the post-EdA OPI.

Qualitative analysis also suggests both gain and non-gain in Bobby’s speech style skill. Similar to Cathy and Isabelle, Bobby also used IfSS-PF inappropriately in the pre-EdA OPI. For example,

153

in the pre-EdA OPI, most of his plain forms were considered as IfSS, which includes, an IfSS-PF for the speech act of thanking, which is shown below.

Excerpt 18 (Bobby, Pre-EdA, rated Intermediate-Low) 1 Interviewer Jaa, intabyuu wa koko made desu. Arigatoo gozaimashita. (Well then, the interview is over. Thank you.) 2 Bobby → Arigatoo. (Thanks.)

Although Bobby has been using masu forms predominantly previously, when the interviewer thanks Bobby by using the masu form (i.e., arigatoo gozaimashita) to indicate the end of the interview in line 1, Bobby responds to it with a plain form equivalent, arigatoo in line 2. When a speaker is performing a speech act of thanking or requesting, he or she clearly has the interlocutor in mind. This case is considered a high addressee-awareness situation instead of low address-awareness situation (Maynard, 1993, p 178). Therefore, the plain form in line 2 is considered as an IfSS-PF. Using IfSS-PFs in such speech acts where masu forms were expected resulted in sounding rude. However, such inappropriate use of IfSS-PFs was not observed in his post-EdA OPI.

Regarding his use of DtSS-PFs, Bobby used DtSS-PFs only for soliloquy-like utterances before EdA experience in Japan, such as by saying wasurechatta (I forgot (how to say it in

Japanese)) to indicate he forgot a word in Japanese, by repeating the interviewer’s utterance sareru (have been done) with a falling intonation to indicate he was trying to recall the meaning of the word, and by saying muzukashii (difficult) to indicate his conviction to himself that it was difficult for him to explain the ongoing topic further. By contrast, in addition to soliloquy-like

154

utterances, Bobby used DtSS-PFs to provide additional information in the post-EdA OPI, as shown in the excerpt below.

Excerpt 19 (Bobby, Pre-EdA, rated Intermediate-Low) 1 Bobby a, amefuto, no, a, daigaku no, amefuto no sukyandaru ga arimashita. (um, American football, well, there was a scandal of college American football.) 2 Interviewer hee, sore wa donna sukyandaru? (Oh, what kind of scandal was it?) 3 Bobby anoo, a, a, amefuto no hito ga, anoo, hoka no hito ni takkuru shimashita. (Well, um, a, a person (a player) from American football, well, did tackle another person (player).) 4 Bobby → anoo…. Aa, muzukashii. (Well… umm, it’s difficult) 5 Bobby Aaa, purei, a, shitte, shitte imasu ka. (Umm, play, a, do you know play?) 6 Interviewer purei? (Play?) 7 Bobby → amefuto no purei. (Play in American football.)

In lines 1 and 3, he uses masu forms to begin explaining current news regarding a university football team in Japan, but he shifts to the plain form muzukashii (difficult) in line 4. This utterance is interpreted as a soliloquy-like utterance because it is produced with a softer voice than the previous utterances. In line 5, he shifts back to the masu form when he asks if the interviewer knows the meaning of purei (play). This shift from plain form in line 4 and masu form in line 5 indicates that he clearly understands the appropriate speech style to communicate with the interviewer. In line 6, the interviewer repeats the word “purei” with a rising intonation, which indicates that she does not know what he means by “purei”. In line 7, he shifts his speech style again to plain form amefuto no purei (play in American football) in order to provide additional information to the previously mentioned word purei. This NPw/oC plain form is

155

considered as DtSS-PF because it provides additional information to previously mentioned purei

(Makino, 2002), and it sounds more naturally produced with the DtSS-PF than it would have been with the masu form (i.e. amefuto no purei desu.).

Although he seems to use DtSS-PFs more actively in the post-EdA OPI, he also used

English (i.e., I forgot so many words…) in the post-EdA OPI to suggest that he forgot some words in Japanese, when he used the DtSS-PF (a, wasuretchatta) appropriately in the pre-EdA

OPI. This is considered regression in his speaking ability.

Regarding his performance in the role-play in an IfSS setting, Bobby predominantly used masu forms before EdA experience, but he used IfSS-PFs predominantly upon his return. This suggests that he gained both linguistic and cultural skills to shift his speech style based on the context.

Examination of Bobby’s style shifting in the pre- and post-EdA OPIs suggests that he was able to communicate appropriately in the formal setting before and after EdA in Japan. His overall use of the plain forms decreased, which might have contributed to sounding a little more formal than before. His appropriate use of IfSS-PFs in a casual setting in the post-EdA OPI suggests that he now acquired both linguistic skills and cultural knowledge to interact effectively in an informal setting. On the other hand, his gain in DtSS-PFs is not straightforward: He was using DtSS-PFs more actively such as for co-construction in the post-EdA OPI than the pre-OPIs in which he used DtSS-PFs only for soliloquy-like utterances. However, his use of English suggests the regression in his ability to use the DtSS-PF in Japanese that he once did.

156

5.3.3.2.7. Anna

Similar to Emma and Bobby, Anna’s pre- and post-EdA OPI ratings did not demonstrate improvement (i.e., Novice-High) in her oral skills. However, the quantitative analysis of her style shifting revealed that she increased the use of plain forms after her EdA experience in Japan.

Qualitative analysis also suggests both gain and non-gain in her style shifting skill, but compared to other participants discussed above, she seems to have the least gain in terms of style shifting skills. For example, in the pre-EdA OPI, all the plain forms were produced with some hesitation, such as produced with a rising intonation or followed by a pause. In addition, all but one plain form in Anna’s pre-EdA OPI were NPw/oC and one was VP “itta (go)” followed by a pause as in Excerpt 20.

Excerpt 20 (Anna, Pre-EdA OPI, rated Novice-High) 1 Interviewer A, soo desu ka. Doko ni itta n desu ka? (Oh, I see. Where did you go?) 2 Anna → Koobe to Oosaka to Kyooto ni itta desu. Kobe and Osaka and Kyoto to went COP (I went to Kobe, Osaka, and Kyoto.) 3 Anna → A, itta. went (Um, I went)

In line 1, the interviewer is asking where she went during her trip in Kansai by using a verbal plain form itta with the extended predicate doko ni itta n desu ka (where did you go?). In her response in line 2, she attempts to say, “I went to Kobe, Osaka, and Kyoto” in the masu form, however, she uses a verbal plain form itta with a copula desu, instead of ikimashita (went). It seems that she is just repeating what the interviewer said (i.e., itta n desu ka), but because she does not use the extended predicate (i.e. n desu), her utterance in line 2 is grammatically

157

incorrect. She realizes her mistake and tries to self-correct in line 3, but this time she just repeats the verbal plain form itta (go). This performance suggests that she wants to show her respect or sound formal in the context by using desu, a copula that has a similar social meaning with a masu form. Her self-correction also suggests that she understands desu cannot be used with a VP.

However, just repeating itta in line 3 suggests that she is not able to conjugate the verbal plain form into masu form, which indicates a lack of procedural knowledge.

Upon her return, she still seems to have difficulty in using plain forms, in general, considering that all of her plain forms were NPw/oC. This is also apparent from her performance in a role-play where the predominant use of IfSS-PFs was expected. During the role-play, she produced two IfSS-PFs, “donna nomi-mono ga ii?” (What kinds of drink are good?) and “donna juusu ga suki?” (What kinds of juice do you like), but her predominant speech style was masu forms, which is demonstrated in Excerpt 21 below.

Excerpt 21 (Anna, Post-EdA, rated Novice-High) 1 Anna A, moshi moshi. (Um, hello.) 2 Interviewer A, moshi moshi, Anna-chan? (Um, hello, is this Anna-chan?) 3 Anna → Hai. (Yes) 4 Interviewer Anoo, Yuki da kedo, (Well, this is Yuki). 5 Anna → A, soo desu ka. (Um, is that so?) 6 Interviewer Un, anoo, konshuu ne, anoo, paatii ga aru n da kedo, konai? (Uh hun, this week, there is a party, why don’t you come?) 7 Anna → etto, aa, nan jikan ga* arimasu ka. (well, umm, how long is it?) 8 Interviewer a, paatii? Etto ne, maa, ni, san jikan gurai kana. (Oh, the party? Well, hmmm, about two or three fours, maybe.) 9 Anna → anoo doko ni arimasu ka. (Well, where is it?)

158

10 Interviewer anoo, watashi no apaato de suru yo. (Well, at my apartment.) 11 Anna → a, soo desu ka…. uun. (Oh, is that so… well.)

In this role-play, Anna was supposed to talk to her friend, Yuki. The setting of the role-play was explained in English. However, Anna was unable to execute this performance with the appropriate speech style. All of her SMPs except line 1 are masu forms, which suggests she does not construct a relationship-building persona appropriate in this context. This demonstrates her non-gain of appropriate IfSS-PFs use even after her yearlong EdA experience in Japan.

Because Anna’s IfSS-PFs performance in an IfSS context was not available before her

EdA experience in Japan, it is difficult to say if she was able to use IfSS-PFs prior to her EdA experience. Furthermore, because her J-CAT score dropped, one might consider that her oral proficiency in terms of using the IfSS-PFs might have also regressed. However, considering that previous studies consistently report EdA students’ increased use of plain forms after EdA experiences, it might be the case that she was not able to use IfSS-PFs prior to her EdA experience, and she had not mastered it yet even after nine months of EdA experience in Japan.

Although Anna was unable to use IfSS-PFs predominantly in the role-play, the quantitative analysis of her utterances revealed that she increased the use of plain forms in her overall utterances. In addition, her use of IfSS-PFs in the role-play (e.g., “donna nomimono ga ii?” “donna juusu ga suki?”) suggests that she started using the IfSS-PFs more actively (i.e., applying what she knew to the new context) in the post-EdA OPI than the pre-EdA OPI where her plain form production was rather reactive (i.e., mere repetition of what she heard).

Considering that Novice-level learners tend to have difficulties in creating with language, Anna’s

159

inability to use IfSS-PFs in a casual setting may also have resulted from her oral language proficiency (i.e., Novice-High). Okamoto (1999) suggests that Japanese L1 speakers use style shifting to avoid sounding too formal or too informal (p. 63). From this perspective, the increased use of plain forms in Anna’s overall speech suggests EdA gain in her oral performance since she was able to sound more approachable which is an important skill for constructing an interpersonal relationship.

5.4. Discussion

The current chapter has focused on the findings on EdA gain measured through two proficiency tests (i.e., J-CAT and OPI) and examining their style shifting in OPIs. In this section,

I discuss the findings in relation to the two research questions.

5.4.1. What did the participants gain or not gain during EdA programs in Japan?

Overall findings suggest that there are individual differences in gain among the seven participants. In general, all participants demonstrated gain in vocabulary and reading, but gain in grammar seems to be more difficult. In addition, increased use of plain form was observed, but it was not consistent. Both J-CAT scores and OPI ratings suggest that yearlong undergraduate EdA experience in Japan is ideal for acquiring language skills to handle simple transactions and communication expected in daily life (Intermediate level). However, skills beyond such tasks, such as discussing topics that are beyond personal interests, are difficult to fully acquire even after yearlong EdA programs in Japan.

160

Let us first discuss gain and non-gain measured with J-CAT and OPI. Figure 9 presents radar charts of the seven participants’ J-CAT scores in four different skills (i.e., listening (L), vocabulary (V), grammar (G), and reading (R)), and the triangles represents the participants’ OPI ratings. A white area in each radar chart indicates the participant’s pre-EdA J-CAT score, and a darker area indicates their post-EdA J-CAT score. Texts next to radar charts represent their proficiency levels (PI = Pre-Intermediate, I = Intermediate, IH = Intermediate-High). Frank’s post-EdA J-CAT score was not available (N/A). The texts within an OPI chart indicate proficiency levels in the OPI rating (AH = Advanced-High, AM = Advanced-Mid, AL =

Advanced-Low, IH = Intermediate-High, IM = Intermediate-Mid, IL = Intermediate-Low, NH =

Novice-High, NM = Novice-Mid, NL = Novice-Low). Two colors on an OPI chart indicate improvement (i.e., Cathy, Frank, Henry, and Isabelle), whereas one color indicates non- improvement demonstrated by OPI ratings (i.e., Anna, Bobby, and Emma). The dark colors indicate improvement.

161

PI > IH I > NA I > IH I > IH

PI > PI PI > I I > IH

Figure 9. J-CAT score radar charts and OPI rating charts

In general, the higher a participant’s proficiency level is, the more balanced the radar

chart becomes. For example, participants with higher language proficiency in the pre-EdA J-

CAT scores (i.e., Frank, Isabelle, Henry, and Emma) demonstrated more balanced skills than the

lower-proficiency participants (Cathy, Bobby, and Anna). The radar charts for low-proficiency

students are much more skewed, especially their vocabulary scores. However, they become less

162

so in post-EdA J-CATs. As mentioned earlier, all participants demonstrated gain in vocabulary.

Among them, the participants whose pre-EdA scores were PI, demonstrated the largest gain among the four skills. This is probably due to their lower scores to begin with, and there was more room for gain than those with the higher proficiency level in pre-EdA J-CATs. Considering it has a greater impact on gains by the participants with the Pre-Intermediate level than the participants with Intermediate-level in the pre-EdA J-CATs, most of the vocabulary that the current participants needed to handle everyday tasks in Japan may have consisted of words or phrases that were at the Intermediate-level (i.e., being able to manage “daily communication”).

Similar to the J-CAT results, most participants were rated as around Intermediate-Low in the OPI tests before EdA programs in Japan. After their return, most of them were capable of communicating at the-Intermediate-Mid level or higher. This also confirms that participants can benefit from attending yearlong undergraduate EdA programs in skills to manage everyday spoken communication. However, it still seems difficult to acquire oral skills beyond the

Intermediate-level even after the yearlong EdA programs. In other words, participation in yearlong undergraduate EdA programs in Japan itself may not be enough to acquire the skills required at the Advanced level, such as narration, description, and discussion skills on topics beyond daily communication. We also have to consider that moving up to the next level in the

OPI rating scale becomes more difficult as the level goes up. This nature of the OPI criteria becomes unfair for EdA participants who left for the EdA program in Japan with higher proficiency, such as Emma in the current study. That being said, Henry demonstrated improvement in the post-EdA OPI even though his oral proficiency was as high as Emma’s.

163

Comparisons of the J-CAT scores and the OPI ratings suggest that skill gains measured by the both proficiency tests seem to be related more or less. For example, participants with

Intermediate-Mid or higher tend to have more balanced radar charts than those with

Intermediate-Low or lower both pre- and post-EdA. In addition, participants who demonstrated improvement in OPI ratings tend to increase their scores in grammar in J-CAT scores more than those who did not. Participants whose gain was not demonstrated by OPI ratings (i.e., Anna,

Bobby, and Emma) had higher scores in grammar in the pre-EdA J-CAT scores than the participants who demonstrated gain in OPI ratings (i.e., Cathy, Frank, Isabelle, and Henry). This finding somewhat contradicts what Brecht et al. (1995) found in their research. Brecht et al.

(1995), whose study involved 658 EdA students, find that knowledge on grammar and reading is a predictor of oral proficiency gain. The difference may have stemmed from the different proficiency tests used in the study done by Brecht et al. and the current study as we discussed the reliability of J-CAT above (i.e., Anna’s case).

EdA gain as seen in the use of style-shifting strategies also suggests there were individual differences among the seven participants. Unlike the two proficiency tests, style-shifting skills do not necessary develop along with the skills measured with J-CAT and/or OPI tests.

Table 12. Summary of EdA gain demonstrated by examining style shifting Name Style Shifting Increase in App. IfSSPF Inap. IfSSPF DtSSPF Increase in fragments Total PF Pre > Post Pre > Post Pre > Post Cathy Yes [+4%] Yes > Yes Yes > No Yes > Yes No [-1%] Frank Yes [+12%] Yes > Yes No > No Yes > Yes Yes [+6%] Isabelle Yes [+6%] No > Yes Yes > No No > No Yes [+9%] Henry Yes [13%] N/A > Yes No > No Yes > Yes Yes [+10%] Emma Yes [13%] Yes > Yes No > No Yes > Yes/? Yes [+6.6%] Bobby No [-4%] No > Yes Yes > No Yes > Yes/No Yes [+5%] Anna Yes [+4%] N/A > No No > No No > No Yes [+6%]

164

Because the format of OPI tests assumes a formal interview setting (and both interviewer and interviewee do not know each other), predominant use of masu forms was expected. Quantitative analysis of the participants’ style shifting revealed that all students used masu-forms predominantly in both the pre- and post-EdA OPIs, which suggests that the participants’ cultural knowledge on appropriate speech style for the given setting before and after their EdA experiences. In addition, the analysis revealed that all but one participant increased the use of plain forms upon their return, which also supports a strong relationship between the use of plain forms and experience living in Japan found in the previous research (Iwasaki 2008, 2010;

Marriot, 1995; and Masuda, 2010).

Qualitative analysis also revealed that the participants used DtSS-PFs for more functions after their EdA experiences. Most of the DtSS-PFs were soliloquy-like utterances (i.e., recalling and exclamatory), but they also started using DtSS-PFs for things such as co-constructing ideas both syntactically and semantically and providing background information (Masuda, 2010). In addition, considering that there was no inappropriate use of IfSS-PFs in high-awareness of the addressee, yearlong EdA experience seems to provide learning opportunities for appropriate use of both IfSS-PFs and DtSS-PFs.

However, although all participants increased the use of plain forms, acquiring natural use of both IfSS- and DtSS-PFs entirely still seems difficult even after yearlong EdA programs in

Japan. For example, some participants’ changes in use of plain forms seemed haphazard (i.e.,

Emma’s use of nan te/to iu and Bobby’s use of the English sentence, which he did not use in the pre-EdA). In addition, all participants, except Henry, used a rising intonation at the end of

NPw/cC to express uncertainty in the formal interview settings, which made their utterances as

165

informal interrogative utterances (i.e., moto-arubaito?). Japanese L1 speakers often use sentence final particles (SFP) (datta) kana or (datta) kke with plain forms to express uncertainty (Okazaki,

2015). Although it highly depends on the context, too much use of such IfSS-PFs can result in sounding too informal. On the other hand, in a casual role-play setting, some words are considered too formal, such as use of ee/hai instead of un, node instead of kara, and watashi instead of boku/ore by a male participant.66 This gap in sounding too informal in a formal discourse setting or too formal in an informal discourse setting may create an awkward atmosphere for a conversation partner.

To develop similar skills—when and how they become aware of speech styles choices as a way of indexically defining a context, a relationship, and so on, L1 speakers of Japanese start very early and take long time to learn such skills. In addition, such skills are very local, every step of the way. This notion of taking extensive time to master such skills suggest that such skills need to be addressed prior to EdA programs in order for EdA participants to fully take advantage of EdA learning environment.

The findings from the qualitative analysis of Anna’s OPI data suggest her improvement in PF use. Given that her OPI and J-CAT scores did not demonstrate improvement in language proficiency improvement, there is more to “gain” in language during EdA in Japan than just the

OPI ratings or J-CAT scores.

66 Choice of first personal pronoun in informal speech is not straightforward because it is deeply influenced by personal preference. However, this particular participant mentioned that he wanted to use boku in IfSS conversation instead of watashi, but he still used watashi in actual performance. 166

5.4.2. Are there any differences in EdA gain between the participants in the EdA training

group and the non-EdA training groups? If so, how?

The pre-departure training alone seems unable to predict gain in style-shifting strategy development as well as gain measured by the J-CAT and the OPI. Tables 13 and 14 reorganized the same content from Tables 11 and 12 based on the participants in the two groups. The participants who attended the EdA training are categorized as Training group and Non-training group (Non-T) in the tables.

Table 13. J-CAT scores categorized by Training and Non-training groups

Name J-CAT T PL L V G R

Group Pre > Post Pre > Post Pre > Post Pre > Post Pre > Post Pre > Post

Bobby 108 > 151 PI > I 29 > 49 11 > 34 41>26 27 > 42 Frank 153 > N/A I > N/A 45 > N/A 40 > N/A 26 > N/A 42 > N/A Emma 158 > 211 I > IH 48 > 58 34 > 48 45 > 47 31 > 58 Training Isabelle 160 > 219 I > IH 44 > 59 41 > 54 36 > 45 39 > 61

Anna 144 > 126 PI > PI 29 >19 15 > 34 70>40 30 > 33 T - Cathy 127 > 222 PI > IH 52 > 71 15 > 49 26 > 45 34 > 57

Non Henry 180 > 217 I > IH 49 > 64 45 > 48 34 > 51 52 > 54

Although Table 13 indicates the EdA training may have had a positive impact on the participants’

EdA gain, Cathy and Henry who were in the non-EdA training group also demonstrated gain.

Results from the OPI ratings and examining style shifting in Table 14 seem more haphazard.

Two participants (i.e. Frank and Isabelle) from the EdA training group demonstrated EdA gain in

OPI ratings, but the other two (i.e. Emma and Bobby) did not. Regarding the non-EdA training group, Cathy and Henry again demonstrated gain, but not Anna.

167

Table 14. OPI ratings and style shift strategy development categorized by Training and Non- training groups Name OPI Style Shifting Pre > Post Increase in App. IfSS- Inap. IfSS- DtSS-PF Increase in Total PF PF PF fragments Group Pre > Post Pre > Post Pre > Post Frank IM > IH Yes [+12%] Yes > Yes No > No Yes > Yes Yes [+6%]

Isabelle IL > IM Yes [+6%] No > Yes Yes > No No > No Yes [+9%] Bobby IL > IL No [-4%] No > Yes Yes > No Yes > Yes [+5%] Yes/No

Training Emma IH > IH Yes [13%] Yes > Yes No > No Yes > Yes [+6.6%] Yes/No

Anna NH > NH Yes [+4%] N/A > No No > No No > No Yes [+6%] T -

Cathy IL > IM Yes [+4%] Yes > Yes Yes > No Yes > Yes No [-1%] Henry IH > AH Yes [13%] N/A > Yes No > No Yes > Yes Yes [+10%] Non

Table 15 reorganized participants into two groups, this time, high-gainers and low-gainers.

Participants who demonstrated gain with all three measurements or most of the measurements were considered as high-gainers (Cathy, Frank, Isabelle, and Henry,), while those whose gains were limited considered as low-gainers (i.e., Emma, Bobby, and Anna).

Table 15. High-gainers and Low-gainers High-gainers Low-gainers

Cathy Emma Frank Bobby Isabelle Anna Name Henry

The comparison of the results shown in Tables 13, 14, and 15 suggests that both groups included those with high-gain and those with low-gain, which further suggests that the EdA training may not have a great impact on the participants’ EdA gain. This raises new questions:

Why did the EdA training, which was carefully designed based on empirical research, not have much impact on the participants’ EdA gain, especially for Bobby and Emma? Why are there

168

individual differences among the participants in both EdA groups? If there are elements other than the EdA training that influenced the participants’ EdA gain, what are they? As we will discuss in the next chapter, investigation of the process of EdA learning suggested that the design of the EdA programs have had the greatest impact on the seven participants’ EdA gain in Japan.

In order to understand the complexities of EdA gain elucidated in this chapter, Chapter Six focuses on the process of EdA learning by examining the seven EdA students’ social network construction.

169

Chapter 6: Understanding the Processes of Education Abroad Learning in Japan

6.1. Introduction

Chapter Five has discussed the seven participants’ EdA gain (i.e. the outcomes of EdA learning) during yearlong undergraduate EdA programs in Japan. The findings revealed that all participants demonstrated EdA gain with at least one assessment, but there were individual differences in the degrees of EdA gain. In addition, there seem to be no differences between the participants in the EdA training group and the non-EdA training group, which also raised new questions discussed at the end of Chapter Five.

As mentioned in Chapter One, some scholars argue that EdA research should shift its focus from the outcome to the process of EdA learning in order to understand what is really going on during EdA experiences (Wilkinson, 1998; and Wang, 2010). Considering that developing strong social networks with locals has been reported as a strong predictor of EdA gain (Baker-Smemoe et al., 2014; and Isabelli-Garcia, 2006), investigating the construction of social networks as well as participants’ EdA experiences may provide helpful insights on the process of EdA gain. Therefore, the current chapter focuses on the seven participants’ social network development and their EdA experiences in relation to their EdA gain. Three research questions guided the discussion in the current chapter:

170

1. What are the commonalities and differences in seven U.S. students’ social

networks during EdA programs in Japan?

2. In what ways did social networking influence the seven U.S. students’ EdA gain

discussed in Chapter Five?

3. Are there any differences in EdA social networks between the participants in the

EdA training group and non-EdA training group? If so, what are they?

The following sections present findings regarding the current participants’ EdA experiences and social networks during EdA in Japan. As discussed in Chapter Four, the concept of individual network of practice (INoP) developed by Zappa-Hollman and Duff (2015) is used to analyze each participant’s social networks (see Section 4.7.2. in Chapter Four for detailed discussion of the INoP analysis).

6.2. Findings

Analysis of the seven participants’ social networks suggests two significant findings.

First, social networks seemed to have influenced the seven participants’ EdA gain discussed in

Chapter Five: The more complex and the more Japanese people involved, the greater the EdA gain, and EdA students who constructed social ties with local Japanese people through repeated negotiation demonstrated improvement in higher EdA gain. We refer to such social ties as negotiated Third Space, which will be discussed in more detail later. Second, the design (or structure) of their EdA program seemed to have the greatest impact on constructing negotiated

Third Space.

171

The following sections present detailed findings of the seven EdA students’ social networks in the order of Cathy, Frank, Isabelle, Henry, Emma, Bobby, and Anna. As mentioned in Chapter Five, the first four participants (Cathy, Frank, Isabelle, and Henry) are high-gainers and constructed social networks that are more complex and included more Japanese locals than the other three (Emma, Bobby, and Anna) who are low-gainers. For Cathy, Frank, and Isabelle, the design of the EdA program seems to have facilitated significantly their social network construction with Japanese people. In the following sections, the findings regarding background information (i.e. goals and motivation of Japanese study and EdA programs) of each participant are also discussed.

6.2.1. Cathy (High-gainer)

Cathy is one of the most successful participants in the current study regarding EdA gain.

She also constructed more complex social networks in terms of numbers of Japanese nodes (i.e. social ties with Japanese people) as well as clusters (i.e., social groups). As we discuss below, her high motivation seems to play an important role in constructing her social network, but her environment (i.e., EdA program, school club, and dorm) also enabled her to maintain her high motivation throughout her EdA experience and establish such complex social ties.

Cathy double majored in computer science and Japanese. She completed level two

Japanese courses at her home institution. This yearlong EdA experience was her first time in

Japan. She was first interested in studying Japanese because of the cultural differences between

Japan and the US. She described her primary reason for participating in a yearlong EdA program in Japan was to become fluent in Japanese, and she was “really looking forward to

172

actual studying abroad experience and in being in the culture with growth in the language that’s completely different from America.” She mentioned that her goals for her EdA experience in

Japan were becoming fluent in Japanese, making a lot of friends, and integrating as much as possible into the Japanese society. She also held realistic expectations regarding the last goal that as a foreigner, “it’s kinda impossible [to integrate fully in the society]” and she would be

“always treated like a foreigner.” She came to that conclusion by watching many YouTube videos on Japanese society and culture. Cathy utilized social media for learning about Japanese society as well as Japanese people’s expectations toward foreigners (especially non-Asian people), which helped her to develop realistic expectations as well as goals for her EdA learning in Japan. She also mentioned that she would use her Japanese roommates as examples and learned from them what she should and should not do because Japanese people did not always explicitly express their feelings.

Cathy attended a private liberal arts college (i.e., University CFI) in the Kanto region and stayed at a university dorm where both Japanese and international students shared a unit. Cathy took intensive Japanese language courses at her host university, which were held from 8:50 in the morning to 12:40 in the afternoon every day. According to Cathy, the program was harder and had much more homework than her home university, which she described as surprising because she thought the Japanese program at her home university was hard. During her EdA experience, Cathy also actively participated in extracurricular activities. She attended four circles, koto, shamisen, hiking, and softball clubs, which she researched on the university website prior to her EdA experience. However, she withdrew from the shamisen club because the practice was seven hours long at the instructor’s house on Saturdays, and she would get extremely tired due to

173

dealing with the unfamiliar instrument as well as the seven hours of a Japanese-only environment, which made her feel that her “brain just couldn’t take it.” But she expressed that she received the most practice in conversational Japanese from interacting with people in these clubs.

She described her EdA experience as distance running in the way that “once you stop, it is difficult to get back to the track,” so she tried her best not to stop but to participate as much as possible in classes and extracurricular activities. She seemed to be the type of student who can set up realistic goals and work hard toward those goals. In addition to her high motivation to maximize her EdA experience, it seems the communities she was involved in and the social network she constructed helped her keep running. For example, when I asked if there were any occasions when she wanted to “stop running” during her EdA experience in Japan, she said there were many occasions that she wanted to quit the koto club. However the reason why she stuck with it was she became close to one of the club members, Toshi, who was an advanced koto player. He encouraged her to continue playing koto and at the end of the semester, Toshi and

Cathy did a duet at a performance event. They practiced koto together for the event outside of club time. Cathy had the strongest tie with Toshi among other local Japanese students. She also received help from other club members. For example, she described that there were times she did not understand the situation but “there are always someone who can help me in English” in the club. She also learned that these people who always helped her were bucho (a leader of the club).

Cathy first thought that Hinata and Kana in the Koto club were helping her often because they might have wanted to practice speaking English with her, but at one point, she realized that they were responsible for her because of their role in the club, not because they wanted to speak

English with her.

174

Language use data also suggests that having such support from the club members helped

Cathy’s participation in the club. In a log in February, Cathy wrote, “[m]ost of the stuff is in

Japanese and I can communicate with everyone relatively well. When I need it, I can ask in

English so there's no pressure.” The clubs she attended during EdA programs seemed to be organized in the way that all club members including Cathy who need language assistance could participate in the club activity fully. As we will see in Isabelle and Bobby’s cases, such structure is crucial for EdA students to continue their participation in the club. For Cathy, having support from club members enabled her to continuously participate in these clubs.

Figure 10 shows her INoP during EdA in Japan, which consists of one main cluster,

University, and seven sub-clusters, namely, Koto club, Softball club, Korean class, Dorm,

Roommates, and Classmates. Three of the clusters, the koto, hiking, and softball clubs, were the university circles she participated in, and two clusters are class related, the Korean class and

Classmates (of the intensive Japanese language course) cluster. All twenty nodes, except one

(i.e., Kim, a Korean teacher), were students at Cathy’s university, and fourteen out of the twenty nodes are Japanese. Regarding social ties, she developed the strongest ties with Isabelle and Lala who were her classmates. They also spent time together outside of class. She explained that because she spent most of her time at school in the intensive Japanese classes as well as the fact that they spoke English as their L1, it was natural for her to develop strong ties with her classmates. Even though Cathy seems to have developed strong ties with locals (some clusters only consisted of social ties with local Japanese students), she described her Japanese as an obstacle, and she felt at ease with her American friends who shared the same native language.

175

Cathy Kozue Kae Yumi Mina Hinata Mami Toshi Koto Room Club Chiki Roki mates Hiking Club Yuki Dorm Ami Koota Cady University Lake So@ball Club Frank

Chie Korean Classmates Class Shun Cathy

Kim Lala Isabelle

Figure 10. Cathy's INoP

All sub-clusters in Cathy’s INoP were within the cluster of University, which consists of several different sub-clusters, Koto Club, Softball Club, Hiking Club, Korean Class, Dorm,

Roommates and Classmates. As shown in her INoP above, nodes in these clusters, koto club, hiking club, and softball club consisted of only local Japanese students. Although Cathy confessed that she did not have much time to socialize with students in her dorm due to her busy schedule with extracurricular activities and classes, she had developed social ties with Japanese students in her dorm thanks to the structure of the dorm, which we discuss in more detail in

Isabelle’s section below.

176

In addition to her realistic expectations and strategies to maximize her EdA experience prior to the experience, she also received both emotive and instrumental support from both

Japanese and non-Japanese people, which helped her to construct an extensive social network with various people within the university setting.

6.2.2. Frank (High-gainer)

Similar to Cathy, both Frank’s EdA gain and social network development were significant. It seems that the design of the EdA program facilitated Frank’s EdA gain as well as social network development, but his higher language proficiency prior to the EdA program also seemed to play an important role to constructing some of his social ties.

Frank started studying Japanese by himself when he was a high school student because he was interested in Japanese pop-culture. At his home university, he first majored in Mathematics and minored in Japanese, but eventually switched his major to Japanese because he found it more interesting.

Similar to Isabelle, Frank also completed the level three Japanese language courses at his home institution and participated in a short-term faculty-led summer program in the Kansai region prior to his yearlong EdA program in Japan. He described his primary reasons for the yearlong EdA experience as learning more Japanese and getting better at it. He wanted to become fluent in the language as much as possible so that he could use the skills for a future career. In terms of his goals for the EdA program, he mentioned that making friends, getting to know people, and learning how to live in Japan. He wanted to learn “how to be a Japanese

177

citizen” because he planned to go back to Japan to get jobs related to Japanese language and culture in the Navy.

As mentioned above, Frank attended the same private liberal art college and took the same intensive Japanese language courses with Cathy and Isabelle. Frank participated in a ballroom dancing club at the university because he always wanted to try it but had not had a chance prior to his EdA experience. By attending the dance club at the university, he expanded his social network outside of school. He met a professor, Jacey, who was a coach at the ballroom dancing club, and she recommended another ballroom dancing club outside of the university.

Frank started going to practice at the external dance club once a week, which he enjoyed very much. In the current study, social ties established outside of the host university is called external social ties. Prior oral language proficiency (i.e., Intermediate-Mid or higher in the pre-EdA OPI tests) seemed to play an important role to establish such ties as we will discuss in Henry and

Emma’s sections as well as in 6.3.1.

Before the third trimester, I talked with Frank online as a part of the concurrent-EdA training. He told me that he struggled with Japanese students who wanted to speak English with him. We also discussed that in order for him to learn various types of Japanese, participating in volunteer activities in the local community would be good. In the third semester, he volunteered at a children’s institution as a part of a university course, where he played with children who lived in the institution. His primary responsibility at the children’s institution was watching the children and playing with them, but Frank wished to develop good relationships with staff who worked there, especially younger staff members who were around the same age as Frank.

However, he did not have much interaction with the staff members, because they always stayed

178

in the staff room to which Frank did not have access, so he “[could not] become friends” with them. Frank’s unsuccessful experience in constructing social ties with locals suggests that L1 speakers’ willingness to interact with EdA students is important to construct intercultural relationships.

Figure 11 represents Frank’s INoP during EdA in Japan, which consists of two main clusters, University and External dance circle, and four sub clusters, University dance circle,

Dorm friends, Roommate, and Pre-existed friend within the University cluster. Twelve out of twenty nodes are Japanese. Five of the Japanese nodes were students at Frank’s university, and seven of the Japanese nodes belonged to External dance circle. Frank also developed the strongest ties with the international students who took the same intensive language courses with him including Cathy and Isabelle. In addition, Frank mentioned that having similar personalities, interests, and English as the L1 were the reasons why he developed strong ties with Lacy,

Isabelle, and Jacey.

179

Kita Frank Mina Davis Yuuki Nao Jadon Room mates Yumi University Kale Dance Circle Dorm Dain Friends Jacey Lacy Cathy Yori University Emi Class mates Aki External Dance Circle Frank Pre-excising Friend Oka Isabelle Tae Kiku Yasu

Figure 11. Frank's INoP

However, English seems to be one of the reasons for not developing strong ties with his roommate, Yuuki. According to Frank, the reason that he did not talk so much with Yuuki was because he was consistently unsure of whether to speak English or Japanese to Yuuki, which was described below.

I thought I should speak Japanese so I could practice, and he was considerably more comfortable with Japanese, but he also wanted to practice English. Also, it was often difficult and awkward for me to try to speak Japanese effectively to him.

Cathy and Isabelle, who attended the same university, also expressed similar concerns. The university where the three participants attended is known for international and global education, and Japanese students’ English skills there tend to be higher than at other universities in Japan.

180

The school also accepts many Japanese students who have grown up abroad. English seems to facilitate a relationship in some situations but become a hindrance in other situations. Japanese people who can speak English play an important role for EdA students to participate in the local community (e.g., Cathy and Isabelle’s different experiences of attending school clubs). However,

Japanese people’s abilities to speak English could also somewhat negatively affect the construction of strong social ties with them. Frank, as well as the other participants, struggled between the feelings of wanting to practice Japanese with local people and knowing that these

Japanese people also wanted to practice English. Although he was successful in establishing a social tie with Yuuki, this conflict also had a negative impact on Frank’s relationship with his roommate.

6.2.3. Isabelle (High-gainer)

Isabelle attended the same EdA program with Cathy and Frank, and she also benefited from it in terms of constructing social ties with local students. However, she struggled with participating in school club activities unlike Cathy and Frank, which is discussed in detail below.

Isabelle started her Japanese study primary because she had a family in Japan and wanted to communicate with them in Japanese. She visited Japan for two weeks when she was a high school student to see her Japanese family. Isabelle completed the level three Japanese courses at her home institution prior to her yearlong EdA experience in Japan. Her primary goal for this

EdA experience was to get better at the Japanese language by being surrounded by Japanese.

During her EdA program in Japan, Isabelle also lived in a university dorm with two roommates, a Japanese student, and an international student. School dorms where Isabelle, Cathy,

181

and Frank lived had mandatory monthly dorm meetings. During the meetings, both English and

Japanese were used. All participants had to participate fully by stating their opinions and casting their votes. Isabelle spent a great amount of time with other local Japanese students who lived in the same dorm by attending the meetings as well as hanging out with them. Through interacting with local students at the dorm, she learned a lot of colloquial Japanese that young people in

Japan used. However, Isabelle also confessed that many Japanese students at her university already spoke English well and wanted to use English with her, so she did not use as much

Japanese as she expected before her EdA experience in Japan.

Isabelle attended the intensive Japanese language course with Cathy and Frank. Similar to Cathy, Isabelle also mentioned that the four-hour of Japanese classes every day made her really tired. In the beginning of the EdA program, Isabelle joined a music band club because she liked playing the guitar. However, after attending a couple of music events where the club members volunteered to set up for the events, she stopped participating in the club due to language difficulty. According to Isabelle, Japanese was exclusively used during the club meetings as well as for instructions regarding tasks at music events, and she had difficulty understanding what was going on most of the time. Because the recruitment flyer of the band club mentioned that international students were welcomed, Isabelle assumed that she could get some English help from the club members in order her to participate in the club activity.

However, even after she asked the club members to provide English translations on the club group chat, no one did, which made it difficult for her to participate in the club events fully.

Eventually, Isabelle stopped going to the club events.

182

Figure 12 presents Isabelle’s INoP she constructed during EdA in Japan, which consists of one main cluster, University and four sub-clusters, Classmates, Dorm, Same floor of dorm, and Roommates. All of the sixteen nodes were students at Isabelle’s university. Eight out of the sixteen nodes are Japanese.

Isabelle Satomi Sayuri Emi Miki Hiroko Room mates Naoko Same Floor of Dorm Dorm Frank Cindy Hime Cathy Yuko Siki Classmates University Kim Mike May Lala Isabelle

Figure 12. Isabelle's INoP

Similar to most of the participants in the current study, Isabelle constructed strong social ties with other international students who attended the same language classes because they spent a lot of time together in the classes. She also mentioned that having similar personalities helped to develop strong ties with Cathy, Hiroko, and Satomi. Like other participants discussed below, language also played an important role in constructing social ties in Isabelle’s INoP. For example,

Isabelle moved into the dorm around the same time as Emi, Hiro, and Yuko, and the three of

183

them became good friends but Isabelle did not. She described that “they are all fluent in Japanese, and when I first got there, I could have small conversations, but I was not as proficient as I was in the end.” By the time she became comfortable using Japanese, Isabelle already made other friends with whom she wanted to spend time, so she did not spend as much time with these three

Japanese students.

Although Isabelle did not have a positive experience at the music club, unlike Cathy’s experience at her school clubs, Isabelle’s living environment (i.e., mandatory monthly dorm meetings and the culture at the dorm) provided her to interactional opportunities with locals, which helped her to construct strong social ties with Japanese students in the dorm.

6.2.4. Henry (High-gainer)

Similar to the three participants discussed above, Henry also demonstrated significant

EdA gain, especially in OPI rating in which he was the only one who improved at a major level

(from Intermediate-High to Advanced-Low level). Unlike the previous three students, Henry attended a different EdA program at a national university, which seemed to be less structured than the University CFI, but his higher language proficiency seemed to have compensated for such lack, and even allowed him to develop social ties both in and outside of the EdA program that may have resulted in his EdA gain.

Similar to most of the participants in the current study, Henry started his Japanese study because he was attracted by the uniqueness of Japanese culture and language, which he did not find in Spanish, which he studied for two years.

184

Prior to this EdA program, Henry completed level five Japanese courses at his home institution. This yearlong EdA program in Japan was his first time visiting and living abroad. He decided to study in Japan because he wanted to work at a Japanese company, especially at a

Japanese travel company, in the future. He thought that this EdA experience on his resume would be good for him. He also considered that studying abroad in Japan for a year while he was a university student would be ideal timing because he did not have to worry about finding a job.

In addition to getting better at the language, including being able to talk about various different subjects, his goals for participating in the EdA program in Japan were to experience the Japanese lifestyle and to construct good social networks. He wanted to construct good social networks in

Japan so that he would have support if he decides to work there in the future.

Henry attended University H, which is a national university in the Tohoku region. Henry lived in a university dorm during his EdA experience in Japan, but unlike the dorms at

University CFI, Henry described that there was not much interaction in his dorm. According to

Henry, this university has several EdA programs, and he enrolled in a program for international students from European and North American countries.

In addition to Japanese courses for international students, Henry also took general education courses with local students because his Japanese was high enough to take these classes.

Henry mentioned that he struggled to keep up with the assignments from these general education courses, which was because his host university was one of top universities in Japan. For example, one of the classes was a research class on classical Japanese poetry and wood prints, in which he did a research project on the history of locations and cultural background of the poems in

Japanese. According to his language use log in December, his English language use during the

185

week (seven days) was approximately 4.7 hours, and most of them were interactive language use, such as talking with his family on Skype and international friends at the university. By contrast, he used Japanese approximately 18.7 hours during the same week, and most of them were non- interactive and class-related, such as attending classes, doing homework and self-study of

Japanese. Although he wanted to attend some clubs, such as Badminton and Igo clubs, he did not have time because he spent most of his time working on class assignments during the semesters.

Because general education courses Henry took were not language courses, the primary focuses of the courses were not learning the Japanese language. By taking such courses, Henry started using Japanese as a tool to learn other contents rather than learning Japanese as a primary purpose of taking courses, which was notable difference with other participants who took language or content courses for international students, including courses for both Japanese and international students (these courses were not general education courses, and were taught in

Japanese and English). Learning Japanese was still a primary focus of the courses rather than a tool to learn other subjects for them. Such extensive use (i.e., not learning) of academic language probably resulted in Henry’s EdA gain demonstrated with the two language proficiency levels discussed in Chapter Five.

Outside of school, Henry spent his breaks traveling all over Japan and visiting historical places. While traveling, he visited local restaurants and stores with other foreign tourists whom he met there. His high Japanese proficiency and knowledge of Japanese places and culture enabled him to communicate smoothly with locals and introduce Japanese culture to the foreign tourists, which he enjoyed a lot and described as “the best experiences.” He also explored his

186

local area often and found a bar in the downtown area where he returned often during his stay and developed a friendship with the bartender.

Figure 13 presents Henry’s INoP, which consists of two main clusters, University and

Bartender, and six sub-clusters, Western program, Faculty, Student/classmates, Student party,

Band members and Tutor, within the University cluster. The Faculty and Student/classmates belonged to Western program. The Band member cluster connected to Abby who belonged to

Classmate/student cluster.

Four out of fifteen nodes are Japanese, within which there are various social statuses.

Henry was one of three participants in the current study who developed social ties outside of the university. In Henry’s case, he went to the same local bar throughout the year and became friends with the bartender. Henry described Peco as “a very likable personality, is easy to talk to, and doesn't mind talking to foreigners.”

187

Tani Henry Kea Tracy May Gabi Cab Faculty Students/ Classmates Dacia Western Program Peco Abby Alix University Band Bartender Henry Members Student Party Tutor Lucy Yo

Ohta Meda Chris

Figure 13. Henry's INoP

Henry also developed a strong tie with a professor, Cab, whom Henry chose to call by his first name. Cab and Henry shared similar personalities as well as personal backgrounds.

Spending plenty of time to get to know each other by going out for drinks also contributed to developing a strong tie with Cab. Henry also listed a Japanese professor, Tani, who was another professor in the EdA program Henry participated. Henry took some of his classes throughout the year and also went drinking with the professor along with other students who attended his class.

However, he described his relationship with the professor as follows, “I couldn't really connect with him beyond an academic level. He was interesting to listen to because he's very knowledgeable, but it felt like I couldn't communicate with him in the same capacity and depth.”

188

He only used Japanese with Tani, but Tani used “a much more elevated level of Japanese” than

Henry was familiar with, which made it difficult to communicate freely and comfortably with

Tani. Although Cab and Tani were both faculty members of the EdA program Henry attended, the strengths of the social ties he developed with the two professors were different. It was clear that the differences resulted from the differences in the primary language used as a primary means of communication as well as personality differences and similarities.

Within the University cluster, Henry met Ohta through a tutoring program offered at the international student center and they met once a week to practice conversational Japanese.

However, the personal relationship between Henry and Ohta did not go beyond that of the tutor and tutee. Henry met Maeda through a party and went drinking with him a couple of times.

Henry described Maeda as “a very busy medical student and didn't have much time to hang out,” so even though they got along well when they could meet up, they did not have enough time together to get closer. Henry mentioned that it was difficult to make Japanese friends at the university because most of the local students were really busy, and they have their own circles already.

6.2.5. Emma (Low-gainer)

Unlike the previous four participants (Cathy, Frank, Isabelle, and Henry), Emma’s EdA gain was categorized lower-gainer in Chapter Five, which means some assessments (i.e., her J-

CAT result and some aspect of style-shifting development) demonstrated her EdA gain, but others did not (i.e., OPI rating and some use of DtSS-PF). Findings in her social network development also suggest her social network development was limited compared to the previous

189

four participants in terms of quantity (numbers of Japanese nodes and clusters) as well as quality

(i.e., social status of Japanese nodes).

Emma started learning Japanese by herself when she was a high school student because she was interested in the culture and the language that was “very different from the United

States.” She also studied in the northern part of Japan as a high school exchange student for nine months. However, she regretted that she did not use Japanese much during her first EdA program in Japan, and she mostly spent time with other exchange students with whom she could communicate well in English. Upon her return, she continued her Japanese study at her home institution where she majored in Japanese and East Asian studies.

Emma completed level four Japanese language courses at her home institution in the summer prior to this yearlong EdA program in Japan. She also participated in a short-term EdA program offered at her home institution during the summer prior to the yearlong EdA program.

In addition to getting better at Japanese, her primary reason for this yearlong EdA program was trying to get more experience “under my belt.” She also mentioned that she would try her best to speak Japanese more than she did in the first time she was in Japan as a high school exchange student without worrying about making mistakes. She also wanted to see differences in life among Japanese cities by exploring the city she stayed for the nine months and going to local festivals.

Emma studied at a private university in the Chubu region in Japan. She took Japanese language courses for international students, which especially helped her reading and writing skills. This was also reflected on her post-EdA J-CAT score discussed in Chapter Five. She also took courses for both local Japanese and international students, in which she became friends with

190

a couple of Japanese students. Emma mentioned that she wanted to have more interaction with them outside of class, but she did not have many opportunities to interact with local Japanese students at school. This is because, in a very similar situation to Anna’s, which will be discussed below, all exchange students belonged to ryuugaku-sei bekka (Center for Japanese Studies), which made it difficult to make Japanese friends. Like other participants mentioned above,

Emma also described that it was difficult for her to speak Japanese with local Japanese students.

Local Japanese students who were interested in talking to international students at parties wanted to talk to her in English, but after the parties, they “[did] not want to talk with [her] anymore.”

Language classes in the Emma’s EdA program seemed to have worked well for her.

However, in terms of helping her to develop social ties with Japanese students, it seemed to have less structured than University CFI, which provided well-structured dorms for Cathy, Frank, and

Isabelle, or University H, which had a tutor system and provided a Japanese student tutor to

Henry,

During a Skype meeting, which was a part of the EdA training, she mentioned her struggles with Japanese people’s behavior. She was often irritated by Japanese people’s surprised reaction followed by a compliment every time she spoke Japanese, even simple Japanese like

“konnichiwa (good afternoon).” She felt that the Japanese people’s behaviors were “uekara mesen (condescending attitude).” We discussed that it might be a part of Japanese culture that people often give compliments especially with someone with whom they have not developed a close relationship yet. We also discussed that Japanese people may think Japanese is too difficult for foreigners to learn, so it might be a true surprise rather than uekara-mesen.

191

She also encountered a different expectation that Japanese people hold toward non-

Japanese speakers. In a Japanese media literacy course, Emma paired with a Japanese student for a class project. When they were trying to decide a topic for their project, she was avoiding expressing her choice directly because “most of the Japanese people I came across… not being direct about your feeling.” She thought it would help her communicate with her partner effectively. However, the consequence made her confused. Her project partner told her, “you are so Japanese because you are not telling me what you want!” Emma assumed that it might have been due to her project partner’s previous experience of studying in the U.S. for a year. Because of her previous experience in the US, she may not have expected an American to behave the way

Emma did. This kind of gap in expectations of behavior as well as language between the local and U.S. students was very similar to what has been found in the fieldwork research discussed in

Chapter Three. Emma did not specifically mention how she attempted to negotiate such an expectation, but data of her social network suggests that she did not construct a social tie with the

Japanese student outside of the class.

Emma also met Kiku, with whom she and her American friend started doing Bible study together at café. Emma was excited about this new interactional opportunity because Kiku only spoke Japanese to them, which Emma thought, “(it) actually started to get somewhere.” She was also excited for new opportunities to meet other Japanese people by attending Kiku’s church.

However, Emma and her friend eventually discontinued the friendship with Kiku’s when things

“got a little weird,” and she “didn't feel comfortable pursuing a friendship with her.”

Throughout her EdA experience, Emma met a lot of exchange students. Most of them were from the classes she attended. Regarding extracurricular activities, she wanted to join some

192

university clubs, but most of the club schedules, such as dates and the starting time, did not work well with her own class schedule, so she did not attend any club activities during her EdA experience.

During the winter break, Emma visited her former host family from when she was a high school exchange student and spent a weekend there. Her host mother was Japanese and her host father was American. The two children could also communicate in English but Japanese was their L1. Based on the language log during the trip, she used both Japanese and English to talk to her host family. She described the trip as a good experience because she was able to communicate better with her former host family this time due to her higher Japanese language proficiency than before. However, their interaction seems to be limited to the trip (and before the trip), and she did not list her host family members in her social network.

Figure 14 presents Emma’s INoP during EdA program in Japan, which consists of two main clusters, University and Bible study, and four sub-clusters within the University cluster.

Compared to other participants’ INoPs, the number of nodes (seven) in Emma’s INoP is relatively small. This might have been due to her personal preference of having fewer friends who were really close than having a lot of friends but not close.

Four out of seven nodes are Japanese. Three of them are students at Emma’s host university. Similar to most of the other participants, Emma also developed strong ties with her roommates because she “spent most of my time with them simply because we lived with each other.” In addition, her strongest ties were with Wendy and Kaily who spoke English as L1

(Wendy is British and Kaily is American), which might have played a crucial role.

193

Emma Kassy Wendy

Room Kaily mates Ren

Yuko Dorm Ela Kiku Japanese Class Bible University study Foreign Miki Policy Class Emma

Figure 14. Emma's INoP

Emma made Japanese friends through attending classes, such as Miki from a foreign policy class and Ela from a Japanese class. Kiku was the one with whom Emma and her roommate Kaily did the bible study together. However, after a few meetings, Emma and Kaily decided to discontinue it due to uncomfortable feelings they experienced from their interactions with Kiku. Although Emma had higher oral language proficiency, her interaction with locals seems to be limited, which suggests having a higher oral proficiency alone does not have a great impact on constructing social ties with locals, but other variables, such as school setting, might have played a greater role in interactional opportunities, which will be discussed more in the discussion session.

194

Unlike other participants who expressed various obstacles in speaking Japanese, Emma mentioned that she did not think of any issues regarding speaking Japanese with people in her social network, which was repeatedly mentioned in her language log throughout the year. It could be due to her higher oral proficiency than other participants, as well as her previous EdA experiences prior to this EdA experience. She also mentioned that her oral proficiency was improving. In a language log in February, Emma wrote, “I did notice that compared to last semester when I did the same thing (i.e., talking with Japanese students about her home university at a school event), I was able to answer more smoothly and tell the students more with what I have learned from class.” Although her OPI rating did not indicate improvement on her oral skills, this log suggests that Emma could feel her language improvement by comparing her performance in two similar events.

6.2.6. Bobby (Low-gainer)

Similar to Emma, the findings regarding Bobby’s EdA gain discussed in Chapter Five suggested limited gain. Bobby’s social network, which will be discussed in detail below, also suggests his limited development compared to higher-gainers. His EdA program (University B) seemed to have had less structure in terms of helping Bobby’s construct social ties with Japanese students (e.g., unsuccessful experience in a school club and classes as well as constructing a relationship with his Chinese roommate).

Bobby was interested in learning foreign languages and took German language courses at his home institution at first, but it was too difficult so he switched to Japanese. He chose

Japanese because he liked Japanese anime and games. In terms of language course completion at

195

his home institution, Bobby also completed the level three Japanese language courses at his home institution and attended a summer short-term EdA program in the Kansai region in May

2017 prior to his academic-year EdA program in the Kanto region. As for his primary reason for the yearlong EdA program, he mentioned that he first started considering a yearlong EdA program in Japan when his academic counselor at his home institution strongly recommended spending a year in Japan if he wanted to major in Japanese. As he continued his Japanese study at the host university, he became more and more “excited to experience the culture.” His biggest goal for the EdA program in Japan this time was to become more fluent in the language. He also wanted to visit places in Japan, especially Akihabara67 because he was “a large fan of video games.” Later in the post-EdA interview, he described the experience of visiting Akihabara was one of the best experiences in his yearlong EdA experience in Japan.

As mentioned above, Bobby attended University B, which is one of the largest private universities in Japan and is located in the Kanto region. During the nine months of his EdA experience at the host university, Bobby stayed at a university dormitory for international students and had a roommate who was an international student from China. Despite having the same roommate for the whole nine months, he did not develop a close relationship with his roommate due to language barriers. According to Bobby, their common language was Japanese, but neither of them was comfortable having a conversation in Japanese. The crucial difference between Bobby’s dorm at University B (as well as Henry’s) and dorms at University CFI is that

67 Akihabara is a very popular shopping district for video games and is regarded as “an otaku cultural center” 196

Bobby’s dorm did not have mandatory dorm meetings, therefore, he did not have to use Japanese to interact with his Chinese roommate.

As mentioned, Bobby struggled with his language classes and school clubs. Bobby took several Japanese language courses including comprehension, kanji, and presentation courses, which he “didn’t like” because the structure of the classes was very different from the language classes at the home institution, and he felt like he did not learn much. In the Skype interview conducted in March when he was still in Japan, he mentioned that he started skipping the classes in November because he was suffering from “depression” due to receiving lower test scores than he expected despite his effort in studying for it, which demotivated him to attend the classes. He started seeing a school counselor due to his depression, but it did not help him much. When the new spring semester started, Bobby tried to attend all of his language classes in the beginning, but again he started having panic attacks when he was called on in class. The school counselor

Bobby was seeing recommended that he should have seen a psychiatrist outside of the school, who eventually recommended not attending the classes due to his panic attacks. After following his psychiatrist’s advice, Bobby stopped attending the classes.

During the Skype meeting in March, we also discussed attending some university clubs.

Bobby joined a university club where Japanese and international students do various outdoor actives together, such as visiting famous historical places or holding BBQ parties. However, after attending the activities a couple of times, Bobby stopped attending the club activities because there were a lot of drinking parties involved, and he did not like drinking much. I recommended that he should attend a club related to his hobbies. In the beginning of the new semester, Bobby and his American friend went to visit a video game circle where people interested in Japanese

197

video games gathered and played the games together. Bobby hoped that he could make some

Japanese friends who shared the same hobby with him by participating in the circle. However, most of the circle members did not seem interested in Bobby and his friend, and only a few of them talked to them. Bobby also described that because he and his friends did not know the

Japanese names of the game characters as well as tricks needed to play the games with the club members, they were not able to participate in playing the games with the Japanese members.

After this experience, Bobby and his friend did not go back to the club circle. As we will see below in the discussion of his INoP, those unsuccessful experiences affected his limited social network development.

Figure 15 shows Bobby’s INoP during his stay in Japan, which consists of one main cluster, University and three sub-clusters, Dorm, Same floor and Patrick’s friends. All eighteen nodes in the Bobby’s INoP were international students who lived in the same dorm.

Unlike most of the participants in the study, Bobby did not develop social ties outside of the dorm probably due to his withdrawal from the language program as well as his unsuccessful experience at the university video game club. His Japanese language use seems to be very limited considering he spent most of his time with international students who communicate better in English than Japanese, which is also supported by data on his language use.

According to his language use data, most of his interactive language use was English when he was logging his language use in December. During that time, his Japanese use for interactive activities was limited to the classroom, such as with his professors and classmates, or strangers, such as waiters at restaurants. During the month of February, his Japanese language

198

use had gotten much smaller, and all most all language use logs were English with friends at his dorm.

Bobby Beryl Nairo Yuri

Baker Patrick's Bob Friends Sage Chay

Dash Patrick Ailla Same Floor Ace Eldar Jet Dorm Cai Saul Cavin Aaron Dana University

Bobby

Figure 15. Bobby's INoP

Although Bobby did not have much interaction with locals, and his academic experiences were not as good as he wished, he enjoyed his stay in Japan by visiting popular sightseeing locations. He described that going to a Shibuya countdown New Year event with his international friends was the most memorable occasion. He also visited Akihabara several times and enjoyed buying video games and comics that were only sold in Japan, which was his dream since he was twelve years old.

199

6.2.7. Anna (Low-gainer)

The findings discussed in Chapter Five suggested that Anna demonstrated the least EdA gain among the seven participants in the current study. As we discussed, it might be partially due to the reliability of one of the proficiency tests (i.e., J-CAT, see Section 5.3.1.). Findings regarding her EdA social network development seemed to have accounted for her limited EdA gain.

Anna started her Japanese study at her home institution because she wanted to go to a graduate school to study military history, which required her to take foreign language courses.

She completed level three Japanese language courses at her home institution and participated in a short-term summer EdA program in Japan before her yearlong EdA exchange program in Japan.

Unlike other EdA participants whose major (or one of the double majors) was Japanese, Anna’s major was history. Her primary reason for attending an EdA program in Japan also pertained to her original purpose in studying Japanese, which she described as “immersion (learning environment) [would be] helpful” to improve her speaking, reading and writing abilities in

Japanese, which is required to go on to the Ph.D. program in history at her home institution. She also mentioned that “[it is] nice to have already had experience living abroad” before starting the

Ph.D. program and hoped that this EdA experience would provide her credentials to study in the

Ph.D. program.

Her goals for her EdA experience in Japan were also tied to her primary reason for participating in an EdA program in Japan: She wanted to maintain good grades overall to apply

200

to the Ph.D. program, to take Japanese history courses in order to learn the history “not from

American perspectives but from Japanese perspective,” and to see things in the city where she stayed, such as history of museums, so that if she decided to do research on these topics later she would already have some experience. She also wanted to learn the language better and to improve her reading and writing because her specialization in history would require her to know pre-modern kanji. She mentioned, “this [EdA experience] will help with fluency, and practice, and just the more exposure I get, more listening…, I think it really helps.” Her descriptions clearly showed that she wanted to utilize this EdA experience to gain the credentials to continue her study at graduate school. Based on her primary reasons and goals for studying Japanese as well as participating in an EdA program in Japan, her motivation can be categorized primarily as instrumental motivation (e.g. for their career) rather than integrative motivation (i.e. interests in the target culture and language).

Anna attended a national university in northern Japan and took classes that were taught in

English and Japanese. According to her, the best class she attended at the host university was a history class on local indigenous people, which was taught by an American professor in English.

They went to historical places and learned history from the indigenous people’s perspective. She was interested in learning the history and language of the indigenous people, and that was a part of her primary reason to choose University A as her EdA destination. Her favorite Japanese language class was a kanji class where the instructor “taught us the thought process, so that I could figure out.” This also relates to her primary reasons for studying Japanese and for participation in the EdA program.

201

Although she enjoyed learning about subjects related to her own specialization, Anna seems to have somewhat negative views on her host university. Despite a large number of the international student population (approximately 1850 international students at the time of the study), she described that the school “essentially segregates international students from the rest of the campus.” For example, all the classes for international students were taught in the international student center unless these classes included Japanese students. She continued, “the international student center was nice, but it was very much an island – Japanese students rarely entered the building unless they had to, and almost never hung out in the lounge unless they already had really good English.” According to Anna, no support was available at her host institution for its international students to connect with local students such as language partners or a tutor system.

She mentioned, “if we wanted to speak Japanese outside of language classes, we had to find opportunities to do so.” It seems the environment at the host university made it difficult for her to access local Japanese students, which is also apparent in her INoP described below.

Figure 16 presents Anna’s INoP during her EdA in Japan, which consists of one main cluster,

University including four sub-clusters, Drinking crowd, Roommate, Study crowd, and HU (home university) friends which all belong to University.

202

Jace Kayla Mary Kim Anna June Hong Room Chris Drinking Teddie mates Crowd Laura Tanya Foy Claire Shin Cynthis Kari Anne HU Friends Study Ava Chole Crowd University Casey JC Anna

Figure 16. Anna's INoP

All Twenty nodes with whom she constructed social ties were international students in the Japanese language program at the host institution except Shin who often spent his time at the international center with other international students. The strongest ties are with June and Tanya who were Anna’s roommates and Cynthis who was a friend from the home university and was also an international student at the host university. June and Tanya were also connected with

Anna as Drinking crowd and Study crowd respectively. Anna described that shared interests and the language (i.e., English) made it easier to become close friends with them.

However, English was also the biggest obstacle for Anna when it came to practicing

Japanese. She mentioned in the questionnaire that the international students with whom she spent most of her time in Japan “didn't want to speak it because they were still learning, and it was 203

harder for them to make conversation.” In addition, local Japanese students who befriended

Anna had no problem speaking English and “didn’t seem to care about speaking Japanese.”

Regarding her Japanese language use, she wrote “I honestly got most of my practice with

Japanese with strangers or people on the street who struck up conversations, not with friends.”

Data on self-reported language use also confirmed that most of the Japanese language use was from her Japanese language class, her self-study of the language, doctor visits, small talk with store clerks, and strangers on the street.

The findings on her social networks and EdA experience suggest that English was the primary means of communication with people in her INoP. On the other hand, she used Japanese mostly with people with whom she did not have social ties, such as strangers or clerks at stores.

6.3. Discussion

The previous sections presented the findings of the seven participants’ social networks along with their EdA learning experience from their perspectives. In this section, I discuss the findings in relation to the three research questions mentioned in the beginning of the chapter. I also discuss negotiated Third Spaces in regard to their influence on EdA gain and their development within the participants’ social networks.

6.3.1. What are the similarities and differences in the seven US students’ social networks

and EdA experiences?

The overall findings suggest four major similarities and one major difference. The first similarity is that a host universities were the primary places for developing social networks for

204

all the participants in the current study. For example, four participants (Anna, Bobby, Cathy, and

Isabelle) constructed social networks exclusively within the university setting. Even for the other three participants (Emma, Frank, and Henry), the university clusters still remain the primary places for constructing social ties. This finding suggests that the host university is an essential place for constructing social networks for EdA students.

Second, being a native speaker (or near-native level) of English seems to play an important role in the construction of strong ties. Being able to speak English freely allowed the

EdA students to feel comfortable and made it easier to discuss any topics comprehensively, which contributed to strengthening the relationships. This type of return resulting from such social ties can be considered as an expressive return, a type of social capital that people engaged in social networking (Lin et al., 2001, p. 21) described in Chapter Two. By contrast, interacting with a Japanese student in English seems to have a negative impact on building social networks even when the Japanese student had a higher English proficiency. As some participants mentioned, such a situation created a conflict between the EdA students who wanted to practice

Japanese and the local Japanese students who wanted to practice English.

In the discussion of social capital, Lin (2001) argues that a person takes instrumental action, which results in instrumental return, to obtain resources not possessed by the actor, whereas people’s action is considered as an expressive action, which results in an expressive return, when the purpose is to maintain resources already possessed by the actor (p. 21). From this point of view, EdA students seem to expect expressive returns by investing in social ties with other EdA students who shared the same L1. Because other EdA students also expected the same type of return in the relationship, the use of English in the interaction accelerates the

205

expressive returns. On the other hand, EdA students seem to have expected instrumental returns and wanted to take instrumental action to obtain resources (i.e., improvement in Japanese skill) toward interaction with local Japanese. However, because the local Japanese students also expected instrumental returns by using English with EdA students, the gap in instrumental actions between the local Japanese and the English-speakers may have negatively affected the construction of social ties.

Third, living in the same dorm and attending school clubs seem to create the best environments to meet and construct social ties with Japanese students, but there were other elements that influenced constructing ties with Japanese nodes. The findings in the current study also suggest that nodes of school club clusters often consisted of primarily local Japanese students (Cathy and Frank). However, as Isabelle and Bobby’s unsuccessful experiences in participating in a club suggest, attending club activities was not easy for some EdA participants, which is discussed in more detail later.

Fourth, pre-EdA oral proficiency level seems to have influenced social network building outside of universities. All participants who established external social ties (i.e., Frank, Henry, and Emma,68) rated as Intermediate-Mid or higher in the pre-EdA OPIs.

68 Although Emma discontinued the relationship, she had developed such a relationship, and I would argue that her pre-EdA language proficiency played an important role in initiating the relationship. 206

Table 16. Pre- and post-EdA OPI ratings Name Oral Proficiency Interview Pre-EdA > Post-EdA Cathy Intermediate-Low > Intermediate-Mid Frank Intermediate-Mid > Intermediate-High Isabelle Intermediate-Low > Intermediate-Mid Henry Intermediate-High > Advanced-High Emma Intermediate-High > Intermediate-High Bobby Intermediate-Low > Intermediate-Low Anna Novice-High > Novice-High

According to the OPI criteria (ACTFL, 2012), Intermediate speakers are able to communicate,

“predictable, familiar topics related to daily activities” (p. 12). Although Bobby, Cathy, and

Isabelle were rated as Intermediate-Low (IL) in the pre-EdA OPIs, speakers at this level still display some difficulties to do so occasionally. Interactional opportunities within a host university are often framed for its students to socialize easily. However, there are no such frameworks in interactional opportunities outside of a university. In order to develop social ties with people in such an international environment, EdA participants might have had to have full command of everyday communication skills that navigate them smoothly in such language environments. From this point, Emma, Frank, and Henry might have been able to utilize interactional opportunities with people from outside of school to initiate and maintain personal relationships, which might have been difficult for, Cathy, Isabelle, Bobby, and Anna at least in the beginning of their EdA experiences in Japan.

One major difference is that success level of network building in school clubs depends on the support from local Japanese students in the clubs, such as whether Japanese student members are interested in international students, and if they can (or are willing to) communicate in English

207

to help international student members. It also depends on the program to a lesser extent to the students’ proficiency. For example, Henry, Emma, and Anna all mentioned that although they wanted to participate in school clubs, they were not able to do so due to several reasons, such as schedule conflicts or not knowing how to attend a club. Except for University CFI, all the EdA participants mentioned that there was no club recruitment event when they started their first semester at the host universities, which made it difficult for them to find a club that they were interested in. This is mainly due to the difference in academic calendars between universities in

Japan and the US (i.e., April is often the first month of an academic year in Japan whereas

August or September is usually the first month of an academic year in the US). However, new students at the university Cathy, Frank, and Isabelle attended can start either in April or in

September, and the university holds club recruitment events both in April and in September.

Therefore, Cathy, Frank, and Isabelle could attend the club recruitment event held at the university in the beginning of their EdA program along with other new Japanese students. This might have made it easier for them to find and attend school clubs. This difference in academic calendars between the university and other universities might have also resulted in differences in opportunities to attend school clubs.

6.3.2. In what way did social networking influence the participants’ EdA gain?

Overall findings suggest that the more Japanese nodes in one’s INoP are and the more complex one’s INoP is the greater the participant’s EdA gain. Table 17 presents information regarding the seven participants’ INoPs. The numbers of Japanese nodes seems to relate to EdA gain by participants who started with lower language proficiency. When we compared Anna and

208

Bobby’s INoPs to other participants’ INoPs, it is clear that Anna and Bobby’s INoPs only consisted of people who spoke English as the primary means of communication. This suggests that EdA students who do not construct a social tie with a node with whom they communicate primarily in Japanese, have limited EdA gain.

Table 17. Summary of INoPs by gainers and low-gainers

Name Number of Nodes Number of Clusters Japanese Non-Japanese Main Sub

Group Student Non-student Student Non-student

Cathy 14 0 5 1 1 7

- Frank 5 7 7 1 2 4 Isabelle 8 0 7 0 1 4

High Gainers Henry 2 2 10 1 2 6

Anna 1 0 19 0 1 4

- Bobby 0 0 18 0 1 1

Low gainers Emma 3 1 4 0 2 4

However, for the participants who started with higher oral language proficiency (e.g.,

Emma and Henry), the complexity (or quality) of Japanese nodes seems to have impacted their oral gain. For example, Emma and Henry were both rated as Intermediate-High in the pre-EdA

OPIs. However, upon their return, Henry demonstrated his improvement in the post-EdA OPI

(i.e., Advanced-Low), but Emma’s post-EdA OPI rating remained the same. By comparing

Emma and Henry’s INoPs, there are some differences that might have influenced the differences in the post-OPI ratings. The first difference is the number of nodes to which Emma and Henry were connected. Emma had eight nodes in her INoP–that was the smallest number of nodes among the participants in the current study. By contrast, Henry’s INoP consisted of fifteen nodes.

Although eleven out of the fifteen nodes in Henry’s INoP were non-Japanese, Henry also used

209

Japanese sometimes with four of the non-Japanese nodes. In total, there were eight nodes with whom Henry used Japanese.

In addition, Emma’s nodes consisted of only students, except Kiku who was in her mid- twenties, but Emma discontinued the relationship. On the other hand, Henry’s nodes consisted of not only local students but also professors and a bartender, who were not students. Having people with different social statuses in the social network might have influenced various types of language that Henry encountered, which might have been absent in Emma’s interactional experience. In Advanced-level OPI, a test taker has to hold the floor in all advanced-level topics, which includes discussing topics that are beyond “oneself”, such as community-level. Henry commented on Professor Tani’s sophisticated language as well as intellectual topics, which made

Henry interested in but also made it difficult for him to communicate. Such interactional experience both in and outside of classes might have helped to improve his oral proficiency level.

By contrast, such interactional experience may have absent in Emma’s INoP. For example,

Emma (as well as Frank) mentioned that topics such as politics or intellectual talk or what they called “deep (fukai) topics” were avoided in Japanese conversation among friends. The current study suggested that constructing social ties with varieties of social statuses might be a crucial key for EdA students who started with higher language proficiency to demonstrate a significant gain in OPI rating.

6.3.3. Are there any differences in EdA social networks between the participants in the

EdA training group and non-EdA training groups? If so, how?

210

As shown in Table 18 below, the EdA training seems not to have a great impact on the students’ social networks during EdA programs in Japan. Table 18 categorizes the seven participants into the EdA training group (i.e. Training) and non-EdA training group (i.e. Non- training), which are shown below. Names of high-gainers (Isabelle, Frank, Cathy, and Henry) are shaded. Both groups included both high-gainers and low-gainers.

Table 18. Summary of INoPs by Training and Non-training groups Name Number of Nodes Number of

Clusters Total Japanese Non-Japanese Main Sub Group Student Non-student Student Non-student

Isabelle 15 8 0 7 0 1 4 Frank 20 5 7 7 1 2 5 Emma 8 3 1 4 0 2 4 Training Bobby 18 0 0 18 0 1 1

Cathy 20 14 0 5 1 1 7 -

Henry 15 2 2 10 1 2 6 Non

training Anna 20 1 0 19 0 1 4

If the EdA training did not have a positive impact on EdA gain or social network development, what influenced the participants’ EdA gain and social network development? The following section focuses on the question.

6.3.4. Negotiated Third Spaces during EdA experience in Japan

When we focus on the participants’ struggles and where they encountered these struggles, there is an interesting phenomenon. The findings in the current chapter revealed that high-gainers

(i.e., Cathy, Frank, Isabelle, and Henry) tended to encounter challenges in interactions with

Japanese people within their INoPs. On the other hand, when we examine low-gainers’ (i.e.,

211

Anna, Bobby, and Emma) struggles and where they experienced them, they were all outside of their INoPs. For example, one of the common difficulties among the seven EdA students in the current study was communicating in Japanese due to their lack of language proficiency. Anna mentioned this difficulty when talking about her experiences at local stores and hospitals, but not with Shin who was the only Japanese in her INoP. This is probably due to the fact that she used

English more than Japanese as a means of communication. Bobby mentioned that it was much easier for him and his friends who were all English-speaking international students to communicate in English than Japanese. In fact, it was due to the language barrier that he did not develop a social tie with his Chinese roommate despite the fact that they stayed in the same unit for the entire EdA program. Emma encountered communication difficulties when interacting with people from the post office and city hall, but Emma expressed no difficulties speaking

Japanese with the Japanese people in her INoP as mentioned above.

When we examined the similar difficulties of Cathy, Frank, Isabelle, and Henry, it revealed that they also experienced such difficulties with people within their INoPs. For example,

Cathy, Frank, and Isabelle encountered this type of difficulty when interacting with their roommates as well as Japanese students who lived in the same dorm. Cathy and Frank also encountered this type of difficulty when interacting with people in their clubs. Henry encountered this type of difficulty with the Japanese professor in his INoP, as mentioned above.

Another common difficulty shared among the participants was using Japanese with local

Japanese students who wanted to speak English. Cathy, Frank, Isabelle, Henry, and Emma shared this difficulty. When we examined this type of difficulty in their INoPs, Cathy, Frank, and

Isabelle again encountered this difficulty with their Japanese roommates as well as the Japanese

212

students who lived in the same dorms. Henry faced this difficulty when he was interacting with

Maeda, the medical student mentioned before. Emma also expressed this difficulty, but the difference with the four participants was that she did not develop a social tie with the persons who wanted to speak English. With all of the Japanese nodes in her INoP, she used Japanese to communicate most of the time (80% or more). As mentioned above, she expressed no difficulties speaking Japanese with those people.

In Figure 17, the translucent ovals in the four gainers’ INoPs indicate the social ties and/or clusters where they encountered struggles. Some of the ovals cover only one node in a cluster (e.g., Henry’s case) but others cover the whole cluster (e.g., Cathy’s case).

Figure 17. Social ties where Cathy, Frank, Isabelle, and Henry faced struggles 213

What do all the findings regarding the EdA students’ struggles in INoP suggest? When an

EdA student faces a challenge while developing social ties, it requires more negotiations consistently with the interlocutor than negotiation with people without social ties. Through repeated negotiations, the gainers may have learned how to communicate effectively in the TS they were in. Different TSs have different characteristics, therefore, the way they negotiated may have been different depending on the TS or social group (i.e., clusters). If an EdA student has opportunities for repeated negotiations in different clusters, it can create more learning opportunities for the EdA student. Having a node (i.e. a person who an EdA student constructed a social tie with) with whom EdA students negotiate consistently in order to achieve personal or mutual goals may be crucial to maximize EdA learning. Based on the findings discussed above, we will call such social ties or clusters negotiated TS.

As we discussed in Chapter One, TS refers to the intercultural communication mode, where people do not exclusively follow norms or expectations of C1 or C2, and either consciously or unconsciously co-construct a unique communication environment. In order for the people in the TS to be productive and meaningful, they need not only to understand the other side’s culture or to recognize expectations from the other side but also to negotiate their expectations so that they benefit both parties. As mentioned above, the current findings indicate constructing such negotiated TSs in INoP is crucial for EdA gain.

The question is how such a TS that allows EdA students to have consistent negotiation can be constructed during an EdA experience in Japan. The discussion above suggests that the structure of EdA programs or host universities might play an important role in creating such TSs

214

for the EdA students. For example, all participants in the current study lived in the dorms.

However, the structure of the dorms at the university Cathy, Frank, and Isabelle attended provided more opportunities for both international and local Japanese students to interact and negotiate than the universities that Henry, Emma, Bobby, and Anna attended. For example, there was a mandatory dorm meeting every month, which lasted one to two hours according to the language use data from Cathy and Isabelle. Frank mentioned that casual events or parties, such as movie nights or a BBQ party, were occasionally held at his dorm. In addition, the university held a club recruitment event at the same time when most of the international students started their study at the university, which helped Cathy, Frank, and Isabelle to participate in clubs

(although Isabelle discontinued her membership in the band music club).

Although there seems to be much less structure in Henry’s university compared to the university that Cathy, Isabelle, and Frank attended, Henry’s university provided a Japanese tutor with whom Henry met for one hour once a week and discussed various topics in Japanese.

However, for Henry’s case, his higher language proficiency may have also matched perfectly with the program structure. For example, he was able to take general education courses at the university with other Japanese students. Even though he had to spend a large amount of time to prepare for these classes, through the learning of the contents he learned advanced language required in the academic and professional fields. Because of such an intensive academic learning environment as well as constructing a social tie with the Japanese professor, his EdA gain was demonstrated in his OPI rating as well as other assessments. However, a student with lower language proficiency may have felt overwhelmed by the language-learning environment.

215

Of course, there must have been many forms of assistance that international student centers at the other host universities offered. For example, Bobby expressed how helpful the school staff and his Japanese teachers were when he contacted them regarding his problem in the classes. However, this assistance and consideration might have been structured in the way that

Bobby did not have to negotiate in order to find a solution (such as providing English translations, etc.). Such assistance was very crucial for Bobby to continue his EdA program.

However, considering his and the other non-gainers’ EdA learning experience, the universities might not have provided opportunities where they could negotiate with local students and staff.

Along with other research (Baker-Smemoe et al., 2014; Isabelle-Garcia, 2006, etc.), the overall findings in the current study also suggest that strong social ties with locals, more specifically social ties in negotiated TS, have a positive effect on the participants’ EdA gain. In addition, the current findings suggest that that the designs of EdA programs have the greatest impact on the construction of such social networks. This finding is significant because the structures or characteristics of EdA programs often ignored in most EdA studies, at least, the ones discussed in Chapter Two. For example, most of these studies did not consider programmatic differences in relation to the outcomes or the process of EdA learning. By contrast,

Dewey et al (2014) examine EdA program as a variable and report that participating in specific

EdA programs is a significant predictor.69 The current study also argues that the structure of EdA programs or host universities has the greatest impact on the EdA students’ social network development, which facilitates EdA gain. Considering the aforementioned points, the next and

69 Dewey et al. (2014) reports that participating in specific EdA programs is a significant predictor of L2 use (pp. 50-51). 216

final chapter, Chapter Seven, includes a discussion on pedagogical implications for home institutions.

217

Chapter 7: Maximizing Education Abroad Learning in Japan

7.1. Summary of the objectives and the findings of the research

This dissertation aimed to investigate (1) types of cultural and linguistic issues U.S. undergraduate students encounter regarding building personal relationships during EdA programs in Japan; (2) ways in which EdA training can address these issues described in (1); and

(3) the degree to which EdA training can mitigate these issues.

The overall findings suggested that social networks played a crucial role in the seven participants’ EdA gain. More specifically, social networks that included negotiated Third Space

(TS) seemed to have a positive impact on their EdA gain. In addition, the findings suggest that the design of the EdA programs may have had the greatest influence on the construction of negotiated TS in EdA participants’ social networks.

One of the most crucial findings regarding the relationships between social networks and

EdA gain was that participants who had negotiated TS in their social networks (i.e., INoPs) demonstrated more EdA gain. The findings on social networks, including participants’ EdA experiences suggest that EdA programs (i.e., the structure of the language programs, dorm, extracurricular activities, etc.) at the host universities seemed to most strongly influence the production of negotiated TS in social networks. Considering that the host universities were the

218

primary place for constructing social networks with both locals and non-Japanese persons, it is understandable that EdA programs had the most impact on EdA students’ social networks. As mentioned in Chapter One, most undergraduate students who participated in EdA programs in

Japan have at least Intermediate-low level proficiency on the OPI scale, while structures and/or frameworks that EdA programs offered enabled their EdA students to engage in intercultural communication and intercultural relationship construction. Without such structures, they might not be able to do so.

The findings regarding social networks during EdA programs in Japan revealed that host universities were the primary place to construct social networks for the participants in the current study (see Chapter Six for detailed discussion of the findings on social networks during EdA programs in Japan). In addition, most of them developed social ties with non-Japanese international students who were fluent in English, which enabled them to receive expressive returns.70 The use of English seemed to strengthen such relationships. By contrast, being fluent in English seemed a hindrance when communicating with local Japanese students. In fact, most of the EdA students in the current study struggled with local Japanese students who wanted to speak English with them when the EdA students wanted to practice Japanese. In terms of establishing external social ties (i.e. social ties that were constructed outside of the host university setting), the findings revealed that only the participants who started with Intermediate-

Mid or higher oral language proficiency established social ties outside of the host universities.

This suggests that EdA students have to be able to fully handle communication in predictable

70 Some examples of expressive returns are physical health, mental health, and life satisfaction (Lin, 2001, p. 21). 219

and familiar contexts and actively engage in communication if they want to construct social ties outside of a school setting.

The findings in EdA gain revealed that most of the participants demonstrated EdA gain at least with one assessment, but there were individual differences in EdA gain among the participants (see Chapter Five for detailed discussion of the findings on EdA gain). In general, gain in the skills measured by the Japanese Computerized Adaptive Test (J-CAT) seemed easier to demonstrate than the oral productive skills measured with the Oral Proficiency Interview

(OPI). EdA gain measured with J-CAT suggested that EdA students could benefit in vocabulary and reading skills regardless of their pre-EdA proficiency, from yearlong EdA programs in Japan.

On the other hand, grammar skills seemed most difficult to gain during EdA experiences.

Examination of style shifting revealed that all participants demonstrated some kind of improvement such as increased use of plain forms, non-use or self-correction of inappropriate informal speech style (IfSS) in situations where the speakers had high awareness of the addressee, or being able to use IfSS appropriately in a casual speech setting. Because the criteria of the two proficiency tests did not include cultural knowledge, the examination of style-shifting development provided the students’ EdA gain in regards to the students’ language performance in a culturally appropriate manner, which is a crucial communication skill to develop personal relationships in L2.

220

Considering the findings discussed above, the next section discusses pedagogical implications71 for U.S. college students who participate in yearlong undergraduate EdA programs in Japan.

7.2. Home institution training for EdA participants in Japan

Although the overall findings suggest that the structure of EdA programs has had the greatest impact on the construction of social networks, appropriate preparation at a home institution can also help EdA students to engage in social network construction with locals, which can facilitate their EdA gain. This section focuses on pedagogical support by a home institution for American college students, more specifically for American undergraduate students who participate in yearlong EdA exchange programs in Japan.

The findings regarding the construction of negotiated TS suggest the necessity of supporting EdA students’ negotiation skill development in Japanese that is appropriate in EdA context. College students who attend yearlong exchange EdA programs usually spend time in less structured learning environments than students who participate in a faculty-led EdA program.

The less structured learning environment may work better for some EdA students who already established some social networks in Japan or those who already have high enough linguistic and cultural skills to actively participate in community activities and construct social ties with locals like Henry did. However, such a less structured learning environment may overwhelm and demotivate students who started the program with lower linguistic and cultural skills. Such

71 Pedagogical implications based on the findings in the first phase of the study were discussed at the end of Chapter Three. Therefore, this chapter focuses on pedagogical implications based on the findings in the second phase of the research primary discussed in Chapter Five and Six. 221

learning environments can become obstacles to taking the initiative to utilize interactional opportunities. The participants in the current study shared the opinion that the structure and content coverage of the language classes often differ from the ones at a home institution, which confused them and required some time to get accustomed. The following sections discuss pre-, concurrent, and post-EdA training for yearlong EdA exchange students.

7.2.1 Pre-departure training duration and content, cast of characters: Focusing on

negotiation skills and the process of EdA learning

Pre-departure training should include contents to help students to develop negotiation skills as well as viable personae in EdA contexts. Although it was already discussed in Chapter

Three as a pedagogical implication, the importance of developing skills to manage the informal conversation, including abilities for style shifting in Japanese, cannot be overemphasized. For more concrete pedagogical support for style shifting, please see Chapter Three. In the current study, the practice of self-introduction focusing on style shifting was done once individually via

Skype due to restrictions of time and location. Although all four participants who went through the pre-EdA training mentioned that it was very helpful, they also reported that having multiple practices to develop accuracy and fluency in interactive performance would have been more beneficial than doing it once.

Because repeated practice is crucial for smooth communication, style-shifting skills should also be practiced in formal instruction at the home institution. Traditional formal

222

instruction or Japanese language textbooks72 often deemphasize the importance of such communication skills and focus on not mixing speech styles within a conversation. However, previous research on Japanese L1 and L2 speakers’ communication, as well as the findings on the current study suggest that such skills are crucial for all L2 learners to communicate effectively.

Although such performance in casual speech style is primarily for students who participate in undergraduate EdA programs in Japan, students who wish to participate in graduate-level programs or to work for Japanese companies in Japan can benefit from such instruction considering that even at a workplace, informal communication skills, for example at nomi-kai (drinking parties), are necessary to develop close relationships with colleagues.

As mentioned before, style shifting can be used as a communication strategy because it enables one to recognize, negotiate, and co-construct expectations, but these expectations differ depending on contexts, and naturally shift even within the same context. Practice with style shifting skills helps students to respond and continue to develop a viable persona in a given encounter. It is also helpful to practice cultural scripts to negotiate local peoples’ expectations.

One possible cultural script, based on the findings on this research, would one in which the non-

Japanese explains his/her difficulty in using plain forms, including their conjugation as well as their primary learning experience of using masu forms at the home institution to a local Japanese person. This is more practical for yearlong EdA exchange programs where training for local

72 NihonGo Now! (Noda et a. forthcoming), which includes style shifting throughout the lessons, would be a great pedagogical material to practice such communication skills. 223

Japanese students is difficult. By practicing such scripts, EdA students can initiate negotiations of Japanese people’s expectations toward their speech style.

In addition, students can benefit from group discussion on potential struggles, such as how to deal with Japanese students who want to speak only English. During the discussion, an instructor shares a scenario on each struggle that the students may encounter during EdA programs. Such discussions must aim to negotiate and co-construct productive TS with locals, which enables students to construct social ties with the locals rather than a confrontational TS where they discontinue future interaction. Due to the complexity of the content, such discussion should be done in English so that all students can engage in discussion. After discussion, the students create their own script in Japanese by incorporating what they have discussed, which is followed by a practice of the performance. Although the instructor should provide feedback on language and cultural aspects during discussion and the process of script making, he or she should not provide an answer for the students. In this way, the students will learn how to deal with struggles by themselves during EdA programs and prepare for it. All scripts are shared in the class, and through these, students can learn other ways to deal with struggles, and it provides them more choices to use during EdA programs in Japan based on what context they encounter.

To optimize students’ learning, pre-departure training should be done in person rather than online, and as a group rather than an individual lesson as done in the current study.

The findings also suggest that it is crucial to understand the EdA learning from the students’ perspective, and pre-EdA training should provide support that meets students’ needs in different EdA programs. One way to do so is to invite former EdA students to share their experiences, including their struggles (which I consider to be the most helpful advice that pre-

224

EdA students can get from former EdA students), in pre-departure training. I found that after interviewing more than thirty former EdA students in the last seven years, these students were often eager to share their experiences in Japan, and many of them thanked me for giving them the opportunity to share their experience (even those with negative experiences). Inviting former

EdA students to pre-departure training is a win-win situation for both former EdA students who want to share their experience in Japan and pre-EdA students who desire to learn about their EdA program from former EdA students’ perspectives. In addition, such former EdA students’ perspectives provide the home institution valuable insights on how EdA programs are structured, which is something that they may not able to learn from direct communication with the EdA programs, or from information on official university websites or EdA program pamphlets.

Choosing the right EdA program based on not only each EdA students’ academic needs but also their interests in other fields is one of the most important supports that a home institution can provide. All participants in the current study mentioned that their yearlong EdA experiences were successful and great despite the struggles that they faced. For example, Anna, who demonstrated the least EdA gain among those in the current study (see Chapter Five), so enjoyed taking history courses on local indigenous people that she wished to incorporate the contents into her PhD research at the home institution. Bobby, who also did not demonstrate EdA gain as much as other participants and stopped attending classes by the end, enjoyed exploring local places that were closely related to his interests and hobbies. If both of them were sent to different

EdA programs in Japan, they might not have had such positive experiences, and might have had to return home in the middle of their EdA programs.

225

However, as mentioned mainly in Chapter Six, both Anna and Bobby struggled with constructing social ties with locals partially due to their lower language proficiency and partially due to how their EdA programs were structured (see Chapter Six). For example, Anna, Bobby, and other participants (Emma and Henry) complained that it was difficult to find and join school clubs because there was no major school event to help them to do so. Having higher oral language proficiency levels, Emma and Henry were able to find other ways to connect with local students, but without such help from the host institutions, it might have been extremely difficult for Anna and Bobby to do so. It might be also possible that their lower oral language proficiency hindered them from finding helpful resources that host institutions had provided. Considering that, it is crucial for home institutions to understand each EdA program at their affiliated

Japanese universities and what resources that EdA students can use to get access to local students based on their pre-EdA language proficiency.

7.2.2. Concurrent-EdA training with the utilization of technology

Once EdA programs start, students often have to deal with struggles mostly by themselves. This is why pre-departure training is crucial. While the main focus of the pre- departure training is to prepare for potential struggles stemming from cultural differences and to determine how to overcome such struggles by training them with communication strategies, the main purpose of the concurrent training is to check the progress of EdA learning and help the

EdA students to find solutions to overcome struggles or concerns regarding Japanese language and culture if needed. Concurrent training can be done by utilizing technology, such as Skype or

Google Hangout.

226

If possible, having students conducting participant observation is also a great way for them to actively engage in their EdA learning (Kawamura, 2007). Such online feedback might not be as effective as in-person (Kawamura, 2018), but using pedagogical materials designed for such actives (e.g., Action! Japan by Noda et al., 2017) may compensate such gaps. Findings from the participant observations can be shared in an online bulletin board with other EdA participants to encourage student discussion. However, considering intense schedules during

EdA programs as mentioned by all participants in the current study, such activities do not have to take place frequently during EdA programs. Rather, such data can be used as materials in the post-EdA training.

7.2.3. Post-EdA training

Providing support post-EdA program is as important as support pre- and concurrent-EdA program to facilitate further learning in Japanese language and culture. Many EdA students experience reverse culture shock (Paige et al., 2002) upon returning to their home country. Post-

EdA training is a great place for the EdA students to overcome reserve culture shock and share their experiences including TSs they experienced and findings from participant observation if it was conducted. This can be done as a course in which students use both English and Japanese, depending on the level of proficiency and difficulty of topics. As a final project, they present their findings and learning experience in Japanese. As a class, students discuss and summarize their findings and learning experience and make an English version of the presentation, which can be used at the pre-departure training for students who attend EdA programs in Japan the following year.

227

As mentioned above, returned students can serve as a student adviser for the students who will attend the same EdA program in Japan, and it can be also a part of the course requirements to provide more specific information and advice to maximize their learning experience at each host university. Such a sequence of activities pre-, concurrent-, and post-EdA program not only helps EdA students to receive academic (instrumental) returns, but also provides them a resource from which they can seek psychological (expressive) returns when they encounter struggles during their EdA program by connecting with former and potential EdA students at their home institution.

7.3. Limitations and future research implications

Although the current research was carefully designed to overcome the limitations found in the previous research, it also included some constraints discussed in Chapter One. In addition to such constraints, discussion of the findings in the current research uncovered some other limitations. This section discusses the limitations regarding style-shifting data, social-network data, and participants, and how future research can address these limitations.

7.3.1. Style shifting as a means of understanding EdA gain

The findings regarding EdA gain revealed that style shifting was a useful measurement for EdA gain in terms of its sensitivity to participants’ gain in culturally appropriate language use.

However, using OPI data for style shifting might not measure such EdA gain effectively. For example, because it was an interview format, most of the participants’ utterances were responses to the interviewer’s questions, which made it difficult to measure other aspects of the participants’

228

communication skills i.e., skills as listeners (e.g. aizuchi). In addition, research has found that

DtSS-PFs were often used by the interviewer, whose primary role was to elicit information from the interlocutor. Furthermore, as Zeng (2015) mentions, some of the techniques used in OPIs might not reflect real communication because the interviewer has a set of agenda to measure the interviewee’s oral skills based on the descriptions of the OPI scale. Such an agenda (i.e., measuring the interviewee’s oral skills) rarely happens in daily communication. Because style shifting is primarily done to make communication go smoothly (e.g. avoiding sounding unfriendly or too friendly in a given context, focusing on contents rather than the relationships), using natural conversation as its data would have provided better insights on EdA gain than using OPI as data for examination of style-shifting development.

Therefore, future research that wishes to employ style shifting as a measurement of EdA gain should collect natural conversation data, both informal and formal, to examine style-shifting development. As briefly discussed in Chapter Two, style-shifting occurring in a plain-form- predominant discourse context differs from that in a masu-form-predominant discourse context.

In addition, the ability of style shifting requires a speaker to select the appropriate speech style based on a given context and to conjugate masu forms and plain forms accurately and smoothly.

Therefore, having both informal and formal conversation data will enable researchers to assess learners’ style-shifting development from different angles to capture the whole picture of EdA gain in realistic communication.

In addition to collecting style-shifting data, analysis of the data used in the current study needs some modifications to better understand the process of style-shifting development during

EdA programs in Japan. For example, the causes and reasons for Emma’s change from the DtSS-

229

PF (i.e., nan da kke) to the IfSS-PF (nan te iu) as a soliloquy-like utterance was unclear.

Stimulated recall can be used as an instrument to obtain students’ perspectives on such changes.

The judgment of IfSS-PF or DtSS-PF was done by myself based on the findings from previous research on L1 speakers’ style shifting. However, expecting different “norms and rules of interaction” in a TS, judgment based only on how L1 speakers’ shift their speech styles in a given context might not reflect style-shifting abilities that facilitate successful intercultural communication in each TS, as revealed in Zeng’s study (2015). Employment of the matched- guise test, a sociolinguistic experimental technique (Lambert et al., 1960), would be a useful instrument to obtain Japanese native speakers’ perspectives on the non-native speakers’ communicative capabilities in both informal and formal settings. In addition, by adding assessments by L1 speakers who are not teachers of Japanese as a foreign language, findings can provide new insights into the teaching of Japanese to be able to communicate successfully in TS outside of the classroom.

7.3.2. Social networks as a means of understanding the process of EdA learning

The current study investigated the participants’ social network development as a means to understand the process of EdA learning. The current research revealed insights on how social network development influenced the participants’ EdA gain or non-EdA gain (see Chapter Six for detailed discussion on the relationships). However, the social network data used in the current study was corrected once after the participants returned from their EdA programs. Because some of the participants mentioned that they had to stop their relationships with locals (e.g. Emma’s case), collecting social network data a couple of times during yearlong EdA programs may have

230

demonstrated changes in their social network better. As such, the data should demonstrate more clear processes of developing social networks in Japan, which is crucial to understanding what actually happens while they are living and learning in Japan.

Another way to improve the reliability of social network data is to obtain other perspectives on EdA students’ social network data. For example, having data based on observation and interviewing Japanese people in the EdA participants’ social networks would enable an analysis of social network development more in-depth (Zappa-Hollman & Duff, 2015).

7.2.4. Representing diversity in EdA learning

Chapter One discussed limitations regarding the participants from their home institution.

To overcome this constraint, the current research provided detailed descriptions of the participants’ perceptions before and after EdA programs in Japan including their previous learning experiences at the home institution. In addition, in order to represent various EdA learning experiences, the participants were recruited at a U.S. university where they offered various yearlong EdA exchange programs in Japan. The attempt to recruit all EdA participants during 2016-2017 academic year was made by sending emails multiple times. However, two of them did not respond to the emails, and one failed to meet two requirements. Therefore, seven out of ten EdA students participated in the current study. Although individual differences were found in terms of EdA gain and development of social networks, all of the participants in the current study successfully completed their EdA programs without returning home in the middle of the EdA programs. From this point of view, data on the current study may only reflect

American undergraduate students who successfully completed their EdA programs. EdA

231

learning experience from the perspectives of students who discontinued in the middle of their

EdA programs might have provided different insights on their EdA learning experiences.

Furthermore, all the participants in the current study happened to be Caucasian who spent most of their lives in the Midwest prior to EdA programs in Japan. Previous studies revealed that EdA students’ differences in the race resulted in different interactional opportunities even among the students with the same nationality (Talburt & Stewart, 1999;

Twombly, 1995). Having students who represent different backgrounds including race might have provided a more diverse EdA learning experience. Such data is also important in designing pedagogical support to maximize American undergraduate students’ EdA learning in Japan.

7.4. Conclusion

The findings suggest that in order to maximize EdA learning, it is crucial to provide pedagogical support to students’ development of relationship-building persona at Japanese undergraduate settings by utilizing style shifting as one communication strategy that EdA students can employ to recognize, negotiate and co-construct expectations with the locals. In addition, it is essential to understand each EdA program from the students’ perspectives and utilizes them as a valuable resource to provide support to students who will participate in the same EdA program. Such support enables EdA students to develop a relationship-building persona to forge intercultural relationships with locals, which facilitates their EdA gain.

232

Bibliography

Agar, M. (1993). Language shock: Understanding the culture of conversation. New York: Wm. Morrow.

American Council on the Teaching of Foreign Languages (2012). AFTCL Proficiency Guidelines 2012. Retrieved from https://www.actfl.org/publications/guidelines-and-manuals/actfl- proficiency-guidelines-2012

Baker-Smemoe, W., Dewey, D. P., Bown, J., & Martinsen, R. A. (2014). Variables Affecting L2 Gains During Study Abroad. Foreign Language Annals, 47(3), 464–486. Retrieved from http://search.ebscohost.com.proxy.lib.ohio- state.edu/login.aspx?direct=true&db=mzh&AN=2014652895&site=eds-live&scope=site

Batstone, R. (2002). Contexts of Engagement: A Discourse Perspective on “Intake” and Pushed out.” System 30(2002), 1-14.

Berg, V. M., Connor-Linton, J. & Paige, M. R. (2009). The Georgetown Consortium Project: Interventions for Student Learning Abroad. Frontiers: The interdisciplinary journal of study abroad, 3, 1-48.

Brecht, R., Davidson, D. E. & Ginsberg. R. B. (1995). Predictors of foreign language gain during study abroad. In B. F. Freed (Ed.), Second language acquisition in a study abroad context (pp. 37-66). Amsterdam/Philadelphia: John Benjamins Publishing Company.

Brecht, R. & Robinson, J. (1995). On the Value of Formal Instruction in Study Abroad: Students Reaction in Context. In B. F. Freed (Ed.), Second language acquisition in a study abroad context (pp. 317-334). Amsterdam/Philadelphia: John Benjamins Publishing Company.

Bourdieu, P. & Passeron, J.C. (1977). Reproduction in education, society and culture, trans. Richard Nice. London: Sage.

Burns, P. D. (1996). Foreign students in japan: A qualitative study of interpersonal relations between north American university exchange students and their Japanese hosts (Order

233

No. 9709577). Available from ProQuest Dissertations & Theses A&I; ProQuest Dissertations & Theses Global. (304307255). Retrieved from http://proxy.lib.ohio- state.edu/login?url=https://search-proquest-com.proxy.lib.ohio- state.edu/docview/304307255?accountid=9783

Christen, M., & Warnick, J. P. (2006). Performed culture: An approach to East Asian language pedagogy. Columbus: National East Asian Language Resource Center.

Cohen, A. D., Paige, M., Shively, R. L., Emert, H. A., & Hoff, J. G. (2005) Maximizing study abroad through language and culture strategies: Research on students, study abroad program professionals and language instructors. Minneapolis, MN: The Learning Abroad Center, Office for International Programs, University of Minnesota. Retrieved from https://carla.umn.edu/maxsa/documents/MAXSAResearchReport.pdf

Cook, H. M. (2006). Japanese politeness as an interactional achievement: academic consultation sessions in Japanese universities. Multilingua 25, 269–292.

Cook, H. M., (2008a). Style Shifts in Japanese academic consultations. In Jones, K., & Ono, T. (eds). Style shifting in Japanese, 9-38, Amsterdam: John Benjamins Pub.

Cook, H. M., (2008b). Construction of Speech Styles: the Case of Japanese Naked Plain Form. In Mori, Junko & Ohta, Amy S. (eds). Japanese applied linguistics, 80-108, London/New York: Continuum.

Cook, H. M. (2008c). Socializing identities through speech style: Learners of Japanese as a foreign language. Bristol: Multilingual Matters.

Davidson, D. E. (2010). Study Abroad: When, How Long, and With What Results? New Data From the Russian Front. Foreign Language Annals, 43, 1, 6-26.

Dewey, D., Ring, S, Gardner, D. & Belnap, R.K. (2013). Social network formation and development during study abroad in the Middle East. System 41, 1, 269-282.

Dewey, D., Bown J., & Eggett, D. (2012a). Japanese Language Proficiency, Social Networking, and Language Use during Study Abroad: Learners’ Perspectives. The Canadian Modern Language Review/La Revue canadienne des langues vivantes, 68, 2, 111–137.

Dewey, Belnap, R.K., & Hillstrom, R. (2012b). Social Network Development, Language Use, and Language Acquisition during Study Abroad: Arabic Language Learners’ Perspectives. Frontiers: The Interdisciplinary Journal of Study Abroad, 2012, 84-110.

DuFon, M. A., & Churchill, E. (2006). Language learners in study abroad contexts. Clevedon, England: Multilingual Matters.

234

Feed, B. (1995). Second language acquisition in a study abroad context. Amsterdam/Philadelphia: John Benjamins Publishing Company.

Freed, B. (1990). Language learning in a study abroad context: effective of interactive and non- integrative out-of-class contact on grammatical achievement and oral proficiency. Linguistic, language teaching and language acquisition: The interdependence of theory, practice and research. (GURT 1990) ed. by J. Atlatis, 459-477. Washington D.C.: Georgetown University Press.

Hammerly, H. (1982). Synthesis in second language teaching: An introduction to linguistics. Blaine, Wash: Second Language.

Hernandez, T. A.. (2010) The Relationship Among Motivation, Interaction, and the Development of Second Language Oral Proficiency in a Study-Abroad Context. Modern language journal, 94(4), 600-617.

Hudson, M. E. (2011). Student honorifics usage in conversations with professors. Journal of Pragmatics, 43, 15, 3689-3706.

Huebner, T. (1995). The effects of overseas language programs: Report on a case study of an intensive Japanese course. In B. F. Freed (Ed.), Second language acquisition in a study abroad context. (p.171-193) Amsterdam/Philadelphia: John Benjamins Publishing Company.

Iino, M. (2006). Norms of interaction in a Japanese homestay setting: Toward a Two-Way flow of linguistic and cultural resources. In DuFon, A.M., & Churchill, E. (Eds.) Language learners in study abroad contexts. (p151-172) NY: Multilingul Matters Ltd.

Ikuta, S. (1983). Speech level shift and conversational strategy in Japanese discourse. Language Sciences, 5, 1, 37-53.

Institute of International Education (2018). Open Doors. Retrieved from: https://www.iie.org/Research-and-Insights/Open-Doors/Data/US-Study- Abroad/Destinations on Dec. 28, 2018.

Isabelli-Garcia, C. (2006). Study Abroad Social Networks, Motivation and Attitudes: Implications for Second Langue Acquisition. Language learners in study abroad contexts,. In DuFon, A.M., & Churchill, E. (Eds.) Language learners in study abroad contexts. (pp.231-258) NY: Multilingul Matters Ltd.

235

Iwasaki, N. (2011). Learning L2 Japanese “Politeness” and “Impoliteness”: Young American Men’s Dilemmas during Studying Abroad. Japanese language and literature, 45(1), 67- 106.

Iwasaki, N. (2008). Style Shifts among Japanese Learners before and after Study Abroad in Japan: Becoming Active Social Agents in Japanese. Applied Linguistics, 31, 1, 45-71.

Iwasaki, N. (2007). Assessing Progress towards Advanced Level Japanese after a Year Abroad: Focus on Individual Learners. Japanese language and literature, 41, 2, 271-296.

Jian, X. (2018). A “Third-Space” View on the Pedagogy of East Asian Languages and Cultures. Symposium of Interdisciplinary Approaches to East Language Pedagogy, Columbus, OH. February 24.

Jorden, E. H. & Noda, M. (1987). Japanese: the Spoken Language part 1. New Haven: Yale University.

Kawamura, H (2007). Participant Observation for Language Learners: A Performance-Based Approach to Language Learning during Study Abroad. Japanese Language and Literature, 41, 2, 333-349.

Kinginger, C. (2009). Language learning and study abroad: A critical reading of research. Basingstoke, England: Palgrave Macmillan.

Kramsch, C. (2009). The Multilingual Subject. Oxford: Oxford University Press.

Labov, W. (1972). Language in the Inner City. Philadelphia, PA: University of Pennsylvania Press.

Labov, W. (1997). Some Further Steps in Narrative Analysis. The Journal of Narrative and Life History 7, 395-415.

Lambert, W. E., Hodgson, R. C., Gardner, R. C., & Fillenbaum, S. (1960). Evaluational reactions to spoken languages. The Journal of Abnormal and Social Psychology, 60(1), 44-51.

Lin, N., Cook, K. S., & Burt, R. S. (2001). Social capital: Theory and research. New York: Aldine de Gruyter.

Makino, S. (2002). When does communication turn mentally inward? A case study of Japanese formal-to-informal switching. In: Akatsuka, N., Strauss, S. (Eds.), Japanese/Korean Linguistics 10. CSLI, Stanford, 121–135.

236

Marriott, H. (1995). The acquisition of politeness patterns by exchange students in Japan. In B. Freed (Ed.), Second Language Acquisition in a study abroad Context, 197-224. Amsterdam: Benjamins.

Masuda, K. (2010). Nihongo gakusyuu-sya no buntai sifuto ni tsuite. In Minami, M. (Ed.) Gengo-gaku to Nihongo-kyooiku no Tenkai to Hyousiki VI, 191-212. Tokyo: Kuroshio Syuppan.

Maynard, S. K. (2000). Jōi no gengogaku: "ba kōshōron" to Nihongo hyōgen no patosu. Tōkyō: Kuroshio Shuppan.

Maynard, S. K. (1993). Discourse modality: Subjectivity, emotion, and voice in the Japanese language. Amsterdam: J. Benjamins Pub. Co.

Milroy, L. (1987). Language and social network. New York: B. Blackwell.

Mori, J. & Ohta, A. S. (2008). Japanese applied linguistics. London/New York: Continuum.

Noda, M. (2007). Performed Culture: Cataloguing Culture Gains during Study Abroad. Japanese Language and Literature 41, 297-314.

Noda, M. (2014). Learning East Asian Languages in Cross Cultural Contexts: College Level [PowerPoint presentation].

Noda, M., Ramdeen, Y., Luft, S., & Mason, T. (2018). Action! Japan: A field guide to using Japanese in the community. New York: Routledge.

Okamoto, S. (1999). Situated politeness: Manipulating honorific and non-honorific expressions in Japanese conversations. Pragmatics, 9, 1, 51-74.

Oxford University Press (2018). English: Oxford Living Dictionary. Available from: https://en.oxforddictionaries.com/

Paige, R. M., & University of Minnesota. (2002). Maximizing study abroad: A students' guide to strategies for language and culture learning and use. Minneapolis, Minn: Center for Advanced Research on Language Acquisition, University of Minnesota.

Polanyi, L. (1995). Language learning and living abroad: Stories from the field. In B. F. Freed (Ed.), Second language acquisition in a study abroad context. (pp.271-291) Amsterdam/Philadelphia: John Benjamins Publishing Company.

Ringbom, H. (1980). On the Distinction between Second-Language Acquisition and Foreign- Language Learning. Education resources information center (ERIC).

237

Siegal, M. (1995). Individual differences and study abroad: Women learning Japanese in Japan. In B. F. Freed (Ed.), Second language acquisition in a study abroad context. (pp.225- 244). Amsterdam/Philadelphia: John Benjamins Publishing Company.

Taguchi, N. (2015). Developing interactional competence in a Japanese study abroad context. Tonawanda, NY: Bristol: Multilingual Matters.

Talmy, S. (2008). The Cultural Productions of the ESL Student at Tradewinds High: Contingency, Multidirectionality, and Identity in L2 Socialization. Applied linguistic, 29(4), 619-644.

Talburt, S. & Stewart, M.A. (1999). What’s the Subject of Study Abroad?: Race, Gender, and “Living Culture.” The Modern language journal, 83(2), 163-175.

Terry, K. M. K. (2017). Contact, Context, and Collocation: The Emergence of Sociostylistic Variation in L2 French Learners during Study Abroad. Studies in Second Language Acquisition, 39, 3, 553-578.

Tobaru, H. (2014). Study abroad in Japan: A case study of three American undergraduates individual and cultural experiences. (Unpublished master’s thesis) University of Findlay, Findlay, Ohio.

Twombly, S. B. (1995). “Piropos” and Friendships: Gender and Culture Clash in Study Abroad. Frontiers: The Interdisciplinary Journal of Study Abroad, 1, 1–27. Retrieved from http://search.ebscohost.com.proxy.lib.ohio- state.edu/login.aspx?direct=true&db=eric&AN=EJ608162&site=eds-live&scope=site

Wang, C. (2010). Toward a Second Language Socialization Perspective: Issues in Study Abroad Research. Foreign Language Annals, 43(1), 50–63. Retrieved from http://search.ebscohost.com.proxy.lib.ohio- state.edu/login.aspx?direct=true&db=eric&AN=EJ890410&site=eds-live&scope=site

Walker, G. (2010). Performed culture: Learning to participate in another culture. In G. Walker (Ed.), The pedagogy of performing another culture, (pp. 1-20). Columbus, OH: National East Asian Languages Resource Center.

Walker, G & Noda, M (2010). Remembering the Future: Compiling Knowledge of Another Culture. In G. Walker (Ed.) The pedagogy of performing another culture, (pp. 22-50). Columbus, OH: National East Asian Languages Resource Center.

Wilkinson, S. (1998). Study Abroad from the Participants' Perspective: A Challenge to Common Beliefs. Foreign Language Annals, 31, 1, 23-39.

238

Yao, K. (2017). Effectiveness of Excuses in Japanese Business Context: Accounts as Conflict- Management Strategies. (Electronic Thesis or Dissertation). Retrieved from https://etd.ohiolink.edu/

Yoshimura, N. & Nakayama, M. (2018). Daini gengo shūtoku kenkyū eno izanai: riron kara jisshō e / Yoshimura Noriko, Nakayama Mineharu = An invitation to second language acquisition research. Tokyo: Kurosio Syuppan.

Zappa-Hollman, S., & Duff, P. A. (2015). Academic English Socialization through Individual Networks of Practice. Tesol Quarterly: a Journal for Teachers of English to Speakers of Other Languages and of Standard English As a Second Dialect, 49, 2, 333-368.

Zeng, Z. (2015). Demonstrating and Evaluating Expertise in Communicating in Chinese as a Foreign Language. (Electronic Thesis or Dissertation). Retrieved from https://etd.ohiolink.edu/

239

Appendix A. Phase 1: Interview Questions with EdA students (English)

1. What was your primary reason for participating in this program? 2. Have you spent any time in Japan prior to this program? (If so, please consider the rest of the questions specifically about your experience during this Kobe program.) 3. During your stay in Kobe so far, who have you spent time with most? (Please describe your relationship to that person.) 4. During your stay in Kobe so far, who have you communicated with in Japanese most? 5. Do you consider your language partner as your friend? Please explain. 6. Have you made Japanese friends other than your language partner? 7. What aspects of your language partners’ behavior confused or surprised you? 8. What aspect of Japanese students’ language use confused or surprised you? 9. What challenges or difficulties do you feel you have had while communicating with your language partner? 10. What communication strategies do you use when communicating with American students? 11. How do you think a pre-education abroad training can better prepare American students’ social network construction during EdA?

240

Appendix B. Phase 1: Interview Questions with local Japanese students

1. どうしてランゲッジパートナー(RA)として参加しようと思ったのですか。(What was your primary reason for participating in this program as a language partner?) 2. 今回の留学生の中で、一番長い時間一緒に過ごしたのは誰ですか？ (Among the EdA students, who have you spent time with most?) 3. 今回の留学生の中で、一番日本語を話したのは誰ですか？(Among the EdA students, who have you communicated in Japanese most?) 4. ランゲッジパートナーとなった相手の OSU の学生を友達だと思いますか。その理由も述べてください。(Do you consider your language partner as a friend? Please explain.) 5. アメリカ人の学生の振る舞いについて困ったこと、またはびっくりしたことはありましたか。それはどのようなことですか。(What aspects of American students’ behavior confused or surprised you?) 6. アメリカ人の学生の言葉遣いについて困ったこと、またはびっくりしたことはありましたか。それはどのようなことですか。(What aspects of American students’ language use confused or surprised you?) 7. アメリカ人と会話している時に大変だったことや苦労したことなどはありますか。 (What challenges or difficulties do you have had while communicating with American students?) 8. アメリカ人と接する時に、気をつけていることや工夫しているのはどんなことですか。(What communication strategies do you use when communicating with American students?) 9. 最近できた外国人の友達はいますか。その人とはどのようにして友達になりましたか。(Did you make any new foreign friend recently? How did you become friend with them?) 10. アメリカ人の学生が日本人と良い関係を気づくようにするためにした方がいいこと、またはしない方がいいことなどありますか。(What things should American students do or not do in order to be friends with Japanese students.)

241

11. アメリカ人留学生の出発前のトレーニングでどのようなことをしてあげれば、留学中に日本人とよい（友好的、生産的な）人間関係が構築できると思いますか。 (How can pre-EdA training support American students’ social network construction?)

242

Appendix C. Phase 2: Pre-EdA Interview Questions with EdA Exchange Students

1. Which part of Japan will you be staying? 2. When do you leave for Japan? How long is your stay going to be? 3. What is your primary reason for studying abroad in Japan? 4. What are your goals for studying abroad in Japan? 5. What strategies will you use achieve these goals? 6. What do you think about Japan? 7. What do you think about the US? 8. Do you have any concerns about your up-coming study abroad year in Japan? 9. What sort of strategies do you expect to learn in the pre-EdA training?

243

Appendix D. Phase 2: Post-EdA Interview Questions with EdA Exchange Students

1. How was your overall education abroad experience? 2. What did you learn most from this education abroad experience? 3. What had you wanted to learn/experience before EA but feel you did not learn/experience as much as during EdA? 4. How would you describe your personal and social environment where you lived in Japan? 5. In what areas did you struggle during EdA? How did you cope with or overcome those areas of struggle? 6. Who do you consider your closest friends in Japan? How did you meet them? 7. What aspects of your Japanese friends or people’s behavior confused or surprised you? 8. What aspects of your Japanese friends or people’s language use confused or surprised you? 9. What challenges or difficulties do you feel you have had while communicating with your Japanese friends or people? 10. What strategies do you use when communicating with Japanese people? 11. (Only for EdA group) How do you think the pre-education abroad training prepared you in your network construction during EdA?

244

Appendix E. Study Abroad Social Interaction Questionnaire (SASIQ)

245

2017/3/20 Qualtrics Survey Software

Study Abroad Social Interaction Questionnaire (SASIQ)

Your Name (First and Last):

In this questionnaire you will be asked about people you spoke with in Japanese and in English while

participating in study abroad in Japan. Please respond carefully to each of the items based on your

recollections of your experience in Japan. Your best recollections are acceptable.

1. In the boxes below, please write, from memory, the names of friends or acquaintances who you spoke Japanese with or native speakers of Japanese with whom you regularly spoke in English who ﬁt the following description in all respects:

You at least occasionally spoke Japanese to them. You know them well enough to have spent at least some time socializing with them.

If you had more than twenty friends with whom you at least occasionally spoke Japanese, please simply list the ﬁfteen with whom you spoke Japanese most regularly.

To help you think about people you could name, think about people you met at school, in the community, through internships, or people you lived with, as well as people you were introduced to through friends or others.

Person 1 Person 2 Person 3 Person 4 https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 1/12

246

2017/3/20 Qualtrics Survey Software

Person 5 Person 6 Person 7 Person 8 Person 9 Person 10 Person 11 Person 12 Person 13 Person 14 Person 15 Person 16 Person 17 Person 18 Person 19 Person 20

2. Please use the drop-down menus to indicate how often you spoke Japanese with each individual (Japanese Use), how often you spoke English with them (English Use), and how well they spoke English (English Proﬁciency).

Japanese Use English Use English Proﬁciency » Person 1 » Person 2 » Person 3 » Person 4 » Person 5 » Person 6 » Person 7 » Person 8 » Person 9 » Person 10

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 2/12

247

2017/3/20 Qualtrics Survey Software Japanese Use English Use English Proﬁciency » Person 11 » Person 12 » Person 13 » Person 14 » Person 15 » Person 16 » Person 17 » Person 18 » Person 19 » Person 20

Japanese Use » Person 1 » Person 2 » Person 3 » Person 4 » Person 5 » Person 6 » Person 7 » Person 8 » Person 9 » Person 10 » Person 11 » Person 12 » Person 13 » Person 14 » Person 15 » Person 16 » Person 17

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 3/12

248

2017/3/20 Qualtrics Survey Software Japanese Use » Person 18 » Person 19 » Person 20

Using the dropdown menu on the left, indicate the category that best describes how you met each person. In the text box on the

right please elaborate on the details of your meeting. The boxes will fit more than it might appear. Please feel free to write as

much as you need to.

How you met this person Please Elaborate How You Met » Person 1 » Person 2 » Person 3 » Person 4 » Person 5 » Person 6 » Person 7 » Person 8 » Person 9 Person » 10 Person » 11 Person » 12 Person » 13 Person » 14 Person » 15 Person » 16

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 4/12

249

2017/3/20 Qualtrics Survey Software How you met this person Please Elaborate How You Met Person » 17 Person » 18 Person » 19 Person » 20

2. Please use the drop-down menus to answer each question.

What percentage of What percentage of that time did you that time did you spend doing activities spend doing activities On average how many in Japanese? (reading, in English. (speaking, hours did you spend writing, speaking, reading, writing, with this person per listening to music, listening to music, month? watching TV, etc.) watching TV etc.) » Person 1 0 0 0 » Person 2 0 0 0 » Person 3 0 0 0 » Person 4 0 0 0 » Person 5 0 0 0 » Person 6 0 0 0 » Person 7 0 0 0 » Person 8 0 0 0 » Person 9 0 0 0 » Person 10 0 0 0 » Person 11 0 0 0 » Person 12 0 0 0 » Person 13 0 0 0 » Person 14 0 0 0 » Person 15 0 0 0 » Person 16 0 0 0 » Person 17 0 0 0 » Person 18 0 0 0

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 5/12

250

2017/3/20 Qualtrics Survey Software What percentage of What percentage of that time did you that time did you spend doing activities spend doing activities On average how many in Japanese? (reading, in English. (speaking, hours did you spend writing, speaking, reading, writing, with this person per listening to music, listening to music, month? watching TV, etc.) watching TV etc.) » Person 19 0 0 0 » Person 20 0 0 0

For each of the people in your list, please indicate the level of your friendship, ranging from mere acquaintance to very close friend/confidant.

Note that in terms of communication, level of friendship ranges from engaging in occasional friendly exchanges (low on the scale) to sharing one's deepest feelings or asking for advice regarding personal challenges (high on the scale). Refer to the diagram below to help interpret the range.

1 2 3 4 5 6 7 8 » Person 1 » Person 2 » Person 3 » Person 4 » Person 5 » Person 6 » Person 7 » Person 8 » Person 9 » Person 10 » Person 11 » Person 12 » Person 13

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 6/12

251

2017/3/20 Qualtrics Survey Software

1 2 3 4 5 6 7 8 » Person 14 » Person 15 » Person 16 » Person 17 » Person 18 » Person 19 » Person 20

Choose three people from your list above that you marked as being the closest of friends (highest score). Please tell why you think you were able to develop good friendships with these people? What allowed you to move up the scale from acquaintance to friend, etc.

Choose three people from your list above that you marked as being lowest in terms of friendship level. Please tell why you think you were not able to develop stronger friendships with these people? Describe anything that may have inhibited friendships with these people.

What were some obstacles that kept you from speaking Japanese with these people?

What did your study abroad program do to help you make native speaking friends or acquaintances while in Japan?

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 7/12

252

2017/3/20 Qualtrics Survey Software

What more could your study abroad program have done to help you make native speaking friends or acquaintances while in Japan?

What sorts of things did you do to make friends with native Japanese speakers?

SOCIAL GROUPS

There are four parts to this question (A-D).

Part A.

For this item you will help us identify which people know each other and how they know each other by grouping together the people you listed according to where they should know each other from (and possibly where you got to know them). For example, if three of the people are host family members, you would group them together by dragging their names to the "Host Family" box. If four of the people worked at your internship site and knew each other as a result, you would group them together by dragging their names to the "Group 1" box and then giving the box "Group 1" the label "Internship Site" in the blank below. Clubs, community organizations, etc. could also be used as group labels.

If people belong to more than one group, place them in their primary group (the group they are most tightly linked to).

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 8/12

253

2017/3/20 Qualtrics Survey Software After dragging people to their groups, please be sure to deﬁne each group in the text ﬁelds that follow (Part B) so we can understand how the people know each other. If you have more groups than there are boxes, please use the next question (Part C) to describe who these groups are and how they are made up (the people and the group names).

Items Host Family/ Group 1 » Person 1 Roommates » Person 2 » Person 3 » Person 4 » Person 5 » Person 6 » Person 7 Group 2 » Person 8 » Person 9 » Person 10 » Person 11 » Person 12 » Person 13 Group 3 Group 4 » Person 14 » Person 15 » Person 16 » Person 17 » Person 18

» Person 19 Group 5 » Person 20

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 9/12

254

2017/3/20 Qualtrics Survey Software

Part B

Label for Group 1 Label for Group 2 Label for Group 3 Label for Group 4 Label for Group 5

Part C If there were more groups than six (Homestay plus ﬁve others), please list the groups and their members here.

Part D If people belonged to more than one group, please list these people and their additional groups here. (Give each name with that person's additional group or groups.)

https://az1.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview 10/12

255

Appendix F. Language Use Log Instructions

How to download and use the Language Use Apps (Ver.2)

1. Access to the app through the link I shared via Line or email.

2. If you don’t have the AppSheet app on your phone, please download it. (Android: https://play.google.com/store/apps/details?id=x1Trackmaster.x1Tra ckmaster&hl=en) or (iPhone: https://itunes.apple.com/us/app/appsheet/id732548900?mt=8)

3. After downloading AppSheet, please open the app. You will see an icon Practice_YourName. Please use this app to practice your language use until Oct. 23, 2017.

4. The first Language Log Use will start Oct 23, 2017 and ends Oct 29, 2017. Every morning around JST 7am, I will send a link for access to Language Use Log app for the day. For example, you will receive a link for an app named Oct23_YourName in the morning of Oct. 23, 2017. Please use the app to log your language use for October 23, 2017. In total, you will receive 8 apps (including the practice app). If you tap the mark with the three lines at the upper left-hand corner and tap App Gallery, you will see all the apps like the first screenshot on the right.

5. You can enter your language use anytime you want, and it doesn’t have to be in a chronological order. But I recommend that you log your language use as soon as you finish whatever language activity. However, if a frequent language use like reading English/Japanese news or messages on your phone or tablet, you should enter the total amount of time spending on the language use as one entry at the end

256

of the day.

6. In order to enter your language log, please tap the plus button on the lower right- hand corner.

7. You will see the screen on the right. Please ignore the first two sections (i.e. TIME & DATE). You don’t have to do anything about them.

8. Please tap the language (i.e. Japanese, English or Other) you use.

9. And choose the type of language you use (i.e. Interactional or non-interactional language). For example, talking to someone would be an interactional language, but if you are talking to yourself (e.g. practicing for CC or speech etc.), that is non- interactional. Another example is, if you are writing a journal or essay, it’s non- interactional. But if you are typing a message in a chat-box or writing emails, this would be another example of interactional language use.

10. After completing TYPE, please choose type of language (TYPE 1-4), namely, speaking, listening, reading and writing. If you use more than one type of language, please list the most used language type on TYPE 1 to the least on TYPE 4.

11. For the PURPOSE, please choose from Homework, For Fun, Self-study, Work, Class, Club (サークル) and Other.

12. In WHO1-2, if you selected interactional language use in #9 above, please provide whom you are interacted with (i.e. Friend, Professor etc.)

13. In NAME(s), if you selected interactional language use in #9 above, please provide the name of the person whom you interacted with.

14. In HOW LONG (min), please provide the length of time you are engaged in the language use (e.g. 14 min. or 90 min. etc.).

257

15. In DESCRIPTION, please provide a brief description of what you did (e.g. talking about politics in Japan etc.).

16. Finally, please input the time you started the language use. Please make sure to use military time.

17. When you finish, please tap SAVE to save the data.

258

Appendix G. Instructions of pre-EdA training role-play

Situation: You want to join a university club, so you visited a classroom where club meetings are held. You arrived there 15 minutes earlier than the meeting time. When you entered the classroom, you saw a Japanese student who is a member of the club. (a) 自己紹介をする (introduce yourself and ask questions to continue the conversation) (b) 共通の趣味を見つける (find similar interests/hobbies) (c) 次合う約束をし、連絡先を聞く(try to make an appointment to meet up sometime soon. Ask her contact information.)

259