L2 PRAGMATIC DEVELOPMENT THROUGH CONVERSATIONAL INTERACTION: HERITAGE LANGUAGE BACKGROUND AND EXPLICITNESS OF FEEDBACK By Seongmee Ahn A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Second Language Studies - Doctor of Philosophy 2013 ABSTRACT L2 PRAGMATIC DEVELOPMENT THROUGH CONVERSATIONAL INTERACTION: HERITAGE LANGUAGE BACKGROUND AND EXPLICITNESS OF FEEDBACK By Seongmee Ahn This dissertation investigates whether and how learners’ heritage language (HL) background and explicitness of feedback influence second language (L2) pragmatic development of Korean referent honorifics through conversational interaction. Specifically, this study focuses on HL learners versus non-heritage language (NHL) learners and recasts versus metalinguistic correction. Although abundant empirical research has demonstrated that the provision of interactional feedback to learners’ error facilitates their morphosyntactic and lexical development, the role of interactional feedback in learning L2 pragmatics has been hardly explored. In addition, while recent years have seen a fast growing interest in the study of HL learners in second language acquisition, and the differential effects of implicit and explicit feedback have been a prolific area of research, little is known whether these different populations of learners react to these two types of feedback differently. Thus, the study seeks answers to the following research questions: (a) does interactional feedback promote L2 pragmatic development?; (b) how does learners’ HL background affect L2 pragmatic development?; (c) how does explicitness of feedback affect L2 pragmatic development?; and (d) how does explicitness of feedback affect L2 pragmatic development of learners with different HL background? A total of 78 intermediate-level English-speaking learners of Korean as a foreign language at two public universities in the United States participated. The learners were randomly assigned to the following four experimental groups and a control group: (a) HL/Implicit group (n=16); (b) HL/Explicit group (n=15); (c) NHL/Implicit group (n=16); (d) NHL/Explicit group (n=16); and (e) control group (n=15). The linguistic target was Korean referent honorifics in the area of pragmatics. Since pragmatics is highly contextsensitive, the photos of the researcher’s own family and acquaintances in Korea were used to provide as natural and authentic contexts as possible. The photos of individuals of differing age and social status were used to elicit Korean honorifics. In this study, the researcher, who served as a Korean native speaker (NS) interlocutor, and the learners partook in two types of tasks: (a) one-way photo description task and (b) two-way story sequencing task. This study employed a traditional pretest and posttest design to measure treatment effects. In the first session, all the learners completed a background questionnaire, followed by the pretest and the first treatment. In the second session, after the second treatment was conducted, the immediate posttest followed. During the treatment sessions, the NS researcher provided the learner with implicit or explicit feedback on erroneous utterances in Korean. Two weeks after the immediate posttest, the delayed posttest was administered. The results showed that: (a) all the experimental groups showed a significant increase from the pretest to the posttests, contrary to the control group; (b) the NHL learners significantly outperformed the HL learners; (c) explicit feedback was significantly more effective than implicit feedback; and (d) explicit feedback was significantly more effective than implicit feedback for the HL learners, whereas no significant difference was found in the effects of two feedback types for the NHL learners. Also, the NHL/Explicit group showed the strongest development, whereas the HL/Implicit group demonstrated the weakest among the experimental groups. There was no significant difference between the HL/Explicit and the NHL/Implicit groups. The findings are discussed in terms of relative efficacy of implicit and explicit feedback in relation to different nature and learning process between the HL and the NHL learners. Copyright by SEONGMEE AHN 2013 To my parents v ACKNOWLEDGEMENTS My doctoral work and dissertation could not have been completed without support from many. I would first like to acknowledge generous support from Michigan State University, which provided me with everything I needed in order to complete my graduate work. The Second Language Studies Program at Michigan State University provided me with a rich research environment as well as many opportunities and intellectual challenges. My sincere and foremost heartfelt thanks go to Professor Susan Gass. She is the one who led me to the Second Language Studies Program at Michigan State University and to the world of Second Language Acquisition, especially input and interaction research. As my advisor and dissertation committee chair throughout my graduate study, her unwavering support, guidance, encouragement, and influence on me are beyond words. I also appreciate her patience with me. She will always remain as my role model researcher. I would also like to express my gratitude to my dissertation committee members. I feel privileged to have an expert in focus on form research, Professor Shawn Loewen, in my committee. I sincerely thank him for teaching me quantitative research methods as well as corrective feedback research. His insightful comments on the earlier stages of this dissertation and statistical analyses were of great help to me. I will not be able to forget his kindness and generosity. I feel fortunate to have met Professor Charlene Polio during my doctoral study. I learned much from her particularly in the area of second language writing as well as interaction research. She did everything to help me learn by reading my research papers many times at every stage and providing me with insightful comments. Her help will remain unforgettable. I am grateful for Professor Paula Winke. Her encouragement, enthusiasm, and attitude toward vi research, teaching, and life as well as kindness have always impressed me. She has become another role model in my life. I will keep her precious advice in mind. I feel also indebted to Professor Ok-Sook Park. She taught me about the Korean language, spent so much time discussing my research, and answered all my never-ending, trivial questions, even late in the evening, despite her busy schedule. Our discussions helped shape incipient ideas of this dissertation. In addition to my committee members, I would like to express my appreciation to others. Professor K. Seon Jeon at Columbus State University offered me opportunities to work on the research projects on the Korean as a foreign/second language and pragmatic development years ago, which spurred my interest in these topics. Our collaboration and endless discussions on the research have become a spring board as well as a solid base for this dissertation. I also thank her for answering all the detailed questions about my research ideas either by phone or in person at any time, and for having become a good friend of mine. I would also like to mention Dr. Dan Reed at the testing office at Michigan State University and thank him for guiding me to the world of language assessment and teaching me by collaborating on a range of assessment projects at the Center for Language Education And Research (CLEAR) for over five years. I must not forget generous support and understanding from instructors of the Korean language programs: especially Dr. Kijoo Ko and Dr. Minsook Kim at University of California, Berkeley, and Dr. Sang-Seok Yoon and Dr. Hangtae Cho at University of Minnesota. Recruiting a sufficient number of participants was the one of the biggest challenges I faced to complete this dissertation. When I was desperate to find research participants, they allowed me to visit their institutions and helped me as much as they could. It would be impossible for me to forget their help. My appreciation also goes to students who offered to help me by participating in this vii dissertation research in spite of their busy school life. I should also mention my thankfulness to Professor Ok-Sook Park, who allowed me to recruit research participants in her classes at Michigan State University for the pilot study of this dissertation. I should also thank my graduate student fellows and colleagues. They made my life more enjoyable and offered to help me as well. They are too many to list here, but I particularly thank Junkyu Lee, Andy Xiaoqing Chen, and Soo Hyun Kim for their generous help and encouragement while in graduate school. This dissertation would not be possible without support of my family. First of all, I feel the deepest appreciation to my parents. My father, Sang Dal Ahn, inspired my curiosity and thirsts toward the new world and knowledge and encouraged me to continue to pursue my dreams. His spirit and philosophy helped shape what and who I am and made me study abroad, far away from home for many years. My mother, Jung Sook Son, was my biggest support during the ups and downs of graduate school. She did everything she could to make this dissertation complete, including taking care of my two-year-old daughter. I would like to dedicate this dissertation to my parents. For my brother, Sungwook Ahn and my sister, Sunghee Ahn -- thank you for believing in me and cheering me up. They are perpetual supports for me in life. Thanks to my little daughter, Natalia Gustin, for growing up so well, and sorry for not being with you as much as I wished while working on this dissertation. Finally, I would like to show my sincerest appreciation to my husband, Rainer Gustin, for everything he has done. Thank you for providing technical support for my computer, taking care of our daughter, understanding my long, frequent absences from home, and supporting me no matter what I choose to do. But for your understanding and support, I simply would not have been able to get anything done. viii TABLE OF CONTENTS LIST OF TABLES ..................................................................................................................... xi LIST OF FIGURES ................................................................................................................... xiii LIST OF ABBREVIATIONS .................................................................................................... xiv CHAPTER 1 INTRODUCTION ............................................................................................... 1 CHAPTER 2 LITERATURE REVIEW .................................................................................... Interactionist Approach to SLA ........................................................................................... Theoretical Constructs ................................................................................................... Factors Influencing L2 Development through Conversational Interaction ................... Pragmatics as a Linguistic Target ........................................................................................ Pragmatic vs. Morphosyntactic and Lexical/Semantic Development ........................... Korean Honorifics as a Pragmatic Target ...................................................................... Korean Referent Honorifics ........................................................................................... Heritage Language Background .......................................................................................... Definitions of Heritage Language Learners ................................................................... Characteristics of Heritage Language Acquisition ........................................................ Heritage vs. Non-heritage Language Learners .............................................................. Explicitness of Interactional Feedback/Instruction .............................................................. Feedback Classification and Implicit vs. Explicit Dimension ....................................... Recasts vs. Metalinguistic Feedback ............................................................................. Implicit vs. Explicit Instruction in Instructed SLA and L2 Pragmatics ......................... Research Questions and Hypotheses ................................................................................... 11 11 11 17 19 19 22 24 33 33 37 40 47 47 50 56 60 CHAPTER 3 METHOD ............................................................................................................ Participants ........................................................................................................................... Heritage Language Learners .......................................................................................... Non-heritage Language Learners ................................................................................... Operationalizations .............................................................................................................. Explicit Feedback........................................................................................................... Implicit Feedback........................................................................................................... Heritage Language Learners .......................................................................................... Non-heritage Language Learners ................................................................................... Linguistic Targets ................................................................................................................ Subject Referent Honorifics ........................................................................................... Object Referent Honorifics ............................................................................................ Instruments ........................................................................................................................... Pretest, Immediate Posttest, and Delayed Posttest Tasks and Materials ....................... Treatment Tasks and Materials ...................................................................................... Design and Procedure .......................................................................................................... 63 63 64 66 66 66 67 68 69 69 70 71 72 74 74 76 ix Research Design............................................................................................................. Research Procedure ........................................................................................................ Coding and Scoring.............................................................................................................. Interrater Reliability ....................................................................................................... Data Analyses ...................................................................................................................... 76 77 79 80 81 CHAPTER 4 RESULTS ............................................................................................................ 86 Role of Interactional Feedback in Pragmatic Development ................................................ 86 Subject Honorifics ......................................................................................................... 86 Object Honorifics ........................................................................................................... 89 Heritage vs. Non-heritage Language Learners .................................................................... 92 Subject Honorifics ......................................................................................................... 92 Object Honorifics ........................................................................................................... 94 Implicit vs. Explicit Feedback ............................................................................................. 97 Subject Honorifics ......................................................................................................... 97 Object Honorifics ........................................................................................................... 100 Heritage vs. Non-heritage Language Learners in relation to Implicit vs. Explicit Feedback 103 Subject Honorifics ......................................................................................................... 103 Object Honorifics ........................................................................................................... 111 CHAPTER 5 DISCUSSION ...................................................................................................... 119 Role of Interactional Feedback in Pragmatic Development ................................................ 121 Heritage vs. Non-heritage Language Learners .................................................................... 125 Implicit vs. Explicit Feedback ............................................................................................. 139 Heritage vs. Non-heritage Language Learners in relation to Implicit vs. Explicit Feedback143 Pedagogical Implications ..................................................................................................... 150 CHAPTER 6 CONCLUSION.................................................................................................... 153 Limitations and Suggestions for Future Research ............................................................... 155 APPENDICES ........................................................................................................................... Appendix A: Language Background Questionnaire ............................................................ Appendix B: Table B-1. Participant Background Information ............................................ Appendix C: Table C-1. Participants’ Self-rated Comfort with Korean Language Skills... Appendix D: Table D-1. Heritage Language Learners' Childhood and Current Interaction in Korean .............................................................................................................................. Appendix E: Table E-1. Participants’ Uses and Exposure of Korean ................................. Appendix F: Figure F-1. Sample Testing Material .............................................................. Appendix G: Figure G-1. Sample Treatment Material ........................................................ Appendix H: Task Instructions for Photo Description Task ................................................ Appendix I: Task Instructions for Story Sequencing Task .................................................. Appendix J: Sample Treatment Sessions ............................................................................. Appendix K: Table K-1. Average Time for Each Experimental Group in Each Task ........ Appendix L: Coding and Scoring Guidelines ...................................................................... 159 160 165 171 173 176 178 179 180 181 182 233 234 REFERENCES .......................................................................................................................... 245 x LIST OF TABLES Table 1. Korean Speech Styles by Sentence Types .................................................................. 25 Table 2. Examples of Korean Speech Styles for Declarative Forms ......................................... 26 Table 3. Characteristics of L1, L2, and Heritage Language Acquisition ................................. 38 Table 4. Research Design ......................................................................................................... 76 Table 5. Research Procedure ..................................................................................................... 78 Table 6. Descriptive Statistics for Experimental and Control Groups in Subject Honorifics .. 86 Table 7. ANOVA Table for Experimental and Control Groups in Subject Honorifics across Tests .......................................................................................................................................... 88 Table 8. Descriptive Statistics for Experimental and Control Groups in Object Honorifics .... 89 Table 9. ANOVA Table for Experimental and Control Groups in Object Honorifics across Tests .......................................................................................................................................... 91 Table 10. Descriptive Statistics for HL and NHL Learner Groups in Subject Honorifics ....... 92 Table 11. ANOVA Table for HL and NHL Learner Groups in Subject Honorifics across Tests 94 Table 12. Descriptive Statistics for HL and NHL Learner Groups in Object Honorifics ........ 95 Table 13. ANOVA Table for HL and NHL Learner Groups in Object Honorifics across Tests 97 Table 14. Descriptive Statistics for Implicit and Explicit Feedback Groups in Subject Honorifics ................................................................................................................................. 98 Table 15. ANOVA Table for Implicit and Explicit Feedback Groups in Subject Honorifics across Tests ............................................................................................................................... 100 Table 16. Descriptive Statistics for Implicit and Explicit Feedback Groups in Object Honorifics ................................................................................................................................. 101 Table 17. ANOVA Table for Implicit and Explicit Feedback Groups in Object Honorifics across Tests ............................................................................................................................... 103 Table 18. Average Frequency of Error Occurrences and Feedback Moves per Learner in Each Experimental Group in Subject Honorifics ............................................................................... 105 xi Table 19. Descriptive Statistics for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Subject Honorifics ................................................................................ 106 Table 20. ANOVA Table for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Subject Honorifics across Tests .................................................................. 109 Table 21. Post Hoc Tests in Subject Honorifics ....................................................................... 110 Table 22. Average Frequency of Errors Occurrences and Feedback Moves per Learner Each Experimental Group in Object Honorifics ................................................................................ 112 Table 23. Descriptive Statistics for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Object Honorifics ................................................................................. 113 Table 24. ANOVA Table for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Object Honorifics across Tests ................................................................... 116 Table 25. Post Hoc Tests in Object Honorifics ......................................................................... 117 Table B-1. Participant Background Information ...................................................................... 165 Table C-1. Participants’ Self-rated Comfort with Korean Language Skills ............................. 171 Table D-1. Heritage Language Learners' Childhood and Current Interaction in Korean ......... 173 Table E-1. Participants’ Uses and Exposure of Korean ............................................................ 176 Table K-1. Average Time for Each Experimental Group in Each Task ................................... 233 xii LIST OF FIGURES Figure 1. Korean Honorific System .......................................................................................... 24 Figure 2. Subject Honorifics ..................................................................................................... 27 Figure 3. Object Honorifics ...................................................................................................... 27 Figure 4. The Bilingual Continuum .......................................................................................... 36 Figure 5. Classification of Corrective Feedback ....................................................................... 50 Figure 6. Visual Representations for Experimental and Control Groups in Subject Honorifics 87 Figure 7. Visual Representations for Experimental and Control Groups in Object Honorifics 90 Figure 8. Visual Representations for HL and NHL Learner Groups in Subject Honorifics ..... 93 Figure 9. Visual Representations for HL and NHL Learner Groups in Object Honorifics ...... 96 Figure 10. Visual Representations for Implicit and Explicit Feedback Groups in Subject Honorifics .................................................................................................................... 99 Figure 11. Visual Representations for Implicit and Explicit Feedback Groups in Object Honorifics ................................................................................................................................. 102 Figure 12. Visual Representations for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Subject Honorifics ................................................................................ 108 Figure 13. Visual Representations for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Object Honorifics ................................................................................. 115 Figure F-1. Sample Testing Material. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation. .................... 178 Figure G-1. Sample Treatment Material .................................................................................... 179 xiii LIST OF ABBREVIATIONS ADV = Adverb DO = Direct object case marker EX = Exclamation H = Honorific form Hum = Humble form IO = Indirect object case marker LOC = Locative case marker OBJ = object referent honorifics PAST = Past tense POSS = Possessive PROG = Progressive S = Subject case marker SH = Special honorific form SUB = subject referent honorifics V = Verbal morpheme xiv CHAPTER 1 INTRODUCTION This study aims to contribute to expanding the current boundary and foci of interaction research that have been prevalent thus far. Much empirical evidence in early interaction research has proven the facilitative role of conversational interaction. Furthermore, since Lyster and Ranta’s (1997) influential work on the identification of corrective feedback patterns in French immersion classrooms in Canada, the interest in the effectiveness of different types of feedback has been revived, which generated a great number of research studies on this topic. More recently, as there has been increasing research interest in investigating learner-internal and learner-external factors that may constrain the efficacy of different feedback types, many interaction studies have been centered on these issues. While more studies attempt to explore factors, such as language aptitude (e.g., Sheen, 2007; Trofimovich, Ammar, & Gatbonton, 2007) and working memory (e.g., Gass, Behney, & Uzum, in press; Goo, 2012; Mackey, Philp, Egi, Fujii, & Tatsumi, 2002; Mackey & Sachs, 2012; Sagarra, 2007) as part of learner-internal factors, some factors, such as learners’ past language learning experience (e.g., Gass & Lewis, 2007; Y. Han, 2010) and their pragmatic development (e.g., Fukuya & Hill, 2005; K. Jeon, 2007; Soler, 2002) as part of learner-external factors, have been hardly examined. Hence, there is still great necessity to continue to expand the current scope of investigation. To this end, the present study investigates factors that may influence second language (L2) development through conversational interaction. In particular, it examines whether and how learners’ heritage language (HL) background and explicitness of feedback affect L2 pragmatic development in the area of Korean honorifics. The role of conversational interaction as a facilitator in L2 development has been confirmed by abundant empirical evidence in both laboratory and classroom settings for the past 1 few decades. Interaction researchers are moving their interest from whether interaction promotes learning towards how interaction leads to learning. To enhance the understanding of this interaction and learning mechanism, many researchers have expressed a need to expand the current scope of research issues and areas in interaction and learning framework (Mackey, 2007; Mackey & Gass, 2006). Within the context of the interactionist approach, myriad factors that may influence the relationship between interaction and learning have become a focus of interest among researchers. For example, the factors include areas of language (e.g., morphology, syntax, lexis, phonology, and pragmatics), characteristics of particular linguistic targets (e.g., perceptual salience, and simple vs. complex features), types of feedback (e.g., implicit vs. explicit and input-providing vs. output-pushing), types of knowledge measured (e.g., implicit vs. explicit), past language learning experiences (e.g., HL learning), instructional settings (e.g., second vs. foreign language classroom), individual differences (e.g., working memory, language aptitude, and motivation) and so forth. Researchers are also interested if interactional processes work in concert or individually in relation to these factors. In addition, Schmidt (1990, 1993, 1994, 1998, 2001) emphasized learners’ attentional resource allocation for noticing mismatches between learners’ nontargetlike forms and their interlocutors’ targetlike forms. Many other researchers have suggested the importance of noticing as a mediator between interaction and learning (Gass, 1997; Long, 1996; Mackey et al., 2002). In consideration of a strong interest in this line of research in recent years, much more empirical research in these areas is needed. Moreover, the majority of interaction research has centered on the L2 development of learners of Indo-European languages, such as English and 2 Spanish, with only a few studies on conversation interaction in Korean as a second/foreign language (e.g., Ahn, 2012; K. Jeon, 2007; Kang, 2009, 2010a, 2010b). Korean is distinguished from these Indo-European languages in many regards. For purposes of this dissertation, Korean has its own unique pragmatic features, including honorifics. Thus, it would be interesting to see how interaction influences learning of Korean, a language typologically distant from IndoEuropean languages. Since Long’s Interaction Hypothesis (1981, 1983, 1996), the facilitative role of conversational interaction in second language acquisition (SLA) has been supported by abundant empirical evidence. However, the majority of research to support the Interaction Hypothesis has heavily focused on learners’ morphosyntactic and lexical development, and L2 pragmatics has been a rarely explored dimension (Loewen, 2012; Mackey, 2007). Although there has been a paucity of studies that investigated the effects of interactional feedback in the field of L2 pragmatics, the following studies are worth mentioning. Both Soler (2002) and Fukuya and Hill (2005) explored the effects of interaction on the development of learners’ pragmatic competence of requests, whereas K. Jeon (2007) studied the impact of interactional feedback on the learning of Korean honorifics. The first two studies involve speech acts that concern the interlocutor’s strategic language use to achieve his or her goal in social interaction and are not directly related to the topic of this dissertation. K. Jeon (2007) examined the role of recasts in L2 pragmatic development with a focus on subject-verb agreement in Korean subject referent honorifics. It was reported that the role of recasts in honorific development was not as facilitative as in the development of other linguistic targets in Korean, such as relative clauses, nouns, and verbs. K. Jeon indicated that this may have been because pragmatics is one of the most difficult aspects of L2 learning (Gass, Behney, & 3 Plonsky, 2013; Gass & Selinker, 2008) and may require more than just implicit feedback. She further argued that since honorifics have a low communicative value, learners may not have felt a need to attend to honorific forms. Alternatively, as the learners spoke English as their primary language, which does not have an honorific system, they may have had difficulties using honorifics in Korean. As K. Jeon acknowledged, however, there was a methodological issue raised in this research. The communicative tasks were based on hypothetical situations with fictional characters in which learners might have spoken differently compared to real-life situations. Because the study dealt with honorifics, an aspect of language that is highly dependent on interlocutor characteristics, the use of fictional characters may not have been the most appropriate means to ascertain authentic use. Thus, given that pragmatics, including honorifics, is highly context-sensitive, sufficient contextual information based on authentic situations, if possible, and higher task essentialness should be supplied in developing communicative tasks in order to elicit learners’ pragmatic use in real life. K. Jeon found that many participants in her study often appeared to avoid the use of honorifics completely or use it inconsistently, although feedback was given to their nontargetlike form in a consistent way. It is assumed that because the relative communicative values of the honorific target may be lower than those of other linguistic targets, such as relative clauses, nouns, and verbs, the honorific form may be less salient and less noticeable to the learners. This dissertation is built upon K. Jeon’s work and intends to extend this line of research by incorporating issues brought up from her research. In addition, Montrul (2008b, 2010a) pinpointed the problem that specialized vocabulary, forms of address, and honorifics in East Asian languages are absent or are not completely acquired by HL learners. She further asserted that Korean honorifics are a major pedagogical 4 concern in teaching Korean as a HL. In the same vein, HL learners tend to lack late-acquired aspects of language, such as politeness markers, and sociolinguistic and pragmatic abilities to perform in an adultlike conversation or academic setting (Krashen, 1998; Montrul, 2008b; Valdés & Geoffrion-Vinci, 1998). Therefore, the present study explores how Korean honorifics as a pragmatic target are developed in the context of HL learning. Recent years have seen a rapidly growing interest in the study of HL acquisition in SLA (Gass & Selinker, 2008; Montrul, 2008b). However, the vast majority of interaction research has involved English as a Second Language (ESL)/ English as a Foreign Language (EFL) and other second/foreign languages, such as Spanish and Japanese, and focused on traditional L2 learners’ linguistic development. There has been little research on other languages (Loewen, 2012), particularly less commonly taught languages (LCTLs), such as Korean, despite the fact that the majority of learners in the LCTL classrooms have HL background (Gambhir, 2001; King, 1998). Thus, little is known about how this learner population uses learning opportunities provided through interaction. A HL learner generally refers to “a language student who is raised in a home where a non-English language is spoken, who speaks or at least understands the language, and who is to some degree bilingual in that language and in English” (Valdés, 2001a, p. 38). There is general agreement that the nature of L2 learning of HL learners differs from that of traditional L2 learners (Campbell & Rosenthal, 2000; Gass, 2007; Valdés, 1995, 2001b). Montrul (2010a) stated that HL learners are typically exposed to ample HL aural input in a naturalistic setting during childhood. Thus, they are inclined to be fluent and possess a relatively good control of early-acquired linguistic features, such as phonology and some vocabulary, in their HL. However, once the majority language becomes dominant upon entering school, they experience reduced HL input, and their HL learning becomes interrupted and incomplete. As a 5 result, fossilization typically occurs. This explicates why the HL learners are subject to lack grammatical accuracy and late-acquired linguistic features, including age- and contextappropriate sociolinguistic registers. With regard to processing strategies, the HL learners tend to use top-down processing strategies (i.e., macro approach) that draw on their existing linguistic knowledge and contextual information (Celce-Murcia & Olshtain, 2000; Z. Han, 2004; H. H. Kim, 2005). They tend to perform language learning tasks globally, on a macro level, and negotiate meaning in various situations. This may be linked to their inherent meaning-based comprehension (i.e., semantic processing) strategies. On the contrary, non-heritage language (NHL) learners resort to bottom-up processing strategies (i.e., micro approach) and attend to subelements, such as grammatical categories, vocabulary, and spelling to be able to perform language learning tasks (Kagan, 2005). Due to such differences as the above, it is apparent that the HL learners are in need of a different pedagogy from the NHL learners (Montrul, 2008, 2010a). Nonetheless, there are only a handful of studies available directly looking at linguistic differences between the HL learners and the NHL learners (e.g., Ke, 1998; H. H. Kim, 2005, 2008; Montrul 2002, 2004; O’Grady, M. Lee, & Choo, 2001; Polinsky, 1995, 2000, 2008). In particular, an inquiry into how the HL learners acquire L2 as opposed to the NHL learners has hardly been investigated within the interactionist approach. Gass and Lewis (2007) are the first researchers who investigated this issue within the interactionist framework and examined to what extent L2 learning process of HL learners differs from that of NHL learners. They found subtle but interesting differences between the two groups of learners. They claimed that compared to the NHL learners, the HL learners focused more on semantic issues than on linguistic form and often perceived morphosyntactic feedback as semantic feedback. The HL learners were less 6 aware of and paid less attention to morphosyntactic feedback. However, because types of feedback were not part of the investigation in this study, it was not possible to know how different feedback types affect L2 learning of these two different groups. As Gass and Lewis clearly indicated, this is an area for future research, and this dissertation aims to explore this issue. Moreover, K. Jeon (2007) reported a similar pattern and confirmed the finding of Gass and Lewis: the HL learners seemed to focus on semantics and often disregarded morphosyntactic features of honorifics, which may have contributed to the low accuracy gains in honorifics. Because the majority of the participants in K. Jeon’s study were HL learners, it was not possible to make a direct comparison between the HL and the NHL learners. It is reasonable to assume that the inclusion of the NHL learners in the present study may bring different results. There is a recent doctoral dissertation that explored the relative efficacy of two feedback types, such as recasts and metalinguistic feedback, in the learning of Mandarin classifiers for Chinese HL and NHL learners (Y. Han, 2010). The findings indicated that: (a) there was no statistically significant difference between the HL and the NHL learners in the same feedback condition; (b) neither feedback type nor language background influenced the findings on the oral imitation test; and (c) explicit feedback was more effective than implicit feedback only for the NHL learners only on the written cloze test. Although Chinese classifiers as part of lexical domain were a linguistic focus, metalinguistic feedback without the provision of a correct form was employed in Y. Han’s study. The metalinguistic feedback may have been sufficiently explicit to draw the learners’ attention to problematic areas. However, as Li (2010b) indicated, it is still possible that the learners may not have been able to produce the correct classifier form because it did not exist in their interlanguage. If explicit correction including the correct form had been provided, the learners might have been more likely to be successful in learning Chinese 7 classifiers. Therefore, lack of significant differences may be partly ascribed to the feedback operationalization (i.e., metalinguistc feedback vs. metalinguistic correction) in her study. This suggests that since different studies can produce different results, more research should be conducted using different research designs, languages, linguistic targets, operationalizations, learner populations, and so on, in order to generalize research findings and thus advance our knowledge in SLA. Although Gass and Lewis (2007) initiated this line of research, more attention should be paid to the nature and process of HL acquisition within the interactionist framework. Therefore, this dissertation intends to contribute to extending the current understanding of SLA by investigating a typologically different language, namely, Korean, and to suggest an appropriate pedagogy that meets special demands of learners of differing language background, namely, HL learners. It is needless to say that the relative efficacy of implicit and explicit feedback has been a highly active area in interaction research (e.g., Carroll & Swain, 1993; R. Ellis, Loewen, Erlam, 2006; Lyster, 1998). Likewise, the notion of explicitness has also drawn much attention in instructed SLA and L2 pragmatics research; however, research on pragmatics has centered on the effects of implicit and explicit “instruction,” which is mostly based on quasi-experimental classroom instruction research, not implicit and explicit “feedback” per se in the context of controlled laboratories. Currently, to my knowledge, there is no research available looking into differential effects of implicit and explicit feedback in L2 pragmatic development through interaction, and this dissertation attempts to explore this issue. Due to a multitude of factors (e.g., linguistic target, salience, research context, feedback operationalization, task type, and proficiency) that constrain the comparative effectiveness between the two feedback types, mixed results have been reported regarding the relative effects 8 of implicit and explicit feedback. For example, while the efficacy of recasts did not appear to be strong in classrooms (e.g., Lyster, 1998a, Lyster & Ranta, 1997; Panova & Lyster, 2002), they seemed to be as effective as explicit feedback in laboratories in some studies (e.g., Kang, 2009, 2010a, 2010b; H. Kim & Mathes, 2001; Loewen & Nabei, 2007). In K. Jeon’s (2007) research, only the role of implicit feedback, mainly in the form of recasts, was investigated; thus, it is unknown if explicit feedback is more effective than implicit feedback to draw learners’ attention to politeness markers in Korean honorifics. Hence, this study examines the relative efficacy of these two different types of feedback in L2 pragmatic development. While meta-analyses indicated that explicit instruction may be more effective than implicit instruction in instructed SLA (Norris & Ortega, 2000; Spada & Tomita, 2010), these studies were based on learners’ grammatical development. E. Jeon and Kaya (2006) undertook a meta-analysis study in L2 pragmatic development and found that although explicit instruction was clearly more effective than implicit instruction, there was no statistically significant difference between implicit and explicit feedback due to lack of appropriate research. However, literature indicated that explicit instruction is generally more effective than implicit instruction in the learning of L2 pragmatics (e.g., House, 1996; Takahashi, 2001; Tateyama, Kasper, Mui, Tay, & Thananart, 1997). Thus far, in spite of much progress made in interaction research, research areas that need to be further explored in future investigation have been identified. To the best of my knowledge, there is no research probing the impact of interactional feedback on the pragmatic development of learners with differing HL background. Moreover, while comparative efficacy of implicit and explicit feedback and instruction has been actively investigated in SLA, it is unknown how these types of feedback interact with learners’ different HL background in the pragmatic development. 9 Therefore, this study attempts to seek out the answers to this question. This dissertation includes the following sections. Chapter 2 reviews previous literature and theoretical foundations according to each variable relevant to the current investigation. Chapter 3 details the particular research methodology used in this study, including participant information, operationalizations, linguistic targets, tasks and materials used for testing and treatments, research design and procedure, and coding and scoring guidelines. Chapter 4 presents the results of the present study, and Chapter 5 discusses the results with reference to prior research and theories and suggests pedagogical implications. Chapter 6 draws a conclusion, indicates limitations of the study, and suggests future research. 10 CHAPTER 2 LITERATURE REVIEW This chapter presents an overview of theoretical constructs and previous research concerning main variables pertaining to this study. More specifically, the following areas are reviewed: a) interactionist approach to SLA; b) pragmatics and Korean honorifics as a linguistic target; c) learners’ HL background; and d) explicitness of interactional feedback/instruction. The justifications to conduct this investigation are also discussed along with research questions and hypotheses. Interactionist Approach to SLA Theoretical Constructs All L2 theorists agree that input (e.g., positive evidence) is an essential condition in SLA; however, their opinions differ with respect to the role of interactional feedback (e.g., negative evidence). Although interactional feedback may make positive evidence available, the primary value of its efficacy lies in the provision of negative evidence. Gass (1997) indicated that there are two sorts of evidence available for language learning. Positive evidence refers to language input or model and includes “the set of well-formed sentences to which learners are exposed” (p. 36), whereas negative evidence refers to corrective feedback and contains “type of information that is provided to learners concerning the incorrectness of an utterance” (p. 37). The relative magnitude of these two kinds of evidence is in dispute in SLA. Central to this argument is whether positive evidence is the only necessary condition and/or whether both positive and negative evidence are required for L2 learning. Differing perspectives to the role of interactional feedback exist in SLA. Krashen (1982) claimed that the presence of comprehensible input is sufficient for language acquisition. Schwartz (1993) contended that feedback can make superficial and temporary changes to 11 learners’ performance, but does not contribute to overall underlying competence. Furthermore, Truscott (1999) strongly maintained that feedback is harmful to learners’ development and thus should be abandoned. This nativist position is based on the grounds that L2 acquisition is equivalent to First Language (L1) acquisition. It is argued that children are born with an innate language faculty, Universal Grammar (UG), which enables them to acquire their L1 successfully. Albeit the absence of negative evidence, children are able to learn an intricate set of abstractions through UG. As a result, children can ultimately reach a full control and mastery of their L1 by mere exposure to abundant language input available in a naturalistic setting. In a similar fashion, positive evidence alone is anticipated to lead adults to successful and complete learning outcome and thus is sufficient for L2 acquisition to occur. Nonetheless, the above nativist perspective may not explain L2 learners’ failure in the development of satisfactory production skills. To exemplify this, I consider the French immersion program in Canada which is regarded as one of the most successful language programs. In this program, English-speaking children received some or all of their instruction in French and an ample amount of French language input for many years. It turned out that French listening and reading comprehension of the French immersion students was superior, but their production abilities in speaking and writing were far lower than expected due to grammatical inaccuracy and sociolinguistic inappropriateness. It is attributed to the fact that although these students received plenty of positive evidence, neither did they produce an adequate amount of French language output, nor did their teachers push them to do it in a grammatically accurate and sociolinguistically appropriate way. This unbalanced development of language skills raised doubts about the nativist view. Swain (1985) argued that Krashen’s comprehensible input (1982) alone is not sufficient and that both comprehensible input and comprehensible output are of 12 necessity to reach a full mastery of the target language. Regarding the role of comprehensible output, Swain further claimed that when learners are involved in communication difficulties, they are pushed to make their language output more precise, coherent, and appropriate with a view to delivering their intended message successfully. Interactionists view input as necessary for L1 and L2 acquisition, but as insufficient for adults to learn L2, especially to achieve nativelike proficiency. Therefore, from this interactionist perspective, both positive and negative evidence are necessary for L2 acquisition. Long’s updated version of the Interaction Hypothesis (1996) clearly points to the facilitative role of interaction, which contains negative evidence: negotiation for meaning, and especially negotiation work that triggers interactional adjustments by the NS or more competent interlocutor, facilitates acquisition because it connects input, internal learner capacities, particularly selective attention, and output in productive ways. (pp. 451-452) As such, negotiation for meaning between the learner and the NS during conversational interaction is more than just a way of communication in the interactionist paradigm. Rather, negotiation for meaning, which often includes negative evidence, ultimately promotes L2 acquisition. An interactional process comprises input, interaction, feedback, noticing, and output. As addressed before, Krashen (1982, 1985) argued in his Input Hypothesis that language input, which serves as positive evidence or model, is an essential component for language acquisition to occur. This is based on the view that adults’ L2 acquisition is driven by a sufficient amount of comprehensible input, and input is the sole condition required for SLA. He proposed “i+1” as an optimal condition for comprehensive input: “i ” represents a current level of learner development, 13 whereas “1” indicates a slightly more advanced level that goes beyond the learner’s present developmental stage. Although this Input Hypothesis has been criticized by some researchers due to unspecificity of 1, it played a considerable role to set the ground to continue to discover the relationship between input and interaction. Although various sorts of interactional feedback exist during interaction, Gass and Lewis (2007) represented two types of feedback: negotiation and recasts. Gass (1997) referred to negotiation as “communication in which participants’ attention is focused on resolving a communication problem as opposed to communication in which there is a free-flowing exchange of information” (p. 107). Although researchers employ slightly different definitions of recasts due to the complexities, recasts are generally defined as correct restatement of a learner’s erroneous production without modification of his or her intended message and regarded as implicit negative feedback. Recasts afford positive evidence, and if the corrective nature of recasts is noticed by the learner despite the ambiguity of such feedback, negative evidence is also available. Conversational interaction, including negotiation for meaning, recasts, and feedback as negative evidence, also plays an important role in learning. While partaking in conversational interaction, learners receive feedback from their native speaker (NS) and/or nonnative speaker (NNS) interlocutors, which provides an opportunity for learners to notice the gap between their erroneous form and the interlocutors’ correct form. The learners may attend to problematic areas and, if so, make subsequent modifications in their interlanguage. Then, learning is likely to occur. Long (1996) in his Interaction Hypothesis indicated the important role of attention in interaction and learning relationship. Attention, as one of the crucial mechanisms in this process, is accomplished partly through negotiation (Gass, 1997) and is assumed to mediate between input 14 and learning (or intake)” (Gass & Mackey, 2007b). Notwithstanding the significant role as a way of creating new knowledge and modifying learners’ interlanguage, attention has been under debate among SLA researchers. Tomlin and Villa (1994) proposed three components of attention: alertness (readiness to deal with incoming stimuli or data), orientation (directing attentional resources to particular information, excluding other information), and detection (cognitive registration of sensory stimuli). They claimed that it is only detection that is necessary for learning, and alertness and orientation only serve to help detection occur. The further argument was made that detection does not involve awareness. Nonetheless, it is generally accepted in SLA that awareness is required for deep processing and learning (Leow & Bowles, 2005). According to Schmidt’s (1990, 1993, 1994) Noticing Hypothesis, noticing, equivalent to apperception (Gass, 1988, 1997), is essential to learning, and no learning occurs without attention. In his model, therefore, noticing entails awareness, and all L2 learning is conscious because input does not become intake for learning unless it is noticed (Robinson, Mackey, Gass, & Schmidt, 2012). Yet, this strong version of noticing was criticized due to the evidence of unattentional learning (Gass, 1997). Schmidt (1998, 2001) later modified his strong version of noticing, indicating that learning without conscious awareness may occur, but it may not play a significant role in learning. In addition, he proposed that awareness can be divided into two levels: understanding and noticing. He noted that understanding as a higher level of awareness, which includes an ability to analyze, compare, and test hypothesis, leads to learning, whereas noticing as a lower level of awareness only leads to intake. Robinson (1995, 1996) supported both Tomlin and Villa’s and Schmidt’s (1990, 1993, 1994) approaches of attention and incorporated them to his model. He considered detection as an initial stage of the process and noticing as a later stage. He defined noticing as “detection plus rehearsal in short-term memory, 15 prior to encoding in long-term memory” (Robinson, 1995, p. 296). Generally, while limited, selective attention and the beneficial effects of attention are widely accepted in SLA, but the role of awareness is still under debate. In addition to attention, another component that is believed to benefit L2 learning during interaction involves output, that is, the language that learners produce. Once learners receive feedback from their interlocutors, the learners are anticipated to incorporate the feedback and produce modified output. The significance of output is pertinent to students’ flawed production skills in the French immersion program discussed earlier. Lack of speaking and writing abilities of the French immersion students is attributed to their limited opportunities for language production. It was found that the students in the program spoke more in their L1, English, while they talked much less in French. Besides, they were not pushed to produce language output in a way that is accurate and appropriate by their teachers. Hence, having drawn attention to the beneficial role of output in SLA, Swain (1995) proposed the Output Hypothesis and noted that: Output may stimulate learners to move from the semantic, open-ended nondeterministic, strategic processing prevalent in comprehension to the complete grammatical processing needed for accurate production. Output, thus, would seem to have a potentially significant role in the development of syntax and morphology. (p. 128) That is, the important role of output lies in the notion that output provides learners with opportunities to switch from semantic processing for language comprehension to syntactic processing for language production. In addition, Swain (2005) presented three functions of output. One of the functions of output relates noticing/triggering that learners may become aware of their linguistic problems, while attempting to express their intended messages. Another function concerns that output affords a forum for hypothesis testing. In other words, output 16 supplies learners with a “trial run” to test their hypothesis. The metalinguistic (reflective) function involves that output stimulates learners to reflect on others’ or their original language production. Gass and Mackey (2007b) also noted that not only does output push learners to produce language in a more targetlike way, but it also promotes automatic and fluent language use. So far, in order to understand theoretical underpinnings behind the current study, theoretical evidence showing that interaction facilitates learning has been demonstrated. More specifically, the components that comprise interaction and learning mechanism, such as input, negotiation, feedback, recasts, noticing, attention, and output, have been reviewed. Subsequently, potential factors that may play roles in interaction-driven learning are discussed with a view to answering the questions of why and how this facilitation occurs. Factors Influencing L2 Development through Conversational Interaction In early interaction research in the 1980s, describing how NNSs interact with other NNSs and/or NSs in terms of patterns, frequencies, and functions of interaction was a main focus of the research. This included illustrating negotiation, such as confirmation checks, clarification requests, and comprehension checks (Long, 1983). Afterwards, the conversation and comprehension relationship was researched from 1980s to early 1990s. Finally, in the mid-1990s, researchers became more interested in the links between interaction and learning by providing data as empirical evidence for the interactionist position (Gass, 2003; Mackey & Gass, 2006). Since the early 1980s, Long’s Interaction Hypothesis has been extensively tested and empirically verified. Interaction facilitates L2 learning in laboratory as well as classroom contexts (e.g., Ishida, 2004; Iwashita, 2003; Leeman, 2003; Long, Inagaki, & Ortega, 1998; Mackey, 1999, 2007; Mackey & Oliver, 2002; Mackey & Philp, 1998; Philp, 2003). Given the enormous 17 progress within the interaction research, Gass and Mackey (2006, 2007b) suggested that it may be an appropriate time to change the term “hypothesis” and refer it to as the Interaction Approach because interaction “is moving towards the status of a theory in the sense that it also attempts to explain why interaction and learning can be linked, using cognitive concepts derived from psychology, such as noticing, working memory, and attention” (Gass & Mackey, 2007b, p. 176). With the advent of the new interactionist paradigm, the focus of interaction research has been expanded from whether interaction relates to learning to how and why interaction facilitates learning. In order to pursue this inquiry, factors which might mediate interaction-driven language learning have become of particular interest among researchers, and more studies have been published in recent years. For instance, some research investigated numerous learner-external factors: different linguistic domains (e.g., K. Jeon, 2007); linguistic targets with differing degrees of perceptual salience (e.g., Mackey, Gass, & McDonough, 2000); learning contexts (e.g., Lyster & Mori, 2006; Sheen, 2004); task types (e.g., Révész & Han, 2006); feedback types (e.g., Ammar & Spada, 2006; Lyster & Izquierdo, 2009; Nassaji, 2009; Yang & Lyster, 2010); knowledge types (e.g., R. Ellis et al., 2006; Y. Han, 2010; Kang, 2009, 2010a, 2010b); past learning experiences (e.g., Gass & Lewis, 2007; Y. Han, 2010; Kang, 2009, 2010a, 2010b); age (e.g., Mackey & Oliver, 2002; Mackey & Sachs, 2012; Oliver, 1995, 2000); gender (e.g., RossFeldman, 2007); L1 (e.g., Gass & Lewis, 2007; Y. Han, 2010); interlocutors (e.g., Adams, 2007; Fujii & Mackey, 2009; Sato & Lyster, 2007); and so forth. Others studied learner-internal factors: working memory (e.g., Gass et al., in press; Goo, 2012; Mackey et al., 2002; Mackey & Sachs, 2012; Sagarra, 2007); language analytical ability (e.g., Sheen, 2007; Trofimovich et al., 2007); motivation; willingness to communicate; attitude toward error correction (e.g., Sheen, 2007); 18 anxiety (e.g., Sheen, 2008); and so on. Although some of the factors have been empirically investigated as a research focus, other factors have been merely suggested for future explorations in the literature. It is important to note that each factor may not work individually; rather, it may work in concert with other factors in this complex interaction-learning relation. The majority of recent empirical studies have concentrated on potential factors that may constrain the effects of feedback or particular sorts of feedback. Meta-analysis studies that have been mushrooming in SLA research lately identified potential constraining factors and reported the relationships between such factors and interaction-driven L2 development. In order to expand the current theoretical underpinnings and scope of interaction research, it is crucial to continue to pursue this line of research. Therefore, this study aims to examine learner-external factors that may affect the interaction and learning relationship and how these factors interact with each other. The factors to be investigated in this study are pragmatics as part of linguistic domains, learners’ HL background as part of past learning experiences, and explicitness of feedback as part of feedback types. Pragmatics as a Linguistic Target Pragmatic vs. Morphosyntactic and Lexical/Semantic Development As addressed earlier, a facilitative role of interaction has been attested by a great deal of empirical evidence since Long’s Interaction Hypothesis (1981, 1983, 1996). Nevertheless, the vast majority of interaction research has exclusively focused on learners’ morphosyntactic and lexical development. Lack of non-morphosyntactic studies has been indicated by researchers within the interactionist paradigm and beyond. Kasper and Rose (2002) drew attention to the heavy concentration of grammar as a focal object in the Interaction Hypothesis-based research: From the perspective of L2 pragmatic development, it is noteworthy that SLA research 19 under the interaction hypothesis has typically asked how interaction and the cognitive processes afforded through particular interactional arrangements contribute to learners’ acquisition of grammar [emphasis original]. (p. 33) Long (1996) also pointed out a similar issue in the Interaction Hypothesis: Negative feedback obtained during negotiation work or elsewhere may be facilitative of L2 development, at least for vocabulary, morphology, and language-specific syntax, and essential for learning certain specifiable L1-L2 contrasts [emphasis added]. (p. 414) In this influential Interaction Hypothesis, at least vocabulary, morphology, and syntax are the areas that have been proven to be influenced by negative feedback during interaction so far; however, it still remains unknown whether other language areas can be affected by such feedback available through interaction. In a similar vein, Mackey (2007) expressed a call for future investigation in phonological and pragmatic development to expand the current notion of the role of interaction: Interestingly, there has been very little interaction research to date that has focused on the acquisition of phonological features or pragmatics, although there is no reason to suspect that these areas would not be impacted by interaction. (p. 3) Simply put, the research interests in the interactionist position should move beyond the current scope of grammar-centered investigation. Whether the facilitative role of interaction is applies to other linguistic domains is an open question that needs to be pursued. There has been little attempt to explore L2 pragmatic development through conversational interaction, and only three studies on this relevant area are available. Soler (2002) researched the effect of teacherstudents’ versus learners’ interaction on the development of learners’ pragmatic competence of requests in the EFL context. The results indicated that pragmatic knowledge may emerge from 20 assisted performance in both teacher-students and peer interaction. Fukuya and Hill (2005) showed the possible applicability of recasts in the development of pragmalinguistic conventions of requests in the EFL setting. While engaged in role-plays, the participants in the pragmatic recast group received recasts, while the control group did not. The results showed that the pragmatic recasts group outperformed the control group on measures of both pragmatic appropriateness and grammatical accuracy. Featuring a pretest and posttests design, K. Jeon (2007) examined the differential effects of interactional feedback, mainly recasts, on the development of four different linguistic targets in Korean (i.e., object relative clauses, nouns, verbs, subject-verb honorific agreement). Unlike this dissertation, it is worth noting that her study focused on the subject-verb honorific agreement in the learning of subject referent honorifics. The findings indicated that although the effect of interactional feedback in the honorific development was statistically significant, it showed the least gains among the linguistic targets under investigation. Since the inclusion or exclusion of referent honorifics does not affect the actual referential meaning of the message (Brown, 2011), referent honorifics are low in a communicative value and are less salient. This may have led to the weakest effect of interaction feedback in comparison with other linguistic areas. In sum, these initial research findings suggest that interactional feedback, at least recasts, appears to contribute to promoting L2 pragmatic development. Nevertheless, ample empirical evidence in this line of research is sorely needed to confirm the current findings. Accompanied by the dramatic progress within the interactionist paradigm, the time is ripe to expand the scope of exploration that has been traditionally prevalent in interaction research. Mackey and Gass (2006) discussed a new direction of interaction research: Having firmly established the interaction-learning link, researchers have begun to extend 21 this knowledge base to new contexts, linguistic forms and languages, and interpretations of feedback [emphasis added]. (p. 171) In line with their claim, the present study intends to explore the role of interactional feedback in the development of Korean honorifics as a new linguistic form and language in relation to HL learning as a new learning context. Korean Honorifics as a Pragmatic Target Communicative competence refers to linguistic forms and knowledge of when, how, and to whom it is appropriate to use certain forms (Hymes, 1972). Lyons (1970) noted that the ability to use a language correctly in a specific situation is as important as the ability to produce grammatically well-formed sentences. Campbell and Wales (1970) also expressed a similar viewpoint, namely, that the ability to produce or understand not entirely grammatical, but more importantly, context-appropriate utterances is the most important linguistic ability. Along with these scholars, the significance of communicative competence has been discussed in the literature. The attainment of communicative competence in learning a language entails learning not only grammatical rules and lexicon, but also relevant sociolinguistic and pragmatic rules. Learners’ ability to use socially appropriate linguistic forms and rules that are pertinent to a certain situation constitutes a central element in communicative competence. Communicative competence mirrors Korean honorifics, one of the significant characteristics of the Korean language. Sohn (1999) defined Korean honorifics as grammatical and lexical forms that reflect the speaker’s socioculturally appropriate respect toward the addressee and the referent. In fact, Korean is an honorific language, and the honorific system constitutes the core of the Korean language, which renders the language unique. Sohn claimed that, in Korean, a speaker can barely form a sentence without determination of his or her social 22 relationship with the addressee and the referent according to cultural variables, consisting of power variables, such as age, kinship, gender, social status, and occupational rank, and solidarity variables, such as different degrees of intimacy/distance and the formality of situation. Korean utilizes a highly complex and systematic honorification as a means to express politeness or deference toward the person being addressed. In Korean, it is conventional to use honorifics when addressing a person to whom the speaker is expected to show deference, such as a parent and/or a teacher (W. Lee, 1991). In this dissertation, Korean honorifics are selected as L2 pragmatic targets because the grammatical pattern of Korean honorifics is the most systematic among all known languages (Sohn, 1999). One might argue that Korean honorifics are just morphological or lexical representations. However, it should be noted that when non-honorifics are utilized in a sentence where honorific use is of necessity, the sentence may still be grammatically correct, but pragmatically inappropriate under the given context. Thus, variables pertaining to the social relationships among the listener, the speaker, and the referent should be considered in forming a sentence in Korean to make it fully accepted for interpersonal relationships. As Korean is a highly context-dependent and situation-based language, and using Korean honorifics is restricted by various pragmatic constraints, the Korean honorific system is indeed one dimension of pragmatics. There is one last note that deserves attention with respect to the distinctiveness of Korean honorifics as a pragmatic focus. The great majority of pragmatics-based research involves speech acts, such as requests, apologies, and refusals. Even though both speech acts and honorifics are situated within the pragmatic domain, they show dissimilar properties. According to Sohn (2001), linguistic politeness can be realized either normative (obligatorily for discernment) or strategic 23 (optionally and volitionally). While the former concerns social indexing and structural mechanism as well as cultural norms and social principles, the latter is associated with the interlocutor’s strategic language use to attain his or her goal in social interaction. The former is relevant to Korean honorifics; the latter pertains to illocutionary and expressive language use, including speech act performances, such as hedges. Therefore, care needs to be used in directly comparing the research findings that involve these two areas. Thus far, evidence in supportive of Korean honorifics as a pragmatic target has been presented, and the subsequent section details Korean referent honorifics as a focal object in the present study. Korean Referent Honorifics As illustrated in Figure 1 below, Korean honorifics are composed of addressee honorifics and referent honorifics, with the latter being divided into subject and object honorifics. Addressee honorifics are utilized when the speaker gives respect to the addressee, whereas referent honorifics are used when a noun phrase of a sentence (e.g., subject, direct object, or indirect object) refers to someone the speaker should give respect to (W. Lee, 1991). Figure 1. Korean Honorific System Honorifics Addressee honorifics Referent honorifics Subject honorifics Object honorifics (W. Lee, 1991, p. 8) 24 Addressee honorifics are expressed by the choice of six different speech styles (i.e., plain, intimate, familiar, semiformal/blunt, polite, and deferential), which attach to a predicate located at the end of a sentence in Korean. These speech styles can be realized in different sentence types (i.e., declarative, interrogative, imperative, and hortative/propositive). The following Table 1, drawn and slightly adapted from Brown (2011) and Sohn (1999), indicates six levels of Korean speech styles; of these, only the deferential and the polite levels are honorific speech styles, whereas the remainders are generally considered as non-honorific speech styles. Hence, while the former is employed among non-intimate adults of comparatively equal rank, the latter is used among intimates and in-group members or in clearly downward directions of address by the speaker to his or her interlocutor. These speech levels are also ordered according to degrees of formality; the deferential level being the highest, while the plain level the lowest. Therefore, the deferential level is used in more formal contexts, such as lectures, news broadcasts, and official announcements, than the polite level within the honorific speech styles (I. Lee & Ramsey, 2000; Sohn, 1999; Strauss & Eun, 2005). Table 1. Korean Speech Styles by Sentence Types Speech level Declarative Interrogative Imperative Hortative Deferential -(su)pnita -(su)pnikka? -(u)psio -sipsita Polite -a/eyo -a/eyo? -a/eyo -a/eyo Semiformal/blunt -(s)o/(s)wu -(s)o? /(s)wu? -o/-wu -psita Familiar -ney -na?/-nunka? -key -sey Intimate -a/e -a/e? -a/e -a/e Plain -(n)ta -ni?/-(nu)nya? -(e)la/-kela -ca 25 Table 2 presents how these different speech levels are actually represented in a sentence for declarative forms. All these sentences shown below are distinguished from one another by different sentence endings. This indicates that the speaker has a choice of different levels of speech styles, depending on the addressee to whom he or she is speaking. Table 2. Examples of Korean Speech Styles for Declarative Forms Speech level Declarative Deferential Example -(su)pnita Haksayng-i o-pnita. Student-S come-V A student is coming. Polite -a/eyo Haksayng-i o-ayo. Semiformal/blunt -(s)o/(s)wu Haksayng-i o-o. Familiar -ney Haksayng-i o-ney. Intimate -a/e Haksayng-i o-a. Plain -(n)ta Haksayng-i o-nta. In the area of referent honorifics, subject honorifics occur when the subject refers to someone who is socially superior to the speaker. I. Lee and Ramsey (2000) addressed that, in subject honorifics, the speaker makes a decision to show respect and use the honorific form since, in the speaker’s judgment, the subject is older and/or has a higher social rank than himself or herself. That is, the social status of the subject is compared against that of the speaker. In contrast, object honorifics are used when the object refers to someone who is socially superior to the person performing the action of the verb (e.g., subject). In object honorifics, the comparison is 26 made not against the speaker, but against the subject of the sentence. Thus, the speaker is excluded in this reference construction. I. Lee and Ramsey indicated these associations in Figure 2 and Figure 3 shown below. The size of the boxes represents the degree of social rank, and the lines connecting the boxes indicate the direction of respect. Figure 2. Subject Honorifics Speaker [ Subject - Object - Predicate ] Listener Subject - Object - Predicate ] Listener Figure 3. Object Honorifics Speaker [ (p. 249) The linguistic targets in the present study are referent honorifics, including subject honorifics and object honorifics. For subject honorifics, the focus is the honorific subject case marker –kkeyse and the honorific verbal morpheme –si in an honorific sentence. In a sentence with an honorific subject who is older and/or higher in social status than the speaker, W. Lee (1991) presented the rule of subject honorifics indicating that the subject should be modified by the honorific subject case marker –kkeyse, and the predicate must be modified by the honorific verbal morpheme –si with a proper inflection. The honorific verbal morpheme –si can be used with any verb stems, but there are some exceptions. Special polite verbs, such as capswusita ‘to eat,’ cwumusita ‘to sleep,’ phyenchanhusita ‘to be sick,’ tolakasita ‘to die,’ and kyeysita ‘to be,’ in which the honorific verbal morpheme –si are embedded, are expected to be used in the polite 27 situations. The example (1) below shows an honorifics sentence with a special polite verb. Example 1. Subject Honorifics with a Special Polite Verb Sensayngnim-kkeyse Teacher(H)-S(H) kyosil-ey kyeyseyyo. classroom-LOC be-V(SH) A teacher is in the classroom. Although W. Lee (1991) noted that the non-honorific subject case marker -i (-ka) in a sentence that requires honorification seems to be an acceptable alternative, using the honorific subject case marker –kkeyse, as in prescriptive grammar, is encouraged and taught in Korean language classrooms to satisfy the subject-verb honorific agreement requirement. One might argue that the honorific subject case marker –kkeyse sounds overly honorific in some occasions, particularly in an intimate relationship. However, by rule, a subject in a sentence who deserves the use of the honorific verbal morpheme –si should also take the honorific subject case marker – kkeyse, and the use of the non-honorific subject case marker -i (-ka) is regarded as exceptional. Although both forms seem acceptable, the use of the honorific subject case marker –kkeyse certainly represents the higher level of deference than the honorific verbal morpheme –si (I. Lee & Ramsey, 2000). In contrast to the example (2) shown below that includes the non-honorific subject case marker -i (-ka), the example (3) illustrates a sentence which requires honorification. Example 2. Subject Non-honorifics Haksayng-i (-ka) kong-ul chanta. Student-S ball-DO kick-V 28 A student kicks a ball. Example 3. Subject Honorifics Sensayngnim-kkeyse kong-ul cha-si-nta. Teacher(H)-S(H) ball-DO kick-V(H) A teacher kicks a ball. For object honorifics, the honorific indirect object case marker –kkey and the humble verbal form (-e)tulita in an honorific sentence are the focus. W. Lee (1991) suggested the rule of object honorifics that the object honorifics are shown by the honorific indirect object case marker –kkey, and the predicate is marked as humble (e.g., (-e)tulita ‘to give’) or honorific (i.e., the honorific verbal morpheme –si with a proper inflection), as applicable. In Korean, there are few predicates that represent humbleness of the subject in object honorifics. For instance, some sentences that involve object honorifics need such humble verbs as mosita ‘to accompany,’ poypta ‘to meet,’ and yeccwuta ‘to ask,’ in addition to (-e)tulita ‘to give.’ As shown below, contrary to the example (4) which includes the non-honorific indirect object case marker -eykey (-hanthey), the example (5) displays a sentence that requires honorification. The indirect object sensayngnim ‘teacher,’ is modified by the honorific indirect object case marker –kkey, and the predicate takes the humble form (-e)tulita ‘to give,’ instead of the non-honorific form cwuta ‘to give.’ Example 4. Object Non-honorifics Haksayng-i (-ka) chinkwu-eykey (-hanthey) chayk-ul 29 cwunta. Student-S friend-IO book-DO give-V A student gives a book to a friend. Example 5. Object Honorifics Haksayng-i (-ka) Student-S sensayngnim-kkey chayk-ul teacher(H)-IO(H) book-DO tuli-nta. give-V(Hum) A student gives a book to a teacher. Such explanations of the Korean honorific system as provided above are generally introduced to learners through Korean language textbooks. Yet, Brown (2010) pointed out the problem that some textbooks at times underrepresent the value of honorifics by relying on one particular level or a limited range of honorifics in dialogues. Therefore, it can be too deferential or insufficiently deferential in certain situations. It may be intended to make language learning easier for foreigners, but the simplification and/or misrepresentation of honorification often prevent the learners from employing diverse honorific forms and instead just make them play it safe (Choo, 1999). Brown maintained that this sort of honorific presentation is not only inauthentic but also inappropriate in real-life situations and thus suggested the representation of an accurate and realistic picture of honorifics in Korean textbooks. Given that Korean honorifics entail a complicated system, many book chapters and volumes have been allotted solely to present details of this system. Research on Korean honorifics has been conducted primarily from linguistic, pedagogical, or sociolinguistic perspectives. Since the honorific system is realized based on the understanding of interpersonal relationships among interlocutors, Korean honorifics have been mainly considered as a social 30 aspect and thusly approached from a sociolinguistic perspective. Consequently, any attempt to explore Korean honorifics from a cognitive viewpoint has been exceedingly rare. Moreover, there is scant evidence to examine and further document this system from a language acquisition point of view. In L1 literature, there has been only limited interest in the honorific development of Korean L1 children who grow up in Korea. Researchers generally agree that Korean honorifics tend to appear very early in L1 acquisition, sometime around two years old (Y. Kim, 1997; K. Lee, 1997; J. Park & M. Kim, 2010). This early appearance is limited to explicit modeling and correction from parents, who use a number of strategies to teach honorifics to their children. However, it appears that a full mastery of honorifics does not follow quickly and that adultlike performance may not occur until university age (L. Brown, personal communication, January 15, 2013). In the study on the acquisition of Korean honorific markers by 297 Korean children aged from three to five, J. Park and M. Kim reported that young Korean children acquired honorific markers in the order of hearer-honorifics, subject honorifics, and then object honorifics. In the area of subject honorifics, for instance, the use of the honorific subject case marker –kkeyse was hardly observed by the age of five. In contrast, 70 % of three-year-old children and 90 % of fiveyear-old children acquired the honorific verbal morpheme –si, which lends support to C. Lee’s (1993) claim that –si appears around the age of two. Consistent with K. Lee’s (1997) and Moon’s (2010) findings, object honorifics appeared later and were developed more slowly than subject honorifics: just a small number of four- and five-year-old children used the honorific indirect object case marker –kkey, and only roughly 20 % of the four- and five-year-old children were capable of utilizing the humble verbal form (-e)tulita accurately. This can be construed that not only the honorific subject case marker –kkeyse but also the honorific indirect object case marker 31 –kkey may not be acquired until school age. A couple of studies on the acquisition of Korean honorifics as a second/foreign language may be also worth mentioning. Wang (1995) examined parental influence on Korean HL learners’ acquisition of honorifics using a narrative story, an interview, and a postsession questionnaire. It showed that the majority of the HL learners had difficulties in using honorifics properly, and such inadequacy was, in part, ascribed to their family background. It was construed that although the learners tried to use honorifics correctly after receiving formal instruction at school, it tended to be difficult to give up habits of non-honorific use that were already formed at home. Hence, home environments where the learners grew up listening to their parents were an essential factor influencing their use of honorifics. Using a written discourse completion test (DCT) and a multiple-choice questionnaire, M. Lee (1997) looked into acquisition of Korean referent honorifics by advanced adult learners of L2 Korean. The focus was on the agreement of the honorific title -nim, the honorific case marker -kkey(se), and the honorific verbal morpheme si in referent honorifics. The findings indicated that the learners recognized that sentences with perfect agreement of honorific elements were good and that they should use honorifics in certain contexts. However, they had difficulties in producing their knowledge in writing and often overused or avoided honorifics. Even though the last two studies depicted above afforded relevant information on the honorific acquisition of Korean HL/L2 learners, they both are exploratory and descriptive in nature, and are small-scaled studies. Indeed, more systematic research on the development of honorifics is necessary based on solid theoretical background to better understand Korean honorific development. More rigorous research design is required to pursue this goal. Additionally, most research on the development of Korean honorifics has focused on the HL 32 learners; therefore, how the NHL learners learn this complex honorific system relative to the HL learners has been hardly explained. Therefore, this dissertation intends to incorporate the issues identified and advance our understanding of how honorifics are acquired. Heritage Language Background Definitions of Heritage Language Learners In the United States, the area of HL has been gaining grounds since 1990s, and the use of the term HL is a recent phenomenon. The HL is broadly defined as an indigenous, colonial, and immigrant language that a speaker has a particular family relevance (Fishman, 2001; Wiley, 2005). As used in the United States at present, the HL generally refers to all non-English languages which include languages that Native Americans speak (Valdés, 2001a). The term, HL, is often used as a “community language,” “native language,” and “mother tongue,” namely, a non-English language that immigrants or their children use. HL learners come from different family backgrounds with varying degrees of HL learning experiences, proficiency, cultural knowledge, motivation, and so forth. In consideration of these vastly diverse HL learner profiles and prior language learning experiences, the HL learners are undoubtedly characterized as a heterogeneous group (Kagan & Dillon, 2012; Kanno, Hasegawa, Ikeda, Ito, & Long, 2008; Kondo-Brown, 2005; Valdés, 1995). This leaves language teachers a great challenge on how best to teach this group of learners together with classic L2 learners in the same classrooms (Carreira, 2003, 2007; Potowski, 2002; Valdés, 1997, 2001a). This is also linked to one of the foremost controversial issues in HL education and research: identifying and defining who HL learners are. Nonetheless, unfortunately, there is no consensus in defining this tremendously diverse HL learner population thus far (Hornberger & Wang, 2008; Wiley, 2001). Scholars varied their perspectives on this issue and presented a range of definitions 33 of HL learners. This is reflected in that the HL learners are also referred to as “native speakers,” “quasi-native speakers,” “residual speakers,” “bilingual speakers,” and “home background speakers” (Valdés, 1997). Although there are differing approaches to define HL learners, two general concepts are proposed. In a broad view, Van Deusen-Scholl (2003) defined HL learners as “a heterogeneous group ranging from fluent native speakers to nonspeakers who may be generations removed, but who may feel culturally connected to a language” (p. 221). Van Deusen-Scholl further made a distinction between “heritage learners” and “learners with a heritage motivation.” The former refers to those learners who have achieved some degree of proficiency in the home language and/or have been raised with strong cultural connections; the latter refers to those learners who “have been raised with a strong cultural connection to a particular language through family interaction” and “seek to reconnect with their family’s heritage through language, even though the linguistic evidence of that connection may have been lost for generations” (p. 222). However, although Van Deusen-Scholl’s comprehensive definition includes both cultural heritage and linguistic heritage, it is not clear whether it is an ethnolinguistic affiliation or a language proficiency that is more salient in identifying who a heritage learner is (Wiley, 2001). Particularly, according to Van Deusen-Scholl’s latter definition of “learners with a heritage motivation,” it is difficult to distinguish these learners with a heritage motivation from typical L2 learners in light of their linguistic competence. Suppose that African-American students grew up in an English-monolingual family, and their Swahili language use at home is removed for generations. They come to a language classroom with no prior knowledge in the target language, like classic L2 learners, to learn their ancestral language. Except that these Swahili HL learners are culturally motivated to discover their lost language and heritage, it may be impossible to 34 differentiate these Swahili heritage learners from the regular L2 learners with regards to their language proficiency. Since identifying students’ linguistic proficiency is important for successful language instruction, language educators use the term HL learner based on a pedagogical perspective. Valdés (2001a) suggested the term in a narrow sense that a HL learner refers to “a language student who is raised in a home where a non-English language is spoken, who speaks or at least understands the language, and who is to some degree bilingual in that language and in English” (p.38). As addressed in this definition, it is apparent that HL learners have varying degrees of HL proficiency, which can be distinguishable from classic L2 learners who are true beginners. Valdés’s view also fits the characteristics of second-generation Korean HL learners (J. Lee & Shin, 2008). For example, second-generation Korean-American learners have been exposed to Korean language input at home since birth and grow up using the language with their parents who are Korean NSs and other family members. These learners are already familiar with some Korean culture and possess a range of Korean language skills, but wish to relearn the Korean language in class. Some may possess limited Korean language skills, but they may be sufficient to carry out a basic conversation without much difficulty; others may be fairly competent and even sound like a NS of Korean. Also, some of these learners are eager to pursue their ethnolinguistic roots, whereas others simply want to fulfill their foreign language requirement to graduate. From the above definition Valdés (2001a) suggested, however, it is important to note that the term bilingual can be interpreted differently depending on one’s perspective. One might argue that a bilingual speaker is expected to have an ability to use both languages with perfectly equivalent proficiency. In other words, he or she should be able to perform not only basic routine 35 tasks in daily lives but also superior-level tasks which may involve complex, abstract, and sophisticated language use in formal situations in their HL language as well as dominant language. Unfortunately, there is general agreement that this perfect bilingualism is too idealistic and even mythical; in fact, no one can achieve such equally balanced bilingualism in every language domain and every context. Hence, Kagan and Dillon (2012) concluded that HL speakers are in fact bilinguals, but typically unbalanced bilinguals, meaning that their proficiencies in the two languages are not equal, with the HL being the weaker of the two. Valdés suggested that bilingualism falls along a continuum with different types of bilinguals and strengths, as shown in Figure 4. Different letter sizes imply different levels of strengths in different bilinguals. For example, a recent immigrant can be represented as Ab; a fourthgeneration of an immigrant can be indicated as Ba. Thus, an English NS who has reading knowledge in German can be more bilingual than an English-monolingual. Notwithstanding this individual is far from being a perfect bilingual, he or she is not evidently a monolingual and is still considered as a bilingual. Figure 4. The Bilingual Continuum (Valdés, 2001a, p. 41) Real Bilingualism: A Continuum Monolingual Monolingual Language A Language B A Ab Ab Ab Ab Ab aB Ba Ba Ba Ba Ba B 36 Thus far, a debate over the definitions of HL and HL learners has been discussed along with a bilingual continuum of the HL learners. In the next section, what characterizes the acquisition of HL relative to L1 and L2 acquisition is shown, and each of the characteristics is explained in detail. Characteristics of Heritage Language Acquisition HL acquisition is situated between L1 and L2 acquisition. This means that HL acquisition shares some characteristics of the two different acquisition systems, while it stands alone with its distinct features (Brinton, Kagan, & Bauckus, 2008; Montrul 2010a). Table 3 details characteristics of L1, L2, and HL acquisition, and characters in italicized bold font represent an intersecting subset between L1 and L2 acquisition that characterizes HL acquisition. As clearly indicated in this table, HL learners are characterized as L1/L2 users (Cook, 2001; Valdés, 2001a). They are exposed to rich aural input and interact with family members in the HL (minority language) at home from infancy. As they begin to acquire the HL at an early age, they establish a good control of some early-acquired linguistic features, including subset of phonology, vocabulary, and grammar structures. During this process, the HL learners commit developmental errors similar to native-speaking children of the target language and foreign language learners. When the two languages (i.e., HL and majority language) conflict, language transfer either from L1 to L2 or vice versa may take place (e.g., Albirini, Benmamoun, & Saddah, 2011; Montrul 2008a, 2010c; Pavlenko & Jarvis, 2002). The HL learners’ proficiency is achieved in varying degrees due to their heterogeneous language background (Kagan & Dillon, 2012). Since the HL learners are unable to master the HL fully or to reach at least the level of their age-matching native-speaking peers, incomplete acquisition of the HL can occur (e.g., Albirini et al., 2011; Montrul, 2008a; Polinsky, 2007). Then, the HL can be possibly fossilized (e.g., Z. Han, 2004; H. 37 H. Kim, 2001; Wang, 1995). The HL learners’ motivation, attitude, and affective factors play an important role in order to continue to acquire or reacquire the HL (e.g., H. H. Kim, 2002; J. Lee & H.-Y. Kim, 2008). Table 3. Characteristics of L1, L2, and Heritage Language Acquisition (Montrul, 2010a, p.12) L1 acquisition L2 acquisition Early exposure to the language. Late exposure to the language. Abundant input in a naturalistic setting (aural Varying amount of input in instructed and/or naturalistic setting (aural and written input). input). Control of features of language acquired very Grammar may be incomplete (no chance to early in life (phonology, some vocabulary, develop other structures and vocabulary). some linguistic structures). Developmental errors. Developmental and transfer errors. Outcome is successful and complete. Outcome is variable proficiency. It is typically incomplete. Fossilization does not occur. Fossilization is typical. No clear role for motivation and affective Motivation and affective factors play a role in factors to develop linguistic competence. language development. More complex structures and vocabulary Experience with literacy and formal instruction. developed at school after age 5, when metalinguistic skills develop. Some receive exposure to the majority language input simultaneously, whereas others 38 become exposed to it when they enter school typically around age five. Scholars label the former as simultaneous bilingual and the latter as sequential bilingual (Hornberger & Wang, 2008; Montrul, 2010a, 2011b). Since the HL use of these learners is mostly confined to home environments and thusly comprises casual and informal talk, they tend to lack grammatical precision as well as age-appropriate and context-appropriate sociolinguistic registers used in a variety of situations. Regardless of their bilingual status, however, they spend a bulk of their time using the majority language, such as English, at school, and experience reduced and/or fluctuated HL input once they start school (Domínguez, 2009; Montrul, 2011a; O’Grady, Kwak, O. Lee, & M. Lee, 2011). The linguistic system of HL reveals that many aspects of HL grammars undergo simplification (e.g., Albirini et al., 2011; Montrul, Foote, & Perpiñán, 2008; O’Grady et al., 2011; Polinsky, 2011) and overgeneralization of the default language form (e.g., Albirini et al., 2011; H. H. Kim, 2001). Even if the HL learners acquire a full competence of the HL, the majority language has already dominated their language use on daily basis. It is most likely that the HL ultimately becomes weaker than the majority language. As they grow older, their HL use possibly results in language loss, language erosion, or language attrition (e.g., Albirini et al., 2011; de Bot & Weltens, 1991; Maher, 1991; O’Grady et al., 2011; Olshtain & Barzilay, 1991; Polinsky, 2011; Seliger & Vago, 1991). Furthermore, the HL learners can be characterized as a childhood language overhearer as opposed to a childhood language user (e.g., speaker) per se. They are inclined to overhear conversations among their family members of older generation (e.g., grandparents, parents, and relatives) in the HL. The HL learners also tend to speak or respond in the majority language while their older family members speak to them in the HL (H. H. Kim, 2008). Even if they speak the HL with their family and relatives, it can be minimal, and often include occasional words and 39 short phrases in the HL embedded in utterances in the majority language (Au, Knightly, Jun, & Oh, 2002). Hence, the HL learners tend to use the HL more receptively than productively, and this results in their lack of output. There has been empirical evidence that relearners (e.g., overhearers) outperformed first-time learners in both phonology and grammar. This suggested that even incomplete or discontinued childhood experience with a language helps adult learners’ development of phonology and morphosyntax (Au & Romo, 1997), typically the most difficult areas for the adult learners (Fledge, 1987; Johnson & Newport, 1989; Newport, 1990, 1991; Oyama, 1976). Furthermore, research indicated that childhood overhearing experience may facilitate the learning of phonology (perception), whereas childhood speaking experience may promote the learning of both phonology (production) and grammar (Au et al., 2002; Au, Oh, Knightly, Jun, & Romo, 2008; Knightly, Jun, Oh, & Au, 2003; Oh, Jun, Knightly, & Au, 2003). This points to the important role of output and indicates that using a language productively, not simply overhearing it, is essential in order to produce the language with nativelike phonology and grammar. The characteristics of the HL acquisition are drawn from both L1 and L2 acquisition, but it differs from each of the L1 and L2 acquisition. The field of HL acquisition stands alone and has its own grounds with unique features. Since this dissertation involves HL learners relative to NHL learners, it may be worthwhile to review how these two groups of learners are similar to and/or different from each other. Heritage vs. Non-heritage Language Learners Although there is no single linguistic profile of HL learners due to their heterogeneous language learning experience (Kagan & Dillon, 2012; Kanno et al., 2008; Kondo-Brown, 2005; Valdés, 1995), the nature and process of HL learning differ from those of foreign language 40 learners (Campbell & Rosenthal, 2000; Valdés, 1995, 2001b). Among a number of variables that distinguish between the HL and the NHL learners, target language proficiency is one of the primary indicators that differentiate these two groups of learners. For example, the HL learners often arrive in a language classroom with great fluency and nativelike pronunciation and intonation in the target language. Some HL learners have limited speaking skills in that language, but their listening skills are generally excellent, and they are capable of understanding even fast and spontaneous speech. Although the HL learners may lack not only grammatical accuracy but also age- and context-appropriate pragmatic skills, they have acquired a good amount of awareness and knowledge of the target language and its culture. They also demonstrate a range of target language vocabulary, although their use of vocabulary may be restricted to home environments (Kagan & Dillon, 2012). Thus, the HL is often referred to as “kitchen language” (Polinsky & Kagan, 2007). The HL learners’ apparently superior target language abilities often intimidate and frustrate regular foreign language learners who start from scratch. Although the HL learners may possess low target language literacy, it is evident that the HL learners have a head start from the beginning with many advantages they have already obtained during childhood. In contrast, typical NHL learners come to a language classroom without a home language background and without prior knowledge. Another variable that differentiates between the HL and the NHL learners is the language learning process. The HL learners are inclined to employ top-down processing strategies (i.e., macro approach) (Celce-Murcia & Olshtain, 2000; Z. Han, 2004; H. H. Kim, 2005). They have a tendency to approach language learning tasks as a whole instead of focusing on sub-elements, such as grammatical categories, vocabulary, and spelling. That is, they negotiate meaning in a large variety of contexts (Kagan, 2005), and this may be associated with their meaning-based 41 comprehension or semantic processing strategies. For example, while writing, they tend to concentrate on content and use their high degree of internal grammar they have acquired in a natural setting. Hence, overall, this group of learners is known to rely on intuition and implicit competence (knowledge) to learn a language implicitly (Carreira & Potowski, 2011). Yet, once the HL learners are in the classroom as adults, it remains unclear whether they are still dependent on intuition as if they would pick up the HL as a child in a natural setting or they adopt new learning strategies that require a metalinguistic analysis as an adult. As Montrul (2010a) noted in her recent review article on HL acquisition, this is clearly an open question that needs future investigation: Perhaps the biggest question is how heritage language learners react to classroom instruction, if they have implicit knowledge of the heritage language acquired in childhood. Once in the classroom, will they continue to learn the heritage language implicitly as L1-acquiring children, or will they now rely on explicit learning, like adult L2 learners? Can they eventually catch up with the missing explicit and metalinguistic knowledge that they did not get at school through reading and writing instruction? (pp. 17-18) In contrast, the NHL learners are prone to utilize bottom-up processing strategies (i.e., micro approach) in language learning. In general, they start learning small and simple things and gradually and slowly increase in volume and complexity. For instance, they initiate writing from a sentence level and gradually advance to a paragraph level. They also heavily rely on their metaliguistic ability and explicit knowledge they have gained as an adult in school and apply their world knowledge and schema to learn a new language (Carreira & Potowski, 2011). Thus, they are accustomed to explicit learning that entails drill, feedback, exercise, and so forth. As 42 such, this unique group of the HL learners differs from the NHL learners in many respects, and subsequent research should reveal additional information on these different types of learners. Unfortunately, however, a large body of studies that characterizes HL acquisition has been anecdotal, descriptive, or intuitional. It may be attributed to the fact that HL acquisition is a relatively young field, which has not been built upon a profound theoretical foundation so far (Lynch, 2003; Montrul, 2008b; Valdés, 1997, 2000c, 2005). A good part of research has been provided by language educators who observe HL learners’ linguistic behaviors in language classrooms and researchers who are in search of HL learners’ pedagogical needs, linguistic profiles, motivation, and identity mainly in the form of surveys or interviews. For the most part, the research is based on sociolinguistic and pedagogical perspectives. Nonetheless, empirical research adopting theories of L1 and L2 acquisition is on the constant increase in recent times to help expand our current understanding of HL acquisition. These recent studies primarily adopt linguistic, sociolinguistic, and psycholinguistic approaches and enable researchers to delve systematically into the differences between these two learner groups. The recent studies on HL acquisition often probed linguistic similarities and differences between the HL and the NHL learners in different language domains. According to BleyVroman’s Fundamental Difference Hypothesis (FDH) (1989a, 1989b, 1990), different learning mechanisms work in learning L1 and L2. Children get access to UG in L1 acquisition (Chomsky, 1986). After a critical period, however, this UG is no longer in use. Hence, postpuberty language learners do not gain an ultimate L2 mastery due to the absence of UG. The FDH explains why every L1 speaker achieves a full spectrum of L1 language use, whereas it is highly unlikely for adult learners to reach a perfect mastery in L2. According to this view, the HL learners should be more advantaged than the NHL counterparts in learning the target language. 43 Research reported that the HL learners generally show more nativelike phonological abilities than the NHL learners. Au et al. (2002), Knightly et al. (2003), and Oh et al. (2003) examined phonetic perception and production abilities of low proficiency Korean and Spanish HL learners. The findings indicated that the HL learners possess more nativelike pronunciation and a more sensitive ability for phonological perception than the NHL learners, although the HL learner group has a distinguishable accent when compared to NSs of the target language. However, the results on morphosyntactic features, although they were too broadly defined, indicated that the HL and the NHL learners did not differ from each other. Research on morphosyntactic differences between the two learner groups showed mixed results. On the one hand, some studies showed that early exposure to the HL does not render an advantage to the HL learners compared with the NHL learners of similar proficiency. Bruhn de Garavito (2002) investigated Spanish verb-raising, but did not find any evidence that the HL learners are different from the NHL learners. Furthermore, O’Grady et al. (2001) conducted a comparative study in the acquisition of relative clauses. It was found that the HL learners were not significantly different from the NHL learners in light of their ability to use morphosyntactic clues to interpret complex sentences that involve relative clauses. Rather, they found fundamental similarities in that both types of learners did better on subject relative clauses than on direct object relative clauses in comprehension, and both of them displayed numerous errors in identifying the head of relative clauses. In his small-scale study, J. Kim (2001) provided further evidence that the HL learners do not enjoy advantages over the NHL peers in learning parametric values that differ from those of their L1 English. The two groups demonstrated L1 interference to a similar extent in learning Korean null subject and wh-in-situ constructions. More recent studies provided additional evidence to advance the above findings. Keating, 44 VanPatten, and Jegerski (2011) studied the effects of early exposure to input at the syntaxpragmatics interface, namely anaphora resolution, in Spanish HL speakers and adult L2 learners of advanced proficiency. They discovered that early input and language use does not confer an advantage to HL speakers relative to L2 learners concerning grammatical properties that are acquired later in life or that need sufficient quantities of uninterrupted input. Rather, the Spanish HL speakers selected more pragmatically inappropriate subject antecedents for overt pronouns than the L2 learners. On the other hand, studies provided evidence that the HL learners have an advantage over the NHL learners at least in some aspects of morphosyntax. For example, Montrul (2005) showed that the HL learners of low proficiency outperformed the NHL learners in the use of Spanish unaccusativity. Montrul (2006) further found that the HL learners acquire and maintain some features of syntax; for example, a null subject parameter, which is acquired early during childhood, whereas the NHL learners learn these early-acquired syntactic aspects gradually as time progress. With regards to morphological errors, both Montrul et al. (2008) and Montrul (2011b) provided evidence that for HL and L2 learners, inflectional morphology is equally problematic, and both groups produced and chose default forms where more morphologically specific forms were needed. An interesting observation was made that while the HL learners outperformed in the oral tasks perhaps due to the NHL learners’ performance errors, the NHL learners were more accurate in the written tasks that entail their metalinguistic knowledge for the completion of the tasks. This evident task effect mirrors the tendency that HL learners have stronger speaking and listening skills, whereas NHL learners outdo in reading and writing skills. Montrul (2010b) provided further evidence that HL learners have more nativelike and implicit knowledge and use than L2 learners of low-intermediate or intermediate proficiency at least in 45 some grammatical areas, such as clitics and word order. Additionally, Montrul (2011b) claimed that while both HL learners and instructed L2 learners showed incomplete knowledge of tenseaspect and mood morphology, they differed in consideration of a linguistic target, age, and a manner of acquisition. The HL learners were more targetlike than the L2 learners in earlyacquired features, such as tense and aspect, although they were not better than the L2 learner counterparts in mood morphology, which is a late-acquired grammatical area. The HL learners were more accurate than the L2 learners in the sentence conjunction judgment task, which requires minimal metalinguistic knowledge, whereas the L2 learners were more accurate than the HL learners in the written tasks including the morphology recognition task that utilizes metalinguistic knowledge. Overall, there is a general understanding that early exposure and language use confer an advantage to HL learners at least in the areas of pronunciation and phonology. Although the majority of experimental research on HL acquisition is centered on the HL learners’ morphosyntactic development, it still remains unclear whether the HL learners are different from the NHL learners in this area. Yet, a putative conclusion thus far can be made that HL acquisition experience during childhood brings advantages to the HL learners in some aspects of grammars (e.g., early-acquired features), but not in others (e.g., late-acquired features). Still, more research that examines HL grammars in detail is called for to reveal intricacies behind HL acquisition. Unfortunately, however, almost all the studies investigating linguistic differences between the HL and the NHL learners exclusively focus on grammars (i.e., phonetics, phonology, semantics, syntax, and morphology). There is no published study on this topic in the area of pragmatics. Therefore, it is unknown if the current findings in grammars can be extended to the acquisition of pragmatics, and this dissertation aims to explore this inquiry. 46 Explicitness of Interactional Feedback/Instruction Feedback Classification and Implicit vs. Explicit Dimension Early interaction research efforts were primarily concerned with how negotiation that occurs during interaction promotes L2 development. While these early studies revealed the important role of negotiation, they did not focus on investigating the relative effects of different feedback types attributable to L2 development. Lyster and Ranta (1997) identified different types of feedback based on French immersion classroom interaction at the primary level, and this seminal work has generated a great deal of work looking at differential effects of feedback types in relation to learning. Although there is no general consensus regarding feedback categories identified in L2 classroom, the following feedback classifications have been widely used in L2 interaction research. Lyster and Ranta proposed an analytic model for an error treatment sequence and identified the following six feedback moves in L2 classrooms: explicit correction, recasts, clarification requests, metalinguistic feedback, elicitation, and repetition. According to Lyster and Ranta, explicit correction refers to the provision of a correct form with a clear indication that an error has occurred. Recasts refer to teachers’ full or partial restatement of learners’ erroneous utterance, minus an error. Clarification requests refer to a request of either a repetition or a reformulation when learners’ utterance is misunderstood or ill-formed. Metalinguistic feedback refers to comments, information, or questions regarding wellformedness of learners’ utterance, without providing a correct form explicitly. Elicitation refers to teachers’ attempt to elicit a correct form from learners by pausing, asking questions, or requesting learners to reformulate their utterance. Repetition refers to teachers’ repetition of learners’ erroneous utterance. These feedback moves were further grouped depending on the degree to which they generate learner self-repair. Prompts push them to self-correct their output 47 (i.e., repair) and include the following four feedback moves: clarification requests, metalinguistic feedback, elicitation, and repetition (Lyster, 2004). Some scholars grouped the above feedback types into two main categories: input-providing (e.g., recasts and explicit correction) as opposed to output-pushing (i.e., output-prompting) feedback (e.g., prompts, such as elicitation, clarification requests, and repetition of error) (R. Ellis, 2006, 2010; Sheen, 2010). The former affords learners with a correct form, whereas the latter withholds the correct form and encourages learners to self-repair through prompts. It is argued that these two feedback categories can be explained based on learners’ different levels of cognitive processing. Inputproviding feedback allows learners to cognitively compare their incorrect form and a correct form available in input in working memory, whereas output-pushing feedback enables them to retrieve their existing knowledge in long-term memory (Yang & Lyster, 2010). Others mainly divided feedback types according to explicitness of feedback: implicit (e.g., recasts, clarification requests, elicitation, and repetition) and explicit feedback (e.g., metalinguistic feedback, explicit correction) (Li, 2010a). Yet, the division of these feedback types is not straightforward and has generated ongoing debate. Since prompts and output-pushing feedback contain both implicit (e.g., clarification requests, elicitation, and repetition) and explicit feedback (e.g., metalinguistic feedback), it is difficult to look into the relative effects of explicitness of feedback. Besides, it is hard to discover specific and relative effects of different types of feedback under the category of the prompts or output-pushing feedback. The classification of implicit and explicit feedback has also raised some issues. Although recasts are regarded as implicit feedback in general, they can be diverse in terms of explicitness (Loewen & Philp, 2006; Sheen, 2006) depending on research contexts (e.g., laboratory vs. classroom) and characteristics of recasts (e.g., length, number of changes, salience, and number 48 of linguistic targets). In Doughty and Varela’s (1998) study, for example, corrective recasts consisted of repetitions of learner output of imprecise segments with rising intonation as well as recasts to provide the contrastive L2 forms. It is evident that the addition of repetitions of learner errors and rising intonation made the corrective recasts more explicit than the simple provision of recasts. Explicit feedback includes a broad range of feedback types. For instance, explicit rejection (Carroll & Swain, 1993), simply informing learners that an error has occurred, is more implicit than metalinguistic correction, which supplies learners with a correct form as well as metalinguistic information (Sheen, 2007). Therefore, researchers seem to be in agreement that dichotomous classification can be problematic, and the implicit and explicit dimension of feedback should fall along a continuum (Loewen & Nabei, 2007; Lyster and Saito, 2010). Of important note is that recasts are not in the absolute end of the implicit dimension; rather, they are toward the implicit end of the continuum. Similarly, metalinguistic correction should be more toward the explicit end of the continuum than simple explicit rejection. In consideration of several feedback schemes which have received considerable attention in interaction research (e.g., R. Ellis, 2006, 2010; Li, 2010a; Lyster, 2004; Lyster & Ranta, 1997; Sheen, 2010), Loewen and Nabei (2007) as well as Lyster and Saito (2010), respectively, summarized the feedback categorizations and proposed a more comprehensive picture of the feedback taxonomy. With some modifications, both of their feedback classifications are combined and presented in Figure 5. In this figure, while only recasts and explicit correction are under the category of reformulations/input-providing/other-repair, the rest of the feedback types are classified as prompts/output-pushing/self-repair. It is clearly shown that recasts, clarification requests, repetition, and elicitation are toward the implicit dimension, whereas explicit correction and metalinguistic feedback are toward the explicit dimension. 49 Figure 5. Classification of Corrective Feedback Clarification Repetition Elicitation Metalinguistic clues requests Metalinguistic clues plus other feedback Prompts/Output-pushing/Self-repair Implicit Explicit Reformulations/Input-providing/Other-repair Recasts Explicit correction Recasts vs. Metalinguistic Feedback Based on the feedback scheme suggested in Figure 5, this study investigates the relative impact of recasts and metalinguistic feedback on learners’ L2 development. Different types of interactional feedback and their differential influence on learning have been actively investigated according to the several feedback classifications addressed earlier. In particular, a great amount of attention has been drawn to the role of recasts in the past decade in part because of research indicating that recasts are shown to be the most frequent type of interactional feedback in the classroom (Braidi, 2002; R. Ellis & Sheen, 2006; Lyster & Ranta, 1997) and provide both positive and negative evidence (Gass, 2003). Recasts are defined slightly differently depending on researchers’ respective viewpoints. For instance, in Lyster and Ranta’s study, recasts are referred to as “the teacher’s reformulation of all or part of a student’s utterance, minus the error” (p. 46). Because there is no indication as to whether the interlocutors’ main attention is drawn to language as an object or to a delivery of message in this definition, reformulated utterances occurred in traditional form-focused instruction can be considered as recasts. In contrast, Long 50 (2006) defined a corrective recast as: a reformulation of all or part of a learner’s immediately preceding utterance in which one or more nontarget-like (lexical, grammatical, etc.) items is/ are replaced by the corresponding target language form(s), and where, throughout the exchange, the focus of the interlocutors is on meaning, not language as object [emphasis original] (p. 77). Because Long specified in this definition that the interlocutors’ primary focus is on meaning, such traditional form-focused reformulations are not regarded as recasts (see Nicholas, Lightbown, & Spada, 2001; R. Ellis & Sheen, 2006, for details). Despite such discrepancies as the above, however, recasts are generally referred to as full or partial reformulation of a learners’ erroneous production without changing the original meaning. A number of studies empirically demonstrated the positive effects of recasts relative to other types of feedback (e.g., R. Ellis et al., 2001; Iwashita, 2003; Long et al., 1998) as well as to no feedback (Z. Han, 2002; Leeman, 2003; Mackey & Philp, 1998), although some research reported that prompts work better than recasts (e.g., Lyster, 1998b; Lyster & Ranta, 1997, Yang & Lyster, 2010). On the whole, however, the recent meta-analyses found that recasts generally prove to be effective in promoting L2 learning (e.g., Lyster & Saito, 2010; Mackey & Goo, 2007). Long (1996) argued the benefits of recasts in that the informational content included in recasts is contextualized, and speakers are likely to be aware of the intended meaning in contexts. Long (2006) further argued that recasts are the ideal interactional feedback because they are unobtrusive; thus, the speakers can attend to errors without breaking the flow of communication. A further claim was made that learners’ ill-form and a correct form through recasts are juxtaposed so that the learners can notice the gap and make cognitive comparisons. In contrast, Lyster (1998a) took the stance that recasts may not be facilitative in learning. Recasts may be 51 interpreted by learners as a response to content rather than form, or alternatives of addressing the identical statement. Lyster claimed that the ambiguity of recasts, which contain both positive and negative evidence, reduces the corrective nature of recasts, and makes recasts less noticeable. Long (2006) also regarded recasts as implicit negative feedback, which can be less salient to learners and may lessen noticing. Yet, it may be premature to draw a conclusion on the efficacy of recasts because various factors, such as research contexts, characteristics of recasts, and linguistic targets, may influence the role and the efficacy of recasts. Recasts are generally found to be more effective in laboratories than classrooms (e.g., content-based immersion programs) (Lyster, 1998a, 1998b, 2004; Lyster & Ranta, 1997; Nicholas et al., 2001). In the laboratories where dyadic interaction takes place in the controlled setting, recasts are focused on learners’ errors on a preselected linguistic item and provided in a consistent and intensive way. Thus, recasts tend to be more salient and easily noticed. In contrast, recasts are generally directed at either an individual learner or a whole class in the classrooms in a less intensive and consistent manner partly because of distractions that may result in decreasing the learners’ noticeability. It is usually the case that it is hard to control classroom interactions involving a number of students. Moreover, research found that shorter recasts with fewer numbers of changes and intonational stress are more explicit and salient, which is more likely to increase the efficacy of such feedback (Egi, 2007a; R. Ellis & Sheen, 2006; Loewen & Philp, 2006; Nicholas et al., 2001; Sheen, 2006). It is also apparent that less salient linguistic targets, such as English articles, are less likely to make recasts effective. In comparison with implicit feedback including recasts, much less attention has been paid to explicit feedback, such as metalinguistic feedback. As noted before, metalinguistic feedback is generally referred to as the provision of metalinguistic information on a learner’s erroneous 52 production. The effectiveness of metalinguistic feedback has been mostly researched in comparison with that of recasts as part of the investigation between implicit and explicit dimension. While some research employed metalinguistic feedback (e.g., R. Ellis, 2007; R. Ellis et al, 2006; Kang, 2009, 2010a, 2010b; Loewen & Nabei, 2007; Lyster, 2004; Lyster & Ranta, 1997), others utilized metalinguistic correction, which may be more toward the explicit end of the continuum (e.g., Li, 2009; Sheen, 2007, 2010b). Metalinguistic correction differs from metalinguistic feedback in that the former entails the explicit provision of a correct form, whereas the latter does not. In R. Ellis and R. Ellis et al.’s work involving past tense –ed, metalinguistic feedback, such as “you need past tense,” was considered sufficient for learners to retrieve their existing knowledge in long-term memory and modify the wrong utterance on their own. On the contrary, in Sheen’s studies focusing on English articles, she noted that affordance of metalinguistic correction was deemed necessary because this is more effective than using metalinguistic feedback. Since English articles are a non-salient feature, it is assumed that simply offering metalinguistic information would not be explicit enough for learners to make a necessary change in their incorrect utterance. As such, explicitness of feedback includes a wide range of feedback types. As a result, there is much work comparing the efficacy of implicit feedback with explicit feedback, employing various operationalizations, contexts, and methodologies. Some included computerdelivered feedback, although conflicting results were reported (see R. Ellis, 2007; R. Ellis et al., 2006, for review). While a number of studies found a superiority of explicit feedback over implicit feedback (e.g., Caroll 2001; Caroll & Swain, 1993; R. Ellis et al., 2006; Nagata, 1993; Rosa & Leow, 2004;), there is some research that did not find a statistical difference (e.g., DeKeyser, 1993; Kang, 2009, 2010a, 2010b; H. Kim & Mathes, 2001; Loewen & Nabei, 2007; 53 Sanz, 2003). Additionally, efforts have been made to synthesize the effects of implicit and explicit feedback in recent meta-analysis research, but they have revealed inconclusive findings. Russell and Spada (2006) attempted to compare the effects of implicit and explicit feedback types on L2 learning, but only Carroll and Swain’s (1993) study was available for analysis due to lack of qualified work. Thus, any generalization of this exploration was not feasible. Mackey and Goo (2007) compared the efficacy of recasts, negotiation, and metalinguistic feedback and observed that recasts yielded larger mean effect sizes than other types of feedback on immediate posttests, short-term delayed posttests, and longer-term delayed posttests. However, again, due to a limited number of qualified studies, it was not considered appropriate to argue over the effects of feedback types. Li (2010a) found a pattern that explicit feedback was more effective than implicit feedback on both immediate and short-delayed tests, but it was reversed on long-delayed posttests. Yet, he failed to find a statistical significance on this observation. There is a body of research that directly examined the relative effects of recasts and metalinguistic feedback per se as part of the comparison between the implicit and the explicit dimension. In short, there has not been a clear answer to this investigation. Some studies revealed a superiority of metalinguistic feedback. Carroll and Swain (1993) compared the effectiveness of four types of feedback (i.e., direct metalinguistic feedback, explicit rejection, recasts, and indirect metalinguistic feedback) in recall production tasks following each feedback session to explore the dative verb development of low-intermediate ESL learners. The findings indicated that direct metalinguistic feedback was the most effective, and the rest of the experimental groups did not make a statistically significant difference. Carroll (2001) compared the same types of feedback used in Carroll and Swain study to look at low-intermediate ESL learners’ ability to form nouns from verbs. It was revealed that only indirect prompting as well as 54 explicit metalinguistic information helped make learners form a generalization, whereas recasts were not effective in making a generalization. R. Ellis et al. (2006) conducted an experiment to look into the relative efficacy of recasts and metalinguistic explanation in two communicative tasks with a focus on English past tense -ed. They also measured learners’ implicit knowledge (i.e., oral imitation test) and explicit knowledge (i.e., untimed grammaticality judgment test and metalinguistic knowledge test). Overall, a clear advantage for metalinguistic explanation over recasts was found on the both delayed imitation and grammaticality judgment posttests. Using recasts and metalinguistic correction, Sheen (2007) researched a differential effect on learners’ accurate use of English articles in her quasi-experimental study. She found that metalinguistic correction was superior to both recasts and no feedback, but there was no statistical difference between the recast and the control group. Still, others did not find a significant difference in the relative effects between the two types of feedback. H. Kim and Mathes (2001) replicated Carroll and Swain’s (1993) study, but included only explicit metalinguistic feedback and recasts. It was found that there was no statistical difference between the two sorts of feedback. Loewen and Nabei (2007) examined the effects of recasts, metalinguistic feedback, and clarification requests and measured EFL learners’ implicit knowledge (i.e., oral production test and timed grammaticality judgment) and explicit knowledge (i.e., untimed grammaticality judgment). They found that there was no statistical difference among the three feedback types on the timed grammatical judgment test, and no effects of feedback were found on the untimed grammaticality judgment test or the oral production test. As they acknowledged, the lack of significant differences may be ascribed to the insufficient treatment time and number of feedback provided. Lyster and his associates undertook a body of studies probing the effects of recasts 55 compared with prompts in the classrooms (e.g., Ammar & Spada, 2006; Lyster, 1998, 2004; Lyster & Ranta, 1997; Lyster & Saito, 2010; Yang & Lyster, 2010). However, as prompts incorporate both implicit and explicit feedback, it was not possible to find out which of these two feedback types contributed to the effects of prompts. Thus far, most studies on the relative effects of recasts and metalinguistic feedback within the implicit-explicit aspect, have addressed morphosyntactic development. Yet, it may be necessary to take into consideration other studies that go beyond the grammar-focused research. The current study seeks answers to the development of Korean honorifics, which concerns a pragmatic aspect. In the subsequent section, the efficacy of instruction in instructed SLA is briefly discussed, and the effects of instructional strategies in the field of L2 pragmatic development follow. Implicit vs. Explicit Instruction in Instructed SLA and L2 Pragmatics Within the instructed SLA research paradigm, Norris and Ortega’s (2000) groundbreaking meta-analysis measuring the effectiveness of L2 instruction attested a superiority of explicit instruction over implicit instruction. Lately, Spada and Tomita (2010) conducted a metaanalysis study and accrued further evidence that explicit types of instruction are more effective than implicit types for both simple and complex features. These research synthesis studies are largely based on learners’ grammatical development. Nonetheless, it is claimed that the development of L2 pragmatic knowledge is not paralleled to that of L2 grammatical knowledge, and pragmatic knowledge is intrinsically different from grammatical knowledge (Bardovi-Harlig, 1999; Bardovi-Harlig & Hartford, 1990; Koike, 1989). For this reason, it may be necessary to review research with a specific focus on instructional effects in interlanguage pragmatic development relevant to the present study. 56 In comparison with other language areas, including grammar, less attention has been paid to pragmatics and even less to instructed pragmatics in SLA. It may be attributed to the claim that L2 pragmatics is a relatively neglected area in SLA (Rose & Kasper, 2001). Yet, literature indicated that exposure alone is not sufficient for successful L2 pragmatic development (Bardovi-Harlig & Hartford, 1993; Bouton, 1996) especially with a goal of attaining a nativelike mastery. A series of research was conducted to explore the role of instruction in learners’ pragmatic development and attested that learners evidently accelerate their pragmatic development by way of instruction (e.g., Billmyer, 1990; Bouton, 1994; Halenko & Jones, 2011; Lyster, 1994; Yoshimi, 2001). Notwithstanding a growing interest in the effects of pragmatic instruction in the last decade, only a small number of studies that directly probed the relative effects of pragmatic intervention are available, most of which explored the implicit and/or explicit dimension. Generally, explicit instruction relates rule formation; in contrast, implicit instruction involves the absence of it (DeKeyser, 2003). Alternatively, the former entails learners’ metalinguistic awareness of rules; the latter involves their inference of rules without awareness (R. Ellis, 2008). In fact, the division of this implicit and explicit dimension is not clear-cut. As with interaction research, it should be noted that explicitness of instruction in pragmatics research is conceptually considered as a continuum, not as dichotomy. Pragmatic intervention studies involve a wide array of instructional techniques that range from the implicit end of the continuum to the explicit end. Explicit instruction typically includes metapragmatic and/or metalinguistic information, conscious-raising of metapragmatic rules, explanation and discussion of rules, prescribed speech act formulae, and metapragmatic judgment tasks, whereas implicit instruction contains simple exposure to pragmatic examples. Unlike interaction research, however, the investigation into the 57 nature of instruction in L2 pragmatics has focused on explicit instruction (e.g., Halenko & Jones, 2011), and scant attention has been paid to implicit instruction. It may be that since pragmatic functions and associated contextual factors are often not salient to learners and are likely to reduce noticeability in spite of lengthy exposure (Kasper & Rose, 2002), explicit instruction may have been of special interest for researchers. As far as the instructional effects of L2 pragmatic development are concerned, E. Jeon and Kaya (2006) in their meta-analysis reported that the results were not conclusive. It was found that explicit instruction yielded clearly larger effects than implicit instruction based on seven studies under investigation. Nevertheless, they suggested that it is premature to draw a conclusion at this point because instructional techniques often lean towards the explicit end of the implicit-explicit continuum, and too much discrepancy appears to exist in terms of operationalization of implicit and explicit instruction among studies in addition to too small number of primary studies investigated. On the whole, however, there seems to be a trend that explicit instruction is more effective than implicit instruction across diverse methods and contexts in L2 pragmatic development. For instance, Rose and Ng Kwai-fun (2001) explored the effects of inductive (i.e., implicit) and deductive (i.e., explicit) approaches in teaching English compliments and compliment responses to English learners in Hong Kong. The deductive group received metapragmatic information through explicit instruction prior to practice activities, whereas the inductive group was engaged in pragmatic analysis activities in which learners were expected to make generalizations of compliment formulas on their own. A self-assessment task, a DCT, and a metapragmatic assessment task were utilized in a pretest-posttest design. The results found a positive effect only for the deductive group in compliment responses. Although both inductive 58 and deductive instruction may promote pragmalinguistic proficiency, only deductive instruction may work to develop sociopragmatic proficiency. Takahashi (2001) looked into the efficacy of input enhancement on Japanese EFL learners’ English requests and employed four different input conditions: explicit teaching, form-comparison, form-search, and meaning-focused conditions. In a pretest-posttest design, an open-ended DCT as well as written immediate retrospective self-report data was utilized. She found that explicit teaching had the strongest impact, followed by the other three implicit conditions in the order of form-comparison, formsearch, and meaning-focused. Tateyama (2001) compared the effectiveness of implicit and explicit instruction with a focus on attention getters, expressions of gratitude, and apologies to learners of Japanese as a foreign language. The explicit group received metapragmatic information, whereas the implicit group withheld it. Learners were engaged in role-play, multiple-choice tasks, and self-report. Although no statistical difference between the two groups was obtained, closer inspection of the multiple-choice tasks revealed that learners who received explicit instruction outperformed the other counterpart in the areas that required higher formality, indebtedness, and severity of offense of the linguistic expressions. It is construed that explicit instruction may be more beneficial for learners in selecting routine formulas relevant to age, social status, and in-group and out-groupness. Alcón Soler (2005) investigated how different instructional paradigms influence learners’ knowledge and ability to use request strategies in EFL context. Explicit instruction constituted direct awareness-raising tasks and written metapragmatic feedback; implicit instruction was composed of typographical enhancement and implicit awareness-raising tasks. The results demonstrated an advantage of the explicit instruction over the implicit one. Nguyen, T. Pham, and M. Pham (2012) evaluated the effects of implicit and explicit form-focused instruction on the development of the speech act set of 59 constructive criticism of Vietnamese EFL learners. Explicit instruction constituted consciousness-raising activities, explicit metapragmatic explanation, and correction of errors of forms and meanings; implicit instruction comprised pragmalinguistic input enhancement and recasts. Learners’ performance was measured using a DCT, a role play, and an oral peer feedback task on the pretests and posttests. They found clear evidence that the explicit instruction worked significantly better than the implicit instruction on all measures. Taken together, although the results were inconclusive, ample evidence in favor of explicit instruction is available in L2 pragmatic development. The current study intends to discover if this tendency is also applicable to the learning of Korean honorifics within the pragmatic domain. Research Questions and Hypotheses The purpose of this dissertation is to investigate the effects of interactional feedback and factors that may play a role in L2 pragmatic development through conversational interaction. More specifically, it examines whether and how learners’ HL background and explicitness of feedback affect the development of L2 Korean honorifics during conversational interaction. The independent variables include learners’ HL background (i.e., HL learners and NHL learners), feedback types with differing explicitness (i.e., implicit and explicit feedback), and tests with three levels (i.e., pretest, immediate posttest, and delayed posttest). The dependent variable is the use of target Korean honorific features in learners’ oral production. This study seeks answers to the following research questions, and the hypotheses are proposed according to existing literature: 1. Does interactional feedback promote L2 pragmatic development? 60 Yes, it is predicted that interactional feedback facilitates L2 pragmatic development. Soler (2002) and Fukuya and Hill (2005) reported that interaction was effective in learning pragmalinguistic conventions of request. Also, K. Jeon (2007) found that interactional feedback facilitated L2 pragmatic development, specifically subject-verb agreement in Korean subject referent honorifics. Considering the facilitative role of interaction found in these studies, interactional feedback may also promote the development of Korean subject and object referent honorifics in the context of HL and L2 learning in this study. 2. How does learners’ HL background affect L2 pragmatic development? NHL learners are expected to learn Korean honorifics better than HL learners, although both groups of learners may find it difficult to learn honorifics due to negative L1 transfer (K. Jeon, 2007). As a result of incomplete acquisition, many HL learners may show fossilization, simplification, and/or overgeneralization (Montrul, 2010a) in the HL, which is likely to impede the learning of honorifics. Montrul (2008b, 2010a) argued that specialized vocabulary, forms of address, and honorifics in East Asian languages either are missing in HL learners or remain imperfectly acquired. She further claimed that Korean honorifics are a major pedagogical concern in teaching Korean as a HL. Thus, the NHL learners may learn Korean honorifics better than the HL learners. 3. How does explicitness of feedback affect L2 pragmatic development? Explicit feedback is anticipated to be more effective than implicit feedback. Although there has been no published study looking at the relative effects of implicit and explicit “feedback” in L2 pragmatic development within interaction research, many studies within L2 pragmatics research have reported that learners who received explicit “instruction” outperformed those who received implicit “instruction” (e.g., House, 1996; Takahashi, 2001; Tateyama et al., 1997). As such, 61 explicit feedback may be more effective than implicit feedback in interaction-driven L2 pragmatic development in this study. 4. How does explicitness of feedback affect L2 pragmatic development of learners with different HL background? It is predicted that explicit feedback is more effective than implicit feedback for the NHL learners; however, it is unknown as to which type of feedback works better for the HL learners in learning Korean honorifics. There have been hardly any studies directly investigating the relative effects of implicit and explicit feedback in the context of HL and L2 learning. In her recent dissertation, Y. Han (2010) made an attempt to explore this issue, but did not find a significant difference between implicit and explicit feedback in HL and NHL learners’ development of Chinese classifiers. Montrul (2010a) asserted that NHL learners learn the target language as an adult by rule-based instruction in a classroom and maximize the use of metalinguistic and explicit knowledge. Thus, they are likely to benefit more from explicit feedback than from implicit feedback. In contrast, HL learners acquire the target language as in L1 acquisition implicitly and through access to UG in childhood prior to a critical period. Once in the classroom, however, it remains unclear with regard to whether the HL learners regain the target language as L1-acquiring children or depend on explicit learning like adult L2 learners. She further stated that this is perhaps the biggest question as to which type of classroom instruction works better for the HL learners. 62 CHAPTER 3 METHOD Chapter 2 addressed theoretical background behind the present study. The major variables underlying this study were introduced with reference to prior research. Issues that need to be further investigated in future research were identified, and rationale for the current investigation was provided. This chapter describes the characteristics of the study participants, operationalizations of the variables, linguistic targets, tasks and materials that elicited learner language production, design, procedure, and administration of the research, coding and scoring protocols, interrater reliability, and methods of statistical analyses to pursue the research agenda. Participants A total of 78 English-speaking learners of Korean as a foreign language at two large public universities in the United States participated in this study; of these, 51 participants were drawn from University of California, Berkeley, and 27 were from University of Minnesota. Among these 78 participants, 38 were HL learners and 40 were NHL learners. There were 54 female and 24 male students, and they ranged in age from 18 to 29 with an average of 20. There were seven freshmen, 20 sophomores, 28 juniors, 20 seniors, and three doctoral students. Their majors were fairly varied: political economy (n = 16); biology (n = 9); business (n = 9); nutritional science (n = 3); economics (n = 3); communications (n = 3); psycholgy (n = 2); social welfare (n = 2); computer science (n = 2); Asian studies (n = 2); statistics (n = 2); architecture (n = 2); physiology (n = 2); rhetoric (n = 2); undecided (n = 4); and others (n = 15). Seventy two participants reported that they had learned a foreign language other than Korean, most of which were Spanish and French, followed by Chinese, Japanese, and German. Much effort was made to recruit participants who possessed a similar level of Korean language proficiency. The majority of the participants were enrolled in the intermediate-level 63 (i.e., second-year) Korean language classes: (a) 66 students were enrolled in the second-year (3 semester) classes at the time of data collection; (b) four students completed the first-year (2 rd nd semester) classes in the prior semester, but were not enrolled in Korean language classes at the th time of data collection; (c) four students finished the second-year (4 semester) classes in the prior semester, but were not taking Korean language classes at that time; and (d) four students had not been enrolled in any Korean language classes. The last four students in (d) were HL learners who had acquired Korean in informal contexts (e.g., home and Korean communities) and were recruited outside the Korean classes on campus. Their pretest scores in producing target Korean referent honorifics were comparable to the average pretest scores in the study. Their verbal abilities to perform communicative tasks in Korean appeared to be at a similar level, approximately intermediate-level, to the rest of the participants as well. Thus, they were deemed to possess a similar proficiency distribution to the rest of the participants and so were included in the study. All the learners scored lower than 70 % on the pretests of Korean referent honorifics; hence, possible ceiling effects were not observed. The researcher served as a Korean NS interlocutor in NS-HL/NHL learner dyad interactions. The language background questionnaire used in this study, and detailed background information of each participant are presented in Appendix A and B, respectively. Heritage Language Learners While HL learners are generally described as heterogeneous (Kagan & Dillon, 2012; Kanno et al., 2008; Kondo-Brown, 2005; Valdés, 1995), HL learners with a relatively homogeneous language background were selected for this study. All the HL learner participants were second-generation Korean-Americans who grew up with exposure to the Korean language 64 at home, and both of their parents were Korean NSs. Among these 38 HL learners, 33 participants were born in the United States. Five were born in Korea, but moved to the United States before the age of 1. The HL participants were first identified based on the language background questionnaires (see Appendix A), and then the researcher checked with each student to confirm his or her HL background. The HL participants’ comfort with six areas of Korean language skills were self-rated and shown in comparison with the NHL counterparts. The results were in line with the findings from HL acquisition literature (J. Lee, 2005; Montrul, 2008b): The HL learners generally reported their stronger confidence in speaking and listening abilities than reading, writing, grammar, and vocabulary abilities (see Appendix C). In addition, a reduced amount of interaction in Korean was generally shown in comparing the HL learners’ childhood and current interaction. Most HL learners indicated that they had interacted in Korean more with their mother than with their father. While they had interacted with their grandparents mainly in Korean, their interaction in Korean with their siblings and friends was minimal (see Appendix D). In terms of current uses and exposure of Korean, many HL learners indicated that they sometimes enjoyed Korean movies and music, but not reading or writing in Korean (see Appendix E). Among 38 HL learners, 20 learners indicated Korean as their L1, while 15 learners considered both Korean and English as their L1. Three HL learners specified English as their L1. Twenty five HL learners have learned Korean at other non-university institutions in the United States, mostly at Korean church-sponsored Saturday schools prior to the age of 14. Three HL learners have studied abroad in Korea for language training at universities for less than four months. Thirty four HL learners have been to Korea other than the purpose of studying Korean for the duration of up to three months. Most of the HL learners indicated that the purpose of the visits was to meet relatives. 65 Non-heritage Language Learners While most of the NHL learner participants were English NSs, some variations were existent in terms of the NHL learners’ L1 background: English (n = 21); English and other language(s) (n = 9); Chinese (n = 5); Japanese (n = 2); Spanish (n = 1); Hmong (n = 1); and Cantonese and Vietnamese (n = 1). In light of the NHL participants’ comfort with the six areas of Korean language skills, they indicated a stronger confidence in reading and writing than speaking abilities (see Appendix C). In terms of current uses and exposure of Korean, many NHL learners reported that they often enjoyed Korean movies and music, but not reading or writing in Korean (see Appendix E). Only three students have learned Korean at other nonuniversity institutions, for example, private language schools and tutors in the United States, and it was after the age of 16. Four NHL learners have studied abroad in Korea for language training at universities for less than four months. Eight NHL learners have visited Korea other than learning Korean up to three months. Most of the NHL learners were to visit Korea for sightseeing. Operationalizations Explicit Feedback The notion of explicitness is generally viewed as a continuum in SLA. As shown in Chapter 2, this study employed two feedback types that are at the extreme end of feedback taxonomy in order to find the best effects of feedback use. In some studies, metalinguistic information without the provision of a correct form was supplied to learners as explicit feedback (e.g., Carroll & Swain, 1993; R. Ellis et al., 2006). In R. Ellis et al.’s study, for instance, since learners had already learned English regular past tense form -ed, the researchers reminded the learners that the past tense had been incorrectly used. Then, the learners simply applied the rule 66 to regular verbs to correct their erroneous form. Thus, a reminder of the rule may have been sufficient for correction to occur. However, in this dissertation, explicit feedback was operationalized as metalinguistic correction, which provides a targetlike form, accompanied by metalinguistic explanation on a learner’s nontargetlike utterance. Because honorifics could be confusing and complicated, unlike English regular past tense form -ed, simple rule explanation may not have sufficed for learners to correct their nontargetlike form. Therefore, providing the targetlike form in addition to the metalinguistic explanation was considered appropriate as explicit feedback in this study. The following example (6) shows how metalinguistic correction is provided in this study. Example 6. Metalinguistic Correction Learner: Sensayngnim-i (-ka) kong-ul chayo. Teacher-S ball-DO kick-V A teacher kicks a ball. NS: Eluney tayhay malhalttaynun -kkeyse kuliko ‘chata’ey -si- lul tehayse ‘cha-sey-yo’lako hayyo. When you talk about a person who is older than you, use -kkeyse and ‘chaseyyo’ by adding si to ‘chata.’ Implicit Feedback Implicit feedback was operationalized as recasts, which reformulate a learners’ nontargetlike utterance without referring to the source of the nontargetlike form. The following example (7)-(8) illustrates recasts provided in this research. 67 Example 7. Recasts Learner: Sensayngnim-i (-ka) kong-ul chayo. Teacher-S ball-DO kick-V A teacher kicks a ball. NS: A, sensayngnim-kkeyse kong-ul cha-sey-yo.  declarative sentence Teacher(H)-S(H) ball-DO kick-V(H) (falling intonation) Oh, the teacher kicks the ball. Example 8. Recasts Learner: Sensayngnim-i (-ka) kong-ul chayo. Teacher-S ball-DO kick-V A teacher kicks a ball. NS: A, sensayngnim-kkeyse kong-ul cha-sey-yo?  interrogative sentence Teacher(H)-S(H) ball-DO kick-V(H) (rising intonation) Oh, does the teacher kick the ball? As shown above, recasts were provided using a complete sentence in either a declarative or an interrogative form in order to make the NS interlocutor’s response as natural as possible during conversational interaction. The corrective part in recasts was not presented with prosodic emphasis, such as stress and intonation, to make the feedback as implicit as possible. Heritage Language Learners Perhaps the most widely used definition of HL learners in HL acquisition literature was adopted in this study because it was considered to best reflect the current Korean HL learner 68 population in the United States. HL learners were operationalized as those who are “raised in a home where a non-English language is spoken, who speaks or at least understands the language, and who is to some degree bilingual in that language and in English” (Valdés, 2001a, p.38). As indicated in this operationalization, the HL learners who had exposure to the HL language at home and who were actually able to use the HL language, though limited, as well as English were included in this study. In other words, those learners whose HL use was generation(s) removed and whose association with the HL language was limited to cultural aspects of heritage were excluded. In this dissertation, all the HL learner participants were second-generation Korean Americans whose parents were both NSs of Korean and who experienced input and interaction in Korean at home during childhood. A small number of HL learner participants who arrived in the United States as infants were deemed to have comparable HL language background to the rest of the HL learner participants and were included in this study. Non-heritage Language Learners As opposed to the HL learners, NHL learners were operationalized as L2 learners who had no prior exposure to the Korean language at home in childhood. The NHL learners were typical adult L2 learners in the foreign language classrooms. Linguistic Targets Korean honorifics are the most systematic and complex among all known languages, and this renders learning L2 Korean difficult (Sohn, 1999). It is not uncommon to observe that even advanced-level learners of Korean have difficulties in using Korean honorifics properly, and that their problems are persistent even after prolonged instruction and residence in Korea. Thus, teaching Korean honorifics poses a great challenge to teachers of Korean (Byon, 2000). Also, learning Korean honorifics can be particularly difficult for learners whose primary language is 69 English because English does not have such an honorific system. Hence, learning Korean honorifics is one of the biggest barriers for English-speaking learners of Korean. For these reasons, Korean honorifics were selected as a linguistic target in this study. Particularly, the study focuses on the effects of interactional feedback on the learning of Korean referent honorifics, consisting of subject and object honorifics, because of the complexity of the Korean honorific system and the limited amount of treatment for each learner. It should be recalled that referent honorifics occur when a subject, a direct object, or an indirect object in a sentence refers to an honorific person. More specifically, subject honorifics are used when the subject in the sentence refers to the honorific person to the speaker; object honorifics are employed when the object refers to the honorific person to the subject. Subject Referent Honorifics In the area of subject honorifics, the honorific subject case marker –kkeyse and the honorific verbal morpheme –si in the honorific sentence are the focus. As repeated in Chapter 2, the example (9) of subject non-honorifics shows contrast to the example (10) of subject honorifics below. Example 9. Subject Non-honorifics Haksayng-i (-ka) kong-ul chanta. Student-S ball-DO kick-V A student kicks a ball. Example 10. Subject Honorifics Sensayngnim-kkeyse kong-ul cha-si-nta. 70 Teacher(H)-S(H) ball-DO kick-V(H) A teacher kicks a ball. Object Referent Honorifics In the case of object honorifics, the honorific indirect object case marker –kkey and the humble verbal form (-e)tulita in the honorific sentence are the focus. As reported in Chapter 2, the example (11) of object non-honorifics displays contrast to the example (12) of object honorifics. Example 11. Object Non-honorifics Haksayng-i (-ka) Student-S chinkwu-eykey (-hanthey) friend-IO chayk-ul cwunta. book-DO give-V A student gives a book to a friend. Example 12. Object Honorifics Haksayng-i (-ka) sensayngnim-kkey chayk-ul Student-S teacher(H)-IO(H) book-DO tuli-nta. give-V(Hum) A student gives a book to a teacher. Feedback is the most effective when learners already have some knowledge about the linguistic target but have not fully acquired yet (Z. Han, 2002; Mackey & Philp, 1998). Also, a comparable level of Korean referent honorific knowledge as well as Korean proficiency is desired for all the participants in order to make their starting point prior to treatment as close as 71 possible. Thus, an effort was made to ensure that the participants had learned Korean referent honorifics prior to data collection. It was confirmed with the instructors that the great majority of the participants had learned Korean referent honorifics via textbooks throughout their first year in the Korean language programs. Instruments The language background questionnaire was adapted from H. H. Kim (2005) and used for this study. The photos of the researcher’s own family members (e.g., parents, grandparents, nieces, nephews, and other relatives) and acquaintances (e.g., teacher and her student) taken in Korea were used to elicit Korean subject and object referent honorifics (see Appendix F and G). Since pragmatics is highly context-sensitive, using the photos of the researcher’s family and acquaintances in Korea was deemed appropriate to provide authentic contexts. Individuals of differing age (e.g., grandparent vs. grandchild) and social status and age (e.g., teacher vs. student) in various situations (e.g., school, visiting grandparents, and Hanna’s home) were chosen to appear in the photos so as to provide rich contexts where target honorific and/or nonhonorific items could be appropriately used. For instance, each photo included one of the following scenes: (a) an adult’s action; (b) interaction between an adult and a child; (c) a child’s action; (d) interaction between children; and (e) scenery. Because the focus of this study is on subject and object honorifics, each of (a) and (b) was used to elicit subject honorifics and object honorifics, respectively, and (c)-(e) served as distractors. Interaction between adults (e.g., adults of similar age and/or social status) was not included because it often involves more subtle and complicated relationships and may require more complex use of honorifics than other types of interaction described above. Thus, it was beyond the scope of this study. Because the use of photos as a data elicitation method makes it difficult to elicit intended 72 stative verbs, only action verbs were selected. This study also employed photos that require learners to make use of action verbs that are commonly used in daily lives and taught in Korean classes. This was intended to minimize the participants’ potential lack of vocabulary knowledge in order to describe the photos as well as to provide both the HL and the NHL learners who might possess different lexical abilities with a fair opportunity. In addition, much effort was made to choose photos that are intended to elicit regular verbs. Rationale behind the exclusive selection of the regular verbs in this study is that the HL learners are known to be better capable of utilizing a proper predicate inflection on irregular verbs than the NHL learners, if the verbs are commonly used at home. The HL learners are generally exposed to the HL input at home, which contains various irregular verbs with a proper inflection. Therefore, using irregular verbs with a proper inflection may be already automatized in the HL learners’ lexicon. The irregular verbs with a proper inflection may be available in the form of a language chunk, and they may not even aware of predicate inflection rules. In contrast, the NHL learners were anticipated to have difficulties in applying the rules to produce irregular verbs with a proper predicate inflection. It may be particularly hard to do so during online processing and production because higher cognitive demands are required. For these reasons, it was considered fair to include only regular verbs that are frequently encountered both at home and in class in order to avoid the potential bias. Much effort was also made to exclude lexicalized honorific verbs and nouns (i.e., special polite verbs and nouns) because lexical honorification is not the focus of the study. However, it was considered inevitable to include the lexicalized humble verb form (-e)tulita ‘to give’ as an exception since object honorifics that involve an indirect object and a direct object in a sentence often require the use of the humble form. There are few other predicates that show humbleness 73 of a subject in object honorifics; for example, mosita ‘to accompany,’ poypta ‘to meet,’ and yeccwuta ‘to ask.’ Yet, it was considered difficult to successfully elicit oral production of these humble verb forms using photos, and thus they were excluded in the current study except for (e)tulita ‘to give.’ Pretest, Immediate Posttest, and Delayed Posttest Tasks and Materials Each test was composed of a one-way photo description task (see Appendix F). Both the learner and the Korean NS interlocutor shared a photo album including a set of photos. In the task, the NS interlocutor first introduced herself, then her family members and acquaintances in the photos and asked the learner to describe what each person in the photos was doing (see Appendix H). The learner was anticipated to produce sentences with subject honorifics (e.g., an adult’s action), object honorifics (e.g., interaction between an adult and a child), and/or nonhonorifics (e.g., a child’s action, interaction between children, and scenery) in Korean while looking at the photos. There were three different versions of this task (e.g., school, visiting grandparents, and Hanna’s home). The order of the versions was randomized to control for any ordering effect. Each version of the task was designed to elicit three types of honorifics/nonhororifics: approximately eight subject honorifics, eight object honorifics, and sixteen distractor non-honorifics including scenery scenes. Those learners who scored lower than 70 % on the pretests were expected to participate in the treatments and posttests to avoid ceiling effects. Treatment Tasks and Materials For the treatment materials, two different kinds of tasks were utilized (i.e., one-way tasks and two-way tasks) due to the possibility of task effects. Because research indicated that task types may affect interaction and learning outcomes (Duff, 1986; Pica, Holliday, Lewis, & Morgenthaler, 1989; Pica, Kanagy, & Falodun, 1993), both one-way and two-way tasks were 74 included to minimize this influence. The treatment tasks contained two versions of a one-way information gap task (e.g., amusement partk and New Year’s Day) and two versions of a twoway information gap task (e.g., department store and birthday party) (see Appendix G). The one-way task was the same as the photo description task used in the testing materials except that different versions of the task were used and that the NS interlocutor provided negative interactional feedback, in the form of either implicit feedback or explicit feedback, on the participants’ nontargetlike utterance. Each version of the task was designed to elicit three sorts of honorifics/non-hororifics: approximately eight subject honorifics, eight object honorifics, and sixteen distractor non-honorifics including scenery scenes. The two-way task included the story sequencing task that the learner and the NS interlocutor each had six photos with a sequential order. Both the learner and the NS interlocutor described the photos to each other and worked together to identify the order of the photos. Then, the learner was asked to create a story based on the 12 photos (see Appendix I). The learner was expected to produce sentences with subject honorifics (e.g., an adult’s action), object honorifics (e.g., interaction between an adult and a child), and/or non-honorifics (e.g., a child’s action, interaction between children, and scenery) in Korean while engaged in the task. The NS interlocutor afforded negative interactional feedback, in the form of either implicit feedback or explicit feedback, on the participants’ nontargetlike utterance. Each version of the task was designed to elicit three types of honorifics/non-honorifics: approximately three subject honorifics, three object honorifics, and six distractor non-honorifics. As research showed that learners could benefit from positive evidence (i.e., input) available in non-feedback turns as well as from negative evidence (i.e., interactional feedback) (Iwashita, 2003), the two-way task in the present study was designed to allow positive evidence not from non-feedback turns but from feedback 75 turns. Thus, the photos to elicit subject and object honorifics were given to the learner, whereas those photos to elicit non-honorifics were provided for the NS interlocutor. This was intended to control the NS’s modeling (i.e., positive evidence) of honorific use since this study is to examine the effects of interactional feedback only, not modeling. An example of treatment sessions for a NHL learner in the implicit feedback condition in this study is transcribed and presented in Appendix J. Design and Procedure Research Design The research design is shown in Table 4 below. The study employed a traditional pretestimmediate posttest-delayed posttest design and featured both control and experimental groups. Based on the learners’ HL background and the feedback types, the participants were randomly assigned to four different experimental conditions and a control group: (a) HL learner & implicit feedback group (n = 16); (b) HL learner & explicit feedback group (n = 15); (c) NHL learner & implicit feedback group (n = 16); (d) NHL learner & explicit feedback group (n = 16); and (e) control group (n = 15). Table 4. Research Design Group Experimental Experimental Experimental Control group 1 group 2 group 3 group 4 group (n = 16) (n = 15) (n = 16) (n = 16) (n = 15) HL Factor Experimental HL NHL NHL HL HL (n = 7) background NHL (n = 8) 76 Table 4 (cont’d) Explicitness Implicit Explicit Implicit Explicit No feedback of feedback Research Procedure The research procedure is outlined in Table 5 below. In session 1, all the participants completed a language background questionnaire (see Appendix A) and performed a pretest and a treatment 1. In session 2, they carried out a treatment 2 and an immediate posttest. The first two sessions were adminstered within a week. Two weeks after the immediate posttest, the delayed posttest was administered. Each of the test sessions lasted about 8-22 minutes for the pretest, 925 minutes for the immediate posttest, and 8-26 minutes for the delayed posttest and was based on the one-way photo description task, whereas each of the treatment sesssions took approximately 23-46 minutes for the photo description task and 14-32 minutes for the story sequencing task. It should be noted that the time each participant took to complete the identical task varied. Because the HL learners were more fluent than the NHL learners, the former generally finished the task more quickly than the latter. Detailed information of average time learners in each of the four experimental groups spent in each task is summarized in Appendix K. In NS-learner dyads, they performed a series of communicative tasks, which served as a pretest, treatments, and two posttests, and the entire sessions were audio-recorded for later transcription. There were two treatment sessions: treatment 1 (i.e., one-way photo description task or two-way story sequencing task) and treatment 2 (i.e., one-way photo description task or two-way story sequencing task). The one-way task and the two-way task each involved two activities. Each treatment session included two activities in order for all the participants to be exposed to a 77 relatively comparable amount of input and interactional feedback regardless of the time each participant took to finish during each session given that the HL learners were expected to complete the tasks at a much faster rate than the NHL learners due to their superior fluency. The order of the two treatment sessions was counterbalanced, whereas the order of the activities within each treatment was randomized to control for any ordering effect. During the treatment sessions, the NS interlocutor provided each participant with interactional feedback in the form of either implicit or explicit feedback on the nontargetlike utterances that involved subject and object referent honorofics. On occasion, however, feedback was given to other linguistic items which are not the focus of this study to serve as a distractor as well as to simulate a real-life conversation. Both implicit and explicit feedback were given in Korean. Table 5. Research Procedure Week Day 1 Procedure 1 Task Background questionnaire Pretest Treatment 1 3 Story sequencing task (14-32 min.) Immediate posttest 1 Photo description task (23-46 min.) Treatment 2 1 Photo description task (8-22 min.) Photo description task (9-25 min.) Delayed posttest Photo description task (8-26 min.) 1 2 3 Since pragmatics is context-dependent, it was considered necessary to provide as natural and authentic contexts as possible. Therefore, the NS interlocutor read task instructions to each participant as if they had been engaged in a normal conversation. The intent was for them to 78 converse with each other while looking at photo albums in a relatively natural setting. Also, the instructions were offered in English to ensure that all the participants could comprehend what they were expected to do completely to perform the tasks (see Appendix H and I). To simulate an authentic conversation, the NS interlocutor provided the participants with Korean vocabulary, except honorifics, and explained it to them when they did not know and/or when they asked her its meaning to be able to complete the tasks. Also, both the NS interlocutor and the participants were allowed to divert from the tasks briefly in order to discuss non-honorifics-related issues in the photos. For example, the researcher’s family home in Korea, and interesting Korean culture related to the photos, instead of solely focusing on what an adult and/or a child were doing in the photos that involved the use of honorifics. Coding and Scoring Coding and scoring guidelines for Korean subject and object referent honorifics were established in consultation with a professor in the Korean language program. Every effort was made to provide the coding and scoring scheme that accommodates different linguistic behaviors of both the HL and the NHL learners so that it would be fair for both groups of learners in measuring their honorific development. The first step was to identify obligatory contexts for the use of subject honorifics (i.e., the honorific subject case marker -kkeyse(nun) and the honorific verbal morpheme –si). Once the total numbers of obligatory contexts were coded and counted, the learner utterances were scored. The scoring of subject honorifics focused on the use of the honorific subject case marker kkeyse(nun) in the honorific subject and the honorific verbal morpheme –si in the verbal ending. For subject honorifics, one point was given when a learner produced the honorific subject case marker -kkeyse(nun), and another one point was awarded for the production of the honorific 79 verbal morpheme –si in an honorific sentence. There was no penalty for the inaccurate inflection in the predicate. Zero point was given if the learner used no target honorific features. Then, all these points earned were combined to compute a total score for subject honorifics. Once the scores were obtained, the ratio of the scores of the use of the target honorific features to the total number of obligatory contexts identified for the honorific subject case marker -kkeyse(nun) and the honorific verbal morpheme –si were calculated for subject honorifics. The scoring of object honorifics focused on the use of the honorific indirect object case marker -kkey in the honorific object and the humble verbal form -(e)tulita in the verbal ending. The coding and score procedures for object honorifics were the same as those of subject honorifics. The detailed coding and scoring procedures are presented with corresponding examples of subject and object honorifics, respectively (see Appendix L). Interrater Reliability The researcher, who was a Korean NS, had a master’s degree in English education and a professional certificate in teaching Korean as a foreign language. She was completing a doctoral degree in second language studies. Another Korean NS who held a master’s degree in language education served as a second rater. The researcher as well as the second rater coded Korean referent honorifics data to assess consistency between the two raters. The researcher coded all data, and the second rater coded 25 % of the data that were randomly selected. The second rater received a training session from the researcher. The researcher explained the coding and scoring guidelines and showed examples of coded and scored sentences to the second rater. Also, the second rater read the guidelines thoroughly and asked the researcher questions. A small set of data was given to the second rater so that she could practice prior to coding and scoring. However, this practice set was not included to calculate interrater reliability. When inconsistent 80 coding occurred between the researcher and the second rater, they resolved the differences through exhaustive discussions. As a result, high agreement rates were achieved. There was a 97.13 % and a 99.40 % agreement rate for identifying subject honorifics and object honorifics, respectively, in the data. A 97.97 % and a 99.68 % agreement rate was obtained for coding and scoring the percentage based on the use of the honorific targets for subject honorifics and object honorifics, respectively, in learner production. Data Analyses This study investigates whether and how learners’ HL background and explicitness of interactional feedback during conversational interaction affect the L2 development of Korean subject and object referent honorifics. In order to find the mean frequency of errors occurred and feedback provided per learner in each experimental group in this study, twenty participants (i.e., five participants in each of the four experimental groups) were randomly selected for analysis. A total number of error occurrences and feedback moves on Korean referent honorifics during the two treatment sessions were counted, and then the mean frequency per learner was computed. Different statistical analyses were conducted to answer research questions. Descriptive statistics were based on percentage of the use of the target honorific features learners produced across the pretest-immediate posttest (i.e., posttest 1)-delayed posttest (i.e., posttest 2). The detailed descriptive statistics are reported according to each of the sub-target features: (a) –kkeyse; (b) -si; (c) –kkey; and (d) -(e)tulida. However, it should be noted that only total percentage of subject honorifics and object honorifics, respectively, in each group was computed for inferential statistics. In other words, (a) –kkeyse and (b) -si- were combined into subject honorifics, whereas (c) –kkey and (d) -(e)tulida were collapsed into object honorifics for further inferential statistics. Thus, the rest of the results were analyzed accordingly: (a) subject honorifics and (b) object 81 honorifics. Three inferential statistical methods were used in order to answer the research questions: independent-samples t-tests, one-way ANOVAs, and mixed design repeated-measures ANOVAs. Research question 1 asked whether interactional feedback promotes the development of Korean subject and object referent honorifics. The four experimental groups were collapsed into one experimental group (n = 63) to compare against the control group (n = 15). The within-group variable was the timing of tests (i.e., pretest, immediate posttest, and delayed posttest), and the between-group variable was the efficacy of feedback (i.e., experimental and control groups). As a first step, one-way ANOVAs were performed on the pretest scores of subject and object honorifics, respectively, to determine whether there were any significant differences in learners’ performance prior to treatment. With respect to the overall effects of interactional feedback throughout the pretest-immediate posttest-delayed posttest period, mixed design repeatedmeasures ANOVAs were performed on the group means of subject and object honorifics, respectively. Once the mixed design repeated-measures ANOVAs detected any significant group differences, the one-way ANOVAs were carried out on each of the posttests for subject and object honorifics, respectively. One-way ANOVAs instead of independent-samples t-tests were used for the following reason. One of the assumptions for the independent-samples t-tests involves equal sample sizes, which was not the case for this research question. However, in comparing two different groups of dissimilar numbers of participants against each other, Field (2005) indicated that one-way ANOVA provides reliable results even in unbalanced designs with unequal sample sizes. Research question 2 asked how learners’ HL background affects the development of Korean subject and object referent honorifics. All learners from the HL/Implicit and the 82 HL/Explicit groups were combined to form the HL learner group (n = 31), whereas the NHL/Implicit and the NHL/Explicit groups were collapsed into the NHL learner group (n = 32). The learners in the control group were excluded for analysis. The within-group variable was the timing of tests (i.e., pretest, immediate posttest, and delayed posttest), and the between-group variable was learners’ HL background (i.e., HL and NHL learners). The statistical analysis procedure for this research question was identical to the first research question except that independent-samples t-tests instead of one-way ANOVAs were used. The independent-samples t-tests were performed because nearly even sample sizes were collected in the two groups compared against each other. Research question 3 asked how explicitness of feedback affects the development of Korean subject and object referent honorifics. The implicit feedback group (n = 32) was established with a combination of the HL/Implicit and the NHL/Implicit groups. Likewise, the explicit feedback group (n = 31) was formed by collapsing the HL/Explicit and the NHL/Explicit groups. Again, the learners in the control group were not included for analysis. The within-group variable was the timing of tests (i.e., pretest, immediate posttest, and delayed posttest), and the between-group variable was feedback types (i.e., implicit and explicit feedback). The same statistical analysis procedure used for the first research question was repeated except that independent-samples t-tests instead of one-way ANOVAs were used. The independent-samples t-tests were administered because almost equal sample sizes were obtained in the two groups compared against each other. Research question 4 asked the impact of explicitness of feedback on the Korean subject and object referent honorific development of learners with different HL background. The HL/Implicit group (n = 16), the HL/Explicit group (n = 15), the NHL/Implicit group (n = 16), the 83 NHL/Explicit group (n = 16) and the control group (n = 15) were compared against each other. The within-group variable was the timing of tests (i.e., pretest, immediate posttest, and delayed posttest), and the between-group variables were both learners’ HL background (i.e., HL and NHL learners) and feedback types (i.e., implicit and explicit feedback). Identical statistical analysis procedure used for the first research question was repeated except that the one-way ANOVAs with post hoc tests were performed on each of the posttests for subject and object honorifics, respectively. Games-Howell was selected as post hoc tests. Field (2005) suggested this procedure if there is any uncertainty about homogeneity of variance. Several assumptions for performing independent-samples t-tests were investigated. The assumptions that scores are independent and that data are measured at an interval level were satisfied. Also, sample sizes in each group were nearly equal. Kolmogorove-Smirnov tests for the groups on the pretest and posttests indicated that some data were not normally distributed (p < .05). However, Field (2005) indicated that for equal sample sizes, violating this assumption has only a small impact on the differences between the assumed and true Type I error rate. Several assumptions of ANOVAs were also checked. The assumption of independence was met because the participants were randomly assigned to each group, and each participant did not influence each other. The assumption of independent measurement was also satisfied because the dependent variable was measured at an interval scale, and the distance between the points of the scales was equal. The Kolmogorove-Smirnov tests for the groups on the pretest and posttests revealed that some of the data showed departure from normality and violation of homogeneity of variance by Levene’s tests (p < .05). However, scholars insisted that when group sizes are equal, ANOVA is fairly robust with respect to violations of the assumptions of normality and homogeneity of variance (Field, 2005; Lindman, 1974). Since sample sizes in each group were 84 nearly equal across the groups in this study, ANOVA was deemed acceptable. Welch’s F statistic and Games-Howell as post hoc tests were used when the assumption of homogeneity of variance was violated. For repeated-measures design, when Mauchly’s tests indicated that the assumption of sphericity was not satisfied, Greenhouse-Geisser corrected F-statistic was reported. Concerning effect sizes, this study employed Cohen’s d for the independent-samples t2 tests as well as partial eta squared or ηp for the one-way ANOVAs and the mixed design repeated-measures ANOVAs. For the d, an effect size of .2 is small, .5 is medium, and .8 2 indicates a large effect. For the ηp , .01 constitutes a small effect, .06 a medium effect, and .14 a large effect (Cohen, 1988). An alpha level of .05 was set, and SPSS 15.0 was used to perform the analyses. 85 CHAPTER 4 RESULTS Chapter 3 included detailed information on research methodology in this study, such as participant information, operationalizations, linguistic targets, testing and treatment tasks and materials, research design, procedure, and administration, coding and scoring schemes, and data analyses. This chapter presents the results to answer the research questions raised in Chapter 2. The results are reported according to each of the research questions. Analyses for subject honorifics and object honorifics were conducted separately. The results of subject honorifics are presented first, and those of object honorifics follow. Within each honorific area, descriptive statistics are reported with visual representations, and then inferential statistics are discussed. Role of Interactional Feedback in Pragmatic Development The first research question concerned whether interactional feedback facilitates Korean referent honorific development. The hypothesis predicted that interactional feedback would promote the development of Korean referent honorifics. Subject Honorifics Table 6 below presents means and standard deviations for the pretest-immediate posttest (i.e., posttest 1)-delayed posttest (i.e., posttest 2) for the experimental and the control groups. Descriptive statistics are based on percentage of the use of the target honorific features in learner production across the tests. Recall that only total percentage of subject honorifics, not percentage of each sub-honorific feature, in each group was computed for inferential statistics. Table 6. Descriptive Statistics for Experimental and Control Groups in Subject Honorifics Pretest Group Feature M Posttest 1 SD M 86 SD Posttest 2 M SD Table 6 (cont’d) Experimental -kkeyse 5.75% 19.81 64.55% 39.64 59.58% 42.19 (n = 63) -si- 8.50% 18.69 66.51% 40.04 55.06% 41.15 Total 7.18% 14.65 65.42% 37.32 57.29% 38.59 Control -kkeyse 1.91% 7.38 .00% .00 .00% .00 (n = 15) -si- 10.60% 14.07 12.65% 19.64 18.01% 30.28 Total 6.27% 8.80 6.43% 9.90 9.07% 15.15 The performance of the experimental and the control groups from the pretest to the posttests is illustrated in Figure 6. Overall, the graphs clearly indicate that learners in the experimental group made a remarkable improvement despite a slight decrease between the posttests. In contrast, the control group hardly showed any evidence of development. Figure 6. Visual Representations for Experimental and Control Groups in Subject Honorifics 100% 90% Mean percentage 80% 70% 60% Experimental Control 50% 40% 30% 20% 10% 0% Pretest Posttest 1 87 Posttest 2 Time To validate these visual representations, the one-way ANOVA was conducted and found no statistical difference between the experimental and the control groups on the pretest, F(1, 76) 2 = .05, p = .82, ηp = .00. Thus, both groups proved to perform equally on the pretest. Table 7 displays the results of the mixed design repeated-measures ANOVA: a main effect for group, 2 2 F(1, 76) = 27.34, p = .00, ηp = .27, a main effect for time, F(1.58, 120.04) = 28.83, p = .00, ηp 2 = .28, and an interaction for group × time, F(1.58, 120.04) = 26.47, p = .00, ηp = .26, were statistically significant. The main effect for group represents the differential effects of the groups on the learners’ improvement of subject honorifics. The main effect for time demonstrates the learners’ differential performance throughout the pretest-immediate posttest-delayed posttest. The significant interaction effect between the groups and time suggests that the groups performed differently from each other over time. Table 7. ANOVA Table for Experimental and Control Groups in Subject Honorifics across Tests Source df 2 p F n ηp Between subjects Group 1.00 27.34 .00 .27 Within subjects Time 1.58 28.83 .00 .28 Group × Time 1.58 26.47 .00 .26 In order to determine whether the difference between the two groups was statistically significant on each posttest, a series of the one-way ANOVAs were performed. The differences between the experimental and the control groups on the immediate posttest, F(1, 76) = 36.52, p 2 2 = .00, ηp = .33, and on the delayed posttest, F(1, 76) = 22.41, p = .00, ηp = .23, were 88 statistically significant. However, the relative values of the effect sizes indicate that the difference between the experimental and the control groups on the immediate posttest was larger than on the delayed posttest. These results confirmed the previous inspection that the experimental group was superior to the control group on both the immediate posttest and the delayed posttest in the development of subject honorifics, and that the superiority of the experimental group over the control group was greater on the immediate posttest than on the delayed posttest. Object Honorifics The following Table 8 shows descriptive statistics for the experimental and the control groups. The descriptive statistics were computed on the basis of percentage of the use of the target honorific features obtained across the pretest-immediate posttest-delayed posttest. Again, note that only total percentage of object honorifics in each group was used for further analysis. Table 8. Descriptive Statistics for Experimental and Control Groups in Object Honorifics Pretest Posttest 1 Posttest 2 Group Feature M SD M SD M SD Experimental -kkey 8.49% 23.39 74.60% 39.08 60.86% 45.45 (n = 63) (-e)tulida 5.90% 15.74 75.36% 30.30 61.31% 36.86 Total 7.20% 14.99 75.00% 30.97 61.11% 36.73 Control -kkey 2.50% 9.68 1.67% 6.46 .83% 3.23 (n = 15) (-e)tulida 7.13% 16.62 7.50% 16.23 6.67% 14.84 Total 4.90% 10.15 4.59% 10.16 3.75% 8.12 89 Figure 7 illustrates graphic patterns for the experimental and the control groups. It is evident that the experimental group improved dramatically from the pretest to the posttests although there was a slight decline from the immediate posttest to the delayed posttest. On the contrary, the control group showed no sign of improvement. Figure 7. Visual Representations for Experimental and Control Groups in Object Honorifics 100% 90% Mean percentage 80% 70% 60% 50% Experimental Control 40% 30% 20% 10% 0% Pretest Posttest 1 Posttest 2 Time First, the one-way ANOVA revealed that the difference in mean percentage between the experimental and the control groups on the pretest was not statistically significant, F(1, 76) = .32, 2 p = .58, ηp = .00, and that the two groups did not differ from each other on the pretest. The mixed design repeated-measures ANOVA shown in Table 9 found that there were a significant 2 main effect for group, F(1, 76) = 51.90, p = .00, ηp = .41, a significant main effect for time, 2 F(1.76, 133.44) = 38.15, p = .00, ηp = .33, as well as a significant interaction for group × time, 90 2 F(1.76, 133.44) = 39.59, p = .00, ηp = .34. The main effect for group indicates the differential effects of the groups on the learners’ improvement of object honorifics. The main effect for time illustrates the learners’ differential performance throughout the pretest-immediate posttestdelayed posttest. Additionally, the significant interaction effect between the groups and time showed that the groups performed differently from each other over time. Table 9. ANOVA Table for Experimental and Control Groups in Object Honorifics across Tests df Source F p ηn p 2 Between subjects Group 1 51.90 .00 .41 Within subjects Time 1.76 38.15 .00 .33 Group × Time 1.76 39.59 .00 .34 The one-way ANOVAs were performed to determine whether the difference between the two groups was statistically significant on each posttest. The difference between the experimental and the control groups on the immediate posttest was statistically significant, F(1, 2 76) = 74.96, p = .00, ηp = .50. The difference between the two groups on the delayed posttest, 2 F(1, 76) = 35.82, p = .00, ηp = .32, was also statistically significant. However, the relative magnitudes of the effect sizes show that the difference on the immediate posttest was larger than on the delayed posttest. These results corroborated that the experimental group outperformed the control group on both posttests in the learning of object honorifics, and that the supremacy of the experimental group over the control group was stronger on the immediate posttest than on the delayed posttest. 91 Heritage vs. Non-heritage Language Learners The second research question involved how learners’ HL background influences the learning of Korean referent honorifics. The hypothesis predicted that the NHL learners would learn Korean referent honorifics better than the HL learners. Subject Honorifics Table 10 displays descriptive statistics for the groups of HL and NHL learners in the development of subject honorifics. The descriptive statistics include means and standard deviations based on percentage of the use of the target honorific features. Overall, larger standard deviations for the HL learner group than for the NHL learner group indicate that the HL learners are generally more heterogeneous in their performance than the NHL counterparts. Table 10. Descriptive Statistics for HL and NHL Learner Groups in Subject Honorifics Pretest Posttest 1 Group Feature M SD HL -kkeyse .40% (n = 31) -si- Posttest 2 M SD M SD 2.25 48.87% 42.82 37.99% 41.73 8.42% 20.51 40.73% 41.46 29.51% 38.11 Total 4.46% 10.28 44.56% 39.21 33.68% 36.04 NHL -kkeyse 10.94% 26.89 79.73% 29.75 80.50% 30.95 (n = 32) -si- 8.58% 17.08 91.49% 15.28 79.82% 26.37 Total 9.81% 17.67 85.63% 21.01 80.17% 24.98 The visual patterns of the descriptive statistics are shown in Figure 8. The NHL learners outperformed the HL learners at the time of posttesting. Whereas both groups improved the use 92 of subject honorifics from the pretest to the immediate posttest, there was a slight drop from the immediate posttest to the delayed posttest, and the greater decrease was observed for the HL learners than the NHL learners. Figure 8. Visual Representations for HL and NHL Learner Groups in Subject Honorifics 100% 90% Mean percentage 80% 70% 60% HL NHL 50% 40% 30% 20% 10% 0% Pretest Posttest 1 Posttest 2 Time To better understand the visual representations, inferential statistical analyses were carried out. First, the results of the independent-samples t-test indicated that the HL and the NHL learner groups were equivalent on the pretest, t(50.12) = 1.47, p = .15, d = .37. As shown in Table 11, the results of the mixed design repeated-measures ANOVA showed that a main effect 2 for group, F(1, 61) = 33.82, p = .00, ηp = .36, a main effect for time, F(1.74, 106.28) = 153.58, 2 p = .00, ηp = .72, and an interaction for group × time, F(1.74, 106.28) = 19.49, p = .00, ηp 2 = .24, were statistically significant. The main effect for group indicates the differential effects of the groups on the learners’ development of subject honorifics. Likewise, the main effect for time 93 illustrates the learners’ differential performance throughout the pretest-immediate posttestdelayed posttest. More importantly, the significant interaction effect between the groups and time indicates that the groups performed differently from each other across tests. Table 11. ANOVA Table for HL and NHL Learner Groups in Subject Honorifics across Tests df Source F p ηx p 2 Between subjects Group 1 33.82 .00 .36 Within subjects Time 1.74 153.58 .00 .72 Group × Time 1.74 19.49 .00 .24 In order to determine whether the difference between the two groups was statistically significant on each posttest, a series of the independent-samples t-tests were performed. The difference between the HL and the NHL learner groups on the immediate posttest was statistically significant, t(45.60) = 5.16, p = .00, d = 1.30. Also, there was a statistically significant difference between the two groups on the delayed posttest, t(53.25) = 5.93, p = .00, d = 1.49. Yet, the comparison of the effect sizes indicates that the difference on the delayed posttest was larger than on the immediate posttest. These results confirmed that the NHL learners outperformed the HL learners on both of the posttests in learning subject honorifics. These results also supported the prior speculation that the NHL learners excelled compared to the HL learners more on the delayed posttest than on the immediate posttest in learning subject honorifics. Object Honorifics Table 12 shown below provides detailed information on descriptive statistics, including 94 means and standard deviations, for the groups of HL and NHL learners in the development of object honorifics. Along with subject honorifics, the HL learner group consistently showed larger standard deviations than the NHL learner group, which suggests that the former shows more heterogeneous performance than the latter. Table 12. Descriptive Statistics for HL and NHL Learner Groups in Object Honorifics Pretest M Posttest 1 Group Feature SD HL -kkey 7.66% 24.30 (n = 31) (-e)tulida 9.98% Total NHL (n = 32) M Posttest 2 SD M SD 58.47% 44.53 48.33% 47.25 19.84 62.20% 32.97 51.61% 34.54 8.82% 17.68 60.31% 32.44 50.00% 36.03 -kkey 9.29% 22.84 90.23% 24.95 72.99% 40.78 (-e)tulida 1.95% 9.04 88.11% 21.04 70.70% 37.11 Total 5.63% 11.90 89.23% 21.79 71.87% 34.62 Figure 9 provides the visual representations for the HL and the NHL learner groups in the development of object honorifics. Similar to subject honorifics, the NHL learners did better than the HL learners on both posttests. Although both groups demonstrated a substantial increase from the pretest to the immediate posttest in the use of object honorifics, the mean percentage underwent a decrease from the immediate posttest to the delayed posttest, and the decline was greater for the NHL learners than the HL learners. 95 Figure 9. Visual Representations for HL and NHL Learner Groups in Object Honorifics 100% 90% Mean percentage 80% 70% 60% HL NHL 50% 40% 30% 20% 10% 0% Pretest Posttest 1 Posttest 2 Time To consolidate this impression from the visual images, the independent-samples t-test was first conducted and found no statistically significant difference between the HL and the NHL learner groups on the pretest, t(61) = -.84, p = .40, d = .21. As presented in Table 13, the results of the mixed design repeated-measures ANOVA revealed that there were a significant main 2 effect for group, F(1, 61) = 8.57, p = .01, ηp = .12, a significant main effect for time, F(1.78, 2 108.82) = 189.77, p = .00, ηp = .76, and a significant interaction for group × time, F(1.78, 2 108.82) = 10.61, p = .00, ηp = .15. The main effect for group suggests the differential effects of the groups on the learners’ development of object honorifics. The main effect for time illustrates the learners’ differential development throughout the tests. The significant interaction effect between the groups and time indicates that the groups performed differently from each other over time. 96 Table 13. ANOVA Table for HL and NHL Learner Groups in Object Honorifics across Tests Source p x ηp 8.57 .01 .12 df F 2 Between subjects Group 1 Within subjects Time 1.78 189.77 .00 .76 Group × Time 1.78 10.61 .00 .15 In order to see whether the difference between the two groups was statistically significant on each posttest, the independent-samples t-tests followed. The difference between the HL and the NHL learner groups on the immediate posttest was statistically significant, t(52.30) = 4.14, p = .00, d = 1.04. The difference between the two groups on the delayed posttest, t(61) = 2.46, p = .02, d = .62, was also found to be significant. Nonetheless, the larger effect size on the immediate posttest than on the delayed posttest indicates that the difference between the two groups was greater on the immediate posttest than on the delayed posttest. These results confirmed that the NHL learners outdid the HL learners on both posttests in the learning of object honorifics, and that the superiority of the NHL learner group over the HL learner group was stronger on the immediate posttest than on the delayed posttest. Implicit vs. Explicit Feedback The third research question asked the relative effects of implicit and explicit feedback on learning Korean referent honorifics. The hypothesis predicted that explicit feedback would work more effectively than implicit feedback in learning Korean referent honorifics. Subject Honorifics Descriptive statistics related to learners’ development of subject honorifics are displayed in Table 14. The descriptive statistics are generated for the implicit and the explicit feedback 97 groups and include means and standard deviations. Table 14. Descriptive Statistics for Implicit and Explicit Feedback Groups in Subject Honorifics Pretest Posttest 1 Group Feature M SD M Implicit -kkeyse 2.73% 13.37 46.14% (n = 32) -si- 5.14% 11.98 Total 4.00% Explicit -kkeyse (n = 31) SD Posttest 2 M SD 41.86 46.86% 43.75 49.87% 44.68 39.97% 39.87 11.64 47.79% 39.61 43.43% 38.68 8.87% 24.62 83.55% 26.51 72.72% 36.75 -si- 11.97% 23.43 83.69% 25.50 70.64% 36.91 Total 10.45% 16.78 83.61% 24.16 71.60% 33.41 The graphic patterns for the implicit and the explicit feedback groups are plotted in Figure 10. It shows that the explicit feedback group surpassed the implicit feedback group in learning subject honorifics although the considerable gains made from the pretest to the immediate posttest fell over time. The greater decline from the immediate posttest to the delayed posttest was found for the explicit feedback group than for the implicit feedback group. 98 Figure 10. Visual Representations for Implicit and Explicit Feedback Groups in Subject Honorifics 100% 90% Mean percentage 80% 70% 60% Implicit Explicit 50% 40% 30% 20% 10% 0% Pretest Posttest 1 Posttest 2 Time In order to confirm these visual inspections, the independent-samples t-test was first conducted and found no statistically significant difference between the implicit and the explicit feedback groups on the pretest, t(53.29) = -1.77, p = .08, d = .45. This indicates that both groups performed similarly prior to treatments. Table 15 below reports the results of the mixed design 2 repeated-measures ANOVA: a significant main effect for group, F(1, 61) = 15.73, p = .00, ηp 2 = .21, a significant main effect for time, F(1.62, 98.50) = 133.61, p = .00, ηp = .69, and a 2 significant interaction for group × time, F(1.62, 98.50) = 7.74, p = .00, ηp = .11. The main effect for group represents the differential effects of the groups on the subject honorific development. The main effect for time illustrates the learners’ differential performance throughout the pretest-immediate posttest-delayed posttest. Also, the significant interaction effect between the groups and time indicates that the groups performed differently from each 99 other as time progressed. Table 15. ANOVA Table for Implicit and Explicit Feedback Groups in Subject Honorifics across Tests Source df F p x ηp 2 Between subjects Group 1 15.73 .00 .21 Within subjects Time 1.62 133.61 .00 .69 Group × Time 1.62 7.74 .00 .11 To discover whether the difference between the two groups was statistically significant on each posttest, the independent-samples t-tests were performed. The differences between the implicit and the explicit feedback groups on the immediate posttest, t(51.53) = -4.35, p = .00, d = 1.10, and on the delayed posttest, t(61) = -3.09, p = .00, d = .78, were statistically significant. Based on the relative magnitudes of the effect sizes, however, the difference between the two groups was greater on the immediate posttest than on the delayed posttest. Overall, the results proved the earlier speculation that the explicit feedback group performed better than the implicit feedback group on both posttests in learning subject honorifics, and that the supremacy of the explicit feedback group over the implicit feedback group was stronger on the immediate posttest than on the delayed posttest. Object Honorifics Table 16 below illustrates descriptive statistics for the implicit and the explicit feedback groups on the development of object honorifics across the pretest-immediate posttest-delayed posttest. 100 Table 16. Descriptive Statistics for Implicit and Explicit Feedback Groups in Object Honorifics Pretest Group Feature Implicit -kkey (n = 32) M Posttest 1 Posttest 2 SD M SD M SD 3.13% 11.88 58.98% 46.25 39.01% 45.15 (-e)tulida 9.53% 20.81 64.68% 33.03 46.62% 35.88 Total 6.37% 14.17 61.92% 35.40 42.89% 36.11 Explicit -kkey 14.02% 30.37 90.73% 20.40 83.41% 33.60 (n = 31) (-e)tulida 2.15% 5.98 86.38% 22.86 76.48% 31.76 Total 8.05% 15.98 88.50% 17.78 79.91% 26.86 The visual graphs for object honorifics in Figure 11 provide an impression that explicit feedback works more effectively than implicit feedback. For both feedback groups, the initial gains from the pretest to the immediate posttest did not sustain entirely on the delayed posttest, and there was a larger decline between the two posttests for the implicit feedback group than for the explicit feedback group. 101 Figure 11. Visual Representations for Implicit and Explicit Feedback Groups in Object Honorifics 100% 90% Mean percentage 80% 70% 60% Implicit Explicit 50% 40% 30% 20% 10% 0% Pretest Posttest 1 Posttest 2 Time The independent-samples t-test was performed to establish whether the difference between the groups was statistically significant on the pretest. This revealed no statistically significant difference between the implicit and the explicit feedback groups, t(61) = -.44, p = .66, d = .11. As the results of the mixed design repeated-measures ANOVA indicate in Table 17, a 2 main effect for group, F(1, 61) = 18.41, p = .00, ηp = .23, a main effect for time, F(2, 122) = 2 2 197.97, p = .00, ηp = .76, and an interaction for group × time, F(2, 122) = 12.62, p = .00, ηp = .17, were statistically significant. The main effect for group indicates the differential effects of the groups on the learners’ improvement of object honorifics. The main effect for time illustrates the learners’ differential performance throughout the tests. The significant interaction effect between the groups and time shows that the groups performed differently from each other over time. 102 Table 17. ANOVA Table for Implicit and Explicit Feedback Groups in Object Honorifics across Tests Source df F p xp2 η Between subjects Group 1 18.41 .00 .23 Within subjects Time 2 197.97 .00 .76 Group × Time 2 12.62 .00 .17 To find out whether the difference between the two groups was statistically significant on each posttest, the independent-samples t-tests were performed. The difference between the implicit and the explicit feedback groups on the immediate posttest was statistically significant, t(46.03) = -3.78, p = .00, d = .95. Likewise, a statistically significant difference between the two groups was found on the delayed posttest, t(57.24) = -4.63, p = .00, d = 1.17. The comparison of the effect sizes indicates that the difference between the two groups was greater on the delayed posttest than on the immediate posttest. These results confirmed the visual representations that the explicit feedback group outdid the implicit feedback group on both of the posttests in learning object honorifics, and that the outperformance of the explicit feedback group was stronger on the delayed posttest than on the immediate posttest. Heritage vs. Non-heritage Language Learners in relation to Implicit vs. Explicit Feedback The fourth research question concerned the relative efficacy of implicit and explicit feedback on the development of Korean referent honorifics in the context of HL and NHL learning. The hypothesis predicted that the NHL learners would benefit more from explicit feedback than implicit feedback in learning Korean referent honorifics. However, the hypothesis was not projected as to which type of feedback would work better for the HL learners. Subject Honorifics 103 Table 18 illustrates the average frequency of errors occurred and feedback provided per learner in each experimental group during the two treatment sessions in subject honorifics. The amount of feedback provided to each group varied according to the number of learner errors and thusly differed among the groups. The learners in the HL/Implicit group showed the largest number of errors and feedback, whereas those in the NHL/Explicit group displayed the smallest. Other groups were placed in the middle of these two groups. This suggests that the learners in the NHL/Explicit group tended to learn more easily and rapidly than those in the HL/Implicit group during the treatments. As the treatments progressed, the number of errors and feedback in the NHL/Explicit group decreased at a faster rate than those of the HL/Implicit group. Therefore, overall, the learners in the NHL/Explicit group made fewer errors and therefore received fewer number of feedback from the NS interlocutor than other groups during the treatment sessions. This indicates that if the learners in the NHL/Explicit group make a greater development through the treatments, it is not because they receive more feedback. Rather, it is likely that the nature of the NHL learners and explicit feedback facilitates a faster rate of learning, which results in fewer number of feedback than those in other groups. 104 Table 18. Average Frequency of Error Occurrences and Feedback Moves per Learner in Each Experimental Group in Subject Honorifics Experimental -kkeyse + -si -kkeyse -si Total group Error Feedback Error Feedback Error Feedback Error Feedback HL/Implicit (n = 16) 25.2 23.2 1.4 1.2 6.8 5.8 33.4 30.2 HL/Explicit (n = 15) 7.8 7.6 2.6 2.6 6.6 6.6 17.0 16.8 NHL/Implicit (n = 16) 6.6 6.6 5.2 4.6 1.8 1.6 13.6 12.8 NHL/Explicit (n = 16) 2.6 2.6 1.8 1.8 3.8 3.6 8.2 8.0 105 Table 19 below presents descriptive statistics for the HL/Implicit, the HL/Explicit, the NHL/Implicit, the NHL/Explicit, and the control groups on their subject honorific development. Descriptive statistics were calculated based on mean percentage. Table 19. Descriptive Statistics for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Subject Honorifics Pretest Group Feature M HL/Implicit -kkeyse .00% (n = 16) -si- Posttest 1 SD Posttest 2 M SD M SD .00 25.00% 36.51 23.44% 35.32 2.19% 4.73 10.50% 23.12 6.16% 13.65 Total 1.12% 2.41 17.30% 24.76 14.82% 21.91 HL/Explicit -kkeyse .83% 3.23 74.33% 33.97 53.52% 43.53 (n = 15) -si- 15.06% 28.03 72.97% 30.88 54.41% 40.31 Total 8.02% 13.92 73.63% 29.71 53.80% 37.81 NHL/Implicit -kkeyse 5.47% 18.80 67.28% 36.53 70.29% 39.22 (n = 16) -si- 8.09% 15.99 89.24% 16.86 73.78% 25.71 Total 6.89% 16.01 78.28% 25.43 72.05% 29.40 NHL/Explicit -kkeyse 16.41% 32.82 92.19% 12.81 90.71% 14.77 (n = 16) -si- 9.08% 18.61 93.75% 13.69 85.86% 26.42 Total 12.73% 19.25 92.98% 12.25 88.28% 16.85 Control -kkeyse 1.91% 7.38 .00% .00 .00% .00 (n = 15) -si- 10.60% 14.07 12.65% 19.64 18.01% 30.28 Total 6.27% 8.80 6.43% 9.90 9.07% 15.15 106 Figure 12 graphically illustrates the mean percentage of the use of the target subject honorific features for all the participating groups. While the experimental groups made an improvement, the control group performed similarly across the tests. Upon closer inspection, the HL/Explicit, the NHL/Implicit, and the NHL/Explicit groups reported much higher rates of improvement than the HL/Implicit and the control groups. Although the mean percentage of all the five groups appeared to increase over time, that of the four experimental groups were dropped from the immediate posttest to the delayed posttest. In particular, a larger decrease was observed for the HL/Explicit group than for other experimental groups. Also, explicit feedback seemed visibly far more effective than implicit feedback for the HL learners. The efficacy of explicit feedback appeared to be stronger than implicit feedback for the NHL learners. However, this contrast was not as strong as that of the HL learner group. Clearly, the NHL/Explicit group made the greatest development, whereas the HL/Implicit group made the least gains from the pretest to the posttests among the experimental groups. The NHL/Implicit and the HL/Explicit groups showed similar developmental rates on the immediate posttest, although the former outperformed the latter on the delayed posttest. To validate the relative developmental rates among the five groups on the posttests, inferential statistical analyses were conducted. 107 Figure 12. Visual Representations for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Subject Honorifics 100% 90% Mean percentage 80% 70% HL/Implicit HL/Explicit NHL/Implicit NHL/Explicit Control 60% 50% 40% 30% 20% 10% 0% Pretest Posttest 1 Posttest 2 Time First, the one-way ANOVA revealed that the difference among the HL/Implicit, the HL/Explicit, the NHL/Implicit, the NHL/Explicit, and the control groups on the pretest was not 2 statistically significant, F(4, 73) = 1.51, p = .21, ηp = .08, which indicates that all the five groups performed similarly on the pretest. Table 20 illustrates the findings of the mixed design 2 repeated-measures ANOVA. A main effect for group, F(4, 73) = 40.58, p = .00, ηp = .69, a 2 main effect for time, F(1.84, 134.16) = 181.24, p = .00, ηp = .71, and an interaction for group × 2 time, F(7.35, 134.16) = 22.15, p = .00, ηp = .55, were statistically significant. The main effect for group indicates the differential effects of the groups on the subject honorific development. The main effect for time accounts for the learners’ differential development throughout the pretest and the posttests. Of note is the significant interaction effect between the groups and time, which indicates that the groups performed differently from each other across tests. 108 Table 20. ANOVA Table for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Subject Honorifics across Tests Source df F p ηx p 2 Between subjects Group 4.00 40.58 .00 .69 Within subjects Time 1.84 181.24 .00 .71 Group × Time 7.35 22.15 .00 .55 In order to find out whether the difference among the five groups is statistically significant on each posttest, the one-way ANOVAs were performed. The results indicated that the learners in the five groups differed significantly from each other on the immediate posttest, 2 F(4, 73) = 49.72, p = .00, ηp = .731, as well as the delayed posttest, F(4, 73) = 29.13, p = .00, 2 ηp = .615. Post hoc multiple comparisons were computed on each posttest to detect where the significant differences lay among the groups. As shown in Table 21, it was revealed that for the HL learners, explicit feedback worked significantly better than implicit feedback on both the immediate and the delayed posttests. However, for the NHL learners, there was no significant difference between implicit and explicit feedback regardless of timing of tests. All the experimental groups, except the HL/Implicit group, performed significantly better than the control group on both posttests. Among these, the NHL/Explicit group showed the best performance as the largest effect size indicated in comparing each of the experimental groups against the control group. In contrast, there was no statistically significant difference between the HL/Implicit and the control groups in the development of subject honorifics. Also, the smallest effect size was reported for the HL/Implicit group among the comparisons of each experimental group against the control group. This demonstrated that the HL/Implicit group showed the worst 109 performance among all the experimental groups. In addition, the results showed that there was no statistically significant difference between the NHL/Implicit and the HL/Explicit groups at the time of both posttesting. This implies that these two groups did not differ from each other in terms of their performance. Of interesting note is that while the NHL/Explicit and the HL/Explicit groups did not significantly differ in light of their performance on the immediate posttest, the NHL/Explicit group significantly outperformed the HL/Explicit group on the delayed posttest. This indicates that the HL/Explicit group showed a larger decrease from the immediate posttest to the delayed posttest. Table 21. Post Hoc Tests in Subject Honorifics Time Posttest 1 Group d = NHL/Explicit .74 NHL/Implicit > HL/Implicit* 2.43 NHL/Implicit = HL/Explicit .17 NHL/Implicit > Control* 3.68 NHL/Explicit > HL/Implicit* 3.87 NHL/Explicit = HL/Explicit .86 NHL/Explicit > Control* 7.74 HL/Implicit < HL/Explicit* 2.07 HL/Implicit = ontrol .57 HL/Explicit Posttest 2 NHL/Implicit > Control* 3.03 NHL/Implicit = NHL/Explicit .68 NHL/Implicit > HL/Implicit* 2.21 110 Table 21 (cont’d) NHL/Implicit = HL/Explicit .54 NHL/Implicit > Control* 2.67 NHL/Explicit > HL/Implicit* 3.76 NHL/Explicit > HL/Explicit* 1.19 NHL/Explicit > Control* 4.93 HL/Implicit < HL/Explicit* 1.27 HL/Implicit = Control .30 HL/Explicit > Control* 1.55 *p<.05 Object Honorifics Table 22 presents the mean frequency of error occurrences and feedback moves per learner in each experimental group during the two treatment sessions in object honorifics. The amount of feedback supplied to each group was contingent on that of errors and varied among the groups. While the learners in the HL/Implicit group showed the highest frequency of errors and feedback, those in the NHL/Explicit group demonstrated the lowest. Other groups were between these two groups. This suggests that if the learners in the NHL/Explicit group show a superior development in this study, it is not because they receive more feedback. Instead, it may be that the nature of the NHL learners and explicit feedback plays a role in their successful development. 111 Table 22. Average Frequency of Error Occurrences and Feedback Moves per Learner in Each Experimental Group in Object Honorifics Experimental -kkey + -(e)tulida -kkey Error -(e)tulida group Error Feedback Feedback HL/Implicit (n = 16) 18.4 17.4 5.2 5.2 1.0 HL/Explicit (n = 15) 4.0 3.8 2.4 2.2 NHL/Implicit (n = 16) 8.6 8.4 3.2 Error Feedback .8 24.6 23.4 6.6 6.2 13.0 12.2 3.0 NHL/Explicit (n = 16) 1.0 1.0 1.4 3.6 3.6 15.4 15.0 1.4 112 Error Total Feedback 7.6 7.4 10.0 9.8 Table 23 shows descriptive statistics, including means and standard deviations, for the HL/Implicit, the HL/Explicit, the NHL/Implicit, the NHL/Explicit, and the control groups on the object honorific development. Descriptive statistics were calculated based on mean percentage. Table 23. Descriptive Statistics for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Object Honorifics Pretest Group Feature HL/Implicit -kkey (n = 16) M Posttest 1 Posttest 2 SD M SD M SD 3.91% 15.63 35.94% 46.52 23.44% 40.02 (-e)tulida 15.94% 25.52 51.58% 34.91 40.89% 28.19 Total 9.99% 18.44 43.83% 33.22 32.31% 30.12 HL/Explicit -kkey 11.67% 31.15 82.50% 27.06 74.88% 40.05 (n = 15) (-e)tulida 3.61% 7.86 73.52% 27.52 63.05% 37.86 Total 7.56% 17.39 77.89% 20.74 68.88% 32.72 NHL/Implicit -kkey 2.34% 6.80 82.03% 33.53 54.58% 45.76 (n = 16) (-e)tulida 3.13% 12.50 77.79% 25.87 52.34% 42.38 Total 2.74% 6.84 80.01% 28.06 53.48% 39.35 NHL/Explicit -kkey 16.23% 30.48 98.44% 4.27 91.41% 24.88 (n = 16) (-e)tulida .78% 3.13 98.44% 4.27 89.06% 18.19 Total 8.51% 15.10 98.44% 3.60 90.26% 14.43 Control -kkey 2.50% 9.68 1.67% 6.46 .83% 3.23 (n = 15) (-e)tulida 7.13% 16.62 7.50% 16.23 6.67% 14.84 Total 4.90% 10.15 4.59% 10.16 3.75% 8.12 113 Figure 13 graphically depicts the mean percentage of the use of the target object honorific items for all the participating groups. The mean percentage of all the four experimental groups escalated over time, but was reduced from the immediate posttest to the delayed posttest. Particularly, the NHL/Implicit group underwent a larger decline than other groups. Unlike the experimental groups, the mean percentage for the control group appeared to remain similar throughout the tests. It seemed that explicit feedback worked more effectively than implicit feedback for the HL learners. Explicit feedback also appeared to be more effective than implicit feedback for the NHL learners; however, this was more obvious on the delayed posttest than on the immediate posttest. It is evident that the NHL/Explicit group showed a superior performance over all the other groups, whereas the HL/Implicit group demonstrated the least gains among the experimental groups as the treatments progressed. Although both the NHL/Implicit and the HL/Explicit groups performed similarly on the immediate posttest, the HL/Explicit group outperformed the NHL/Implicit group on the delayed posttest. To corroborate these visual inspections, inferential statistical analyses were carried out. 114 Figure 13. Visual Representations for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Object Honorifics 100% 90% Mean percentage 80% 70% HL/Implicit HL/Explicit NHL/Implicit NHL/Explicit Control 60% 50% 40% 30% 20% 10% 0% Pretest Posttest 1 Posttest 2 Time As a first step, the one-way ANOVA found that the difference in mean percentage among the HL/Implicit, the HL/Explicit, the NHL/Implicit, the NHL/Explicit, and the control groups on 2 the pretest was not statistically significant, F(4, 73) = .66, p = .62, ηp = .04, which indicates that all the groups performed equally on the pretest. Moreover, as the results of the mixed design repeated-measures ANOVA show in Table 24, a main effect for group, F(4, 73) = 28.14, p = .00, 2 2 ηp = .61, a main effect for time, F(2, 146) = 229.73, p = .00, ηp = .76, and an interaction for 2 group × time, F(8, 146) = 23.36, p = .00, ηp = .56, were statistically significant. The main effect for group indicates the differential effects of the groups on the learners’ object honorific development. The main effect for time illustrates the learners’ differential performance throughout the pretest-immediate posttest-delayed posttest. Of note is the significant interaction effect between the groups and time, which suggests that the participating groups performed 115 differently from each other over time. Table 24. ANOVA Table for HL/Implicit, HL/Explicit, NHL/Implicit, NHL/Explicit, and Control Groups in Object Honorifics across Tests Source df p F x ηp 2 Between subjects Group 4 28.14 .00 .61 Within subjects Time 2 229.73 .00 .76 Group × Time 8 23.36 .00 .56 In order to find out if there is any significant difference among the five groups on each posttest, the one-way ANOVAs were performed. The findings indicated that the five groups of learners differed significantly from each other on the immediate posttest, F(4, 73) = 43.03, p 2 2 = .00, ηp = .70, and the delayed posttest, F(4, 73) = 22.40, p = .00, ηp = .55. To locate the source of the differences, post hoc multiple comparisons were conducted. As shown in Table 25, the results from the post hoc analysis found that as far as the HL learners are concerned, explicit feedback worked significantly better than implicit feedback in the learning of object honorifics on the immediate posttest and the delayed posttest. For the NHL learners, while there was no statistically significant difference between the two types of feedback on the immediate posttest, explicit feedback worked significantly better than implicit feedback on the delayed posttest. All the experimental groups significantly surpassed the control group. Among these, the NHL/Explicit group performed the best on both posttests, as indicated in the largest effect size among the comparisons of each experimental group against the control group. In contrast, the HL/Implicit group showed the least development across tests, as shown in the smallest effect size 116 in comparing each experimental group against the control group. Irrespective of timing of posttests, there was no statistical difference between the NHL/Implicit and the HL/Explicit groups. Also, the NHL/Implicit group significantly outperformed the HL/Implicit group on the immediate posttest, whereas the NHL/Implicit and the HL/Implicit groups were not significantly different in their development on the delayed posttest. This indicates that the NHL/Implicit group underwent a larger decline in the development between the posttests. Table 25. Post Hoc Tests in Object Honorifics Time Posttest 1 Group d = NHL/Explicit .92 NHL/Implicit > HL/Implicit* 1.18 NHL/Implicit = HL/Explicit .09 NHL/Implicit > Control* 3.53 NHL/Explicit > HL/Implicit* 2.31 NHL/Explicit > HL/Explicit* 1.40 NHL/Explicit > Control* 12.48 HL/Implicit < HL/Explicit* 1.22 HL/Implicit > Control* 1.58 HL/Explicit Posttest 2 NHL/Implicit > Control* 4.49 NHL/Implicit < NHL/Explicit* 1.24 NHL/Implicit = HL/Implicit .60 NHL/Implicit = HL/Explicit .42 NHL/Implicit > Control* 1.72 117 Table 25 (cont’d) NHL/Explicit > HL/Implicit* 2.45 NHL/Explicit = HL/Explicit .86 NHL/Explicit > Control* 7.32 HL/Implicit < HL/Explicit* 1.16 HL/Implicit > Control* 1.28 HL/Explicit > Control* 2.73 *p<.05 118 CHAPTER 5 DISCUSSION While an ample amount of research has been conducted within the interactionist paradigm in the past, most research has examined L2 learners’ morphosyntactic development predominantly in English. Thus far, there has been little evidence of how pragmatic development takes place through interactional feedback, how HL learners develop their HL language as opposed to NHL learners, and how feedback of different levels of explicitness interacts with these variables. Therefore, this dissertation aims to investigate the impact of learners’ HL background and explicitness of interactional feedback on the L2 pragmatic development in Korean. Korean referent honorifics, including subject and object honorifics, were chosen as a pragmatic target. Implicit feedback took the form of recasts, whereas explicit feedback took the form of explicit correction. All the HL learner participants were second-generation KoreanAmericans who had experienced exposure to and use in Korean at home during childhood, whereas all the NHL learners were typical adult L2 learners without prior exposure to Korean at home. A total of 78 intermediate-level English-speaking learners of Korean participated in this study and were randomly assigned to one of the following groups: (a) HL learner/Implicit feedback group; (b) HL learner/Explicit feedback group; (c) NHL learner/Implicit feedback group; (d) NHL learner/Explicit feedback group; and (e) control group. Using a pretestimmediate posttest-delayed posttest design, the participants in each group performed a series of communicative tasks with the Korean NS interlocutor and received feedback on their nontargetlike production during treatments. The following summarizes the research findings in response to each of the research questions and hypotheses in this study. The first research question asked whether interactional feedback facilitates learning Korean referent honorifics. The hypothesis was supported given that the efficacy of interactional 119 feedback was statistically significant in the development of Korean referent honorifics. The experimental group that showed the statistically significant effects of interactional feedback was in a drastic contrast to the control group whose improvement was hardly observed. The second research question asked how learners’ HL background influences the development of Korean referent honorifics. The hypothesis was upheld because the results showed that the NHL learners significantly outperformed the HL learners in learning honorifics. The unexpected pattern was found that the HL learners performed object honorifics better than subject honorifics. The third research question concerned the relative effects of implicit and explicit feedback on learning Korean referent honorifics. The results found that explicit feedback was significantly more effective than implicit feedback in learning both subject and object honorifics; therefore, the projected hypothesis was supported. The fourth research question involved the efficacy of implicit and explicit feedback on the learning of Korean referent honorifics for learners with differing HL background. Although the NHL learners appeared to benefit more from explicit feedback than from implicit feedback, there was no statistically significant difference in the relative effectiveness of the two feedback types in learning Korean honorifics except on the delayed posttest in object honorific development. Thus, overall, the hypothesis was not supported. Although a clear prediction was not formed in respect of the relative effects of the two feedback types for the HL learner population, this study found that explicit feedback significantly worked better than implicit feedback for the HL learners. Furthermore, the NHL learners in the explicit feedback condition showed the best performance, whereas the HL learners in the implicit feedback condition demonstrated the worst other than the group of learners who received no feedback over time. The 120 remaining experimental groups fell between these groups. The following sections discuss each of the above findings with reference to theoretical underpinnings and empirical evidence drawn from previous studies. Role of Interactional Feedback in Pragmatic Development The results showed the positive role of interactional feedback in the development of Korean honorifics. Low pretest scores indicate that most of the participants did not produce the target honorific features at all, presumably as a consequence of their habit of not using honorifics or their low confidence in honorific usage. Some made an attempt to use honorifics, but, in most cases, failed to use them properly. However, learners in the experimental group generally made a statistically significant improvement, although learners in the control group barely showed signs of development. Recall that according to Long’s Interaction Hypothesis (1996), negative feedback may facilitate L2 development, at least in the field of vocabulary and morphosyntax. However, there is no information as to whether the feedback influences pragmatic development. This dissertation provides empirical evidence that negative feedback during negotiation also promotes L2 pragmatic development, at least in the area of Korean referent honorifics, and thus contributes to the extension of the current boundary of Long’s Interaction Hypothesis. Perhaps this is one of the most significant theoretical contributions made in this study. Still, much more empirical evidence is required to further update the Interaction Hypothesis. Within the interactionist paradigm, to my knowledge, there are only three studies that explored the impact of feedback on the development of pragmatics. The facilitative role of interactional feedback in pragmatic development attested in this study corroborates that of prior studies. Fukuya and Hill (2005), Soler (2002), and K. Jeon (2007) reported that interactional feedback contributes to the learning of pragmatics. Fukuya and Hill’s work involves Chinese 121 EFL learners’ development in the speech act of requests, and Soler’s study concerns EFL learners’ pragmatic competence of requests. Only K. Jeon’s investigation on Korean honorifics is directly pertinent to this dissertation. In K. Jeon’s study, the weakest effects of interactional feedback, though statistically significant, were found in the development of honorifics compared to morphosyntactic and lexical targets. It was argued that interactional feedback, primarily in the form of recasts, was facilitative in the development of honorifics, but was not as facilitative as in other linguistic areas because communications between conversants are possible without the use of honorifics, and thus Korean honorifics are non-salient and have a low communicative value. In contrast to K. Jeon (2007), this dissertation found a strong impact of interactional feedback on honorific development. When the efficacy of recasts was computed in isolation in order to make a direct comparison with her study, the feedback was still fairly effective. A closer inspection of the two studies may explain the different results. The majority of the participants were HL learners in her study, whereas a balanced number of HL and NHL learners took part in the present study. Because this dissertation found that the NHL learners appeared to learn honorifics somewhat better than the HL counterparts, a much higher proportion of the NHL learners in the present study may have enhanced the positive learning outcomes. Furthermore, different scoring guidelines can explain the different results between the two studies. For example, the scoring schemes in K. Jeon’s study were based on the accuracy of subject-verb honorific agreement in subject honorific development. In other words, one point was awarded only if both the honorific subject case marker -kkeyse(nun) and the honorific verbal morpheme si were correctly used and agreed in an honorific sentence. There was a penalty for learners’ inflectional errors in the predicate. In contrast, this dissertation focuses on the use of the target honorific features, not subject-verb honorific agreement, and thus each of the target features was 122 coded and scored separately. For example, one point was awarded to the use of the honorific subject case marker -kkeyse(nun), and another one point was given to the presence of the honorific verbal morpheme –si in an honorific sentence. There was no penalty for pronunciation, vocabulary, grammar, and morphology errors, including predicate inflection, as long as the target honorific features appeared in an honorific sentence. The scoring guidelines in the current study may be less conservative than those of K. Jeon, but the different scoring systems had to be established in order to accommodate dissimilar linguistic behaviors of each of the HL and the NHL learner populations and assess their honorific development in a fair and unbiased manner. These less conservative scoring guidelines may have made the effects of interactional feedback stronger in this study. As a final point, the use of real-life situations and authentic photos with the appearance of the researcher’s genuine family members and acquaintances may have encouraged the learners to use honorifics. As K. Jeon admitted, hypothetical situations and fictional characters appeared in the picture drawings in her study may not have made learners feel a strong need to use honorifics. As such, the current study suggests that the use of Korean honorifics in the pragmatic domain is truly context-sensitive (Byon, 2002; J. Lee & Shin, 2008). In fact, if an honorific person had been in presence during the experiment, the learners’ honorific use might have been increased. The strong effects of interactional feedback in learning pragmatics demonstrated in this study are surprising given that pragmatics is one of the most difficult areas for L2 learners (Gass et al., 2013; Gass & Selinker, 2008), and that such a significant improvement was achieved through a relatively short duration of treatments in comparison with instructional intervention studies in L2 pragmatics. Kasper (2001) noted that pragmalinguistic knowledge is associated with mappings of form, meaning, force, and context, and often with appropriateness of linguistic 123 forms. If pragmalinguistic problems are limited to short utterance segments, it may be reasonably easy to locate and recast. However, sociopragmatic knowledge is related to action-relevant context factors and communicative action, but not necessarily to specific linguistic forms. Therefore, sociopragmatic problems may be harder to find and correct. Interactional feedback was found to be effective in this study perhaps because the honorific form as a linguistic target clearly pertains to appropriate pragmalinguistic choices, which are easily identified and corrected. The strong impact of interactional feedback on L2 pragmatic development in this study is consistent with the positive effect of instructional intervention in L2 pragmatics research. Research suggested that sole exposure is not sufficient to achieve a nativelike mastery in L2 pragmatic development (Bardovi-Harlig & Hartford, 1993; Bouton, 1996). This is probably true as far as the acquisition of the Korean honorific system is concerned. The great majority of classroom intervention studies in L2 pragmatics revealed that administration of instruction is more effective than no instruction, and pragmatics is indeed teachable in the classrooms (see Bardovi-Harlig, 2001; Kasper, 2001, for review). Empirical evidence showed that instruction accelerates learners’ pragmatic development (Billmyer, 1990; Bouton, 1994; Halenko & Jones, 2011; Lyster, 1994; Yoshimi, 2001). Similarly, many aspects of L2 pragmatics are not acquired without the assistance of instruction, or they are learned more slowly. Thus, instructional intervention may be not only facilitative but also necessary for the acquisition of L2 pragmatic ability (Kasper & Rose, 2001). However, the majority of the intervention studies in interlanguage pragmatics are based on the learning of speech acts, that is, strategic use of pragmatics, and instructional intervention merely subsumes feedback as one of its components. Thus, the results in this study may not be completely paralleled to those in L2 pragmatics studies, and caution should be used when making general statements regarding pragmatic learning. 124 Heritage vs. Non-heritage Language Learners The differing levels of performance between the HL and the NHL learners in this study may be attributable to different language learning processing strategies each group of learners employs. It is speculated that the strong meaning-based orientation of the HL learners may have yielded their underperformance in the learning of honorifics. Gass and Lewis (2007) found that the HL learners focused more on semantic issues than on linguistic form and often disregarded interactional feedback directed at morphosyntax. Consistent with their findings, the current study implies that the HL learners may be more interested in non-linguistic features and content rather than linguistic form involving honorifics during conversation. It is possible that their primary interest may have been to communicate with the NS interlocutor while engaged in conversational interaction, and they may have been more likely to regard communicative tasks in the experiment as real-life conversations than the NHL learners. For example, the HL learners tended to make comments on or ask questions about Korean culture reflected in the photos, the researcher’s and their own family background, as well as people, objects, and scenery in the photos. Example (13) below is drawn from the data in the current study and shows a typical interactional pattern between the HL learners and the NS interlocutor. It appears that the HL learner focused more on content than on form during conversation given that his main interest was centered on discovering what it was in the cup, for example, orange juice, water, or barley tea. Hence, the HL learner disregarded the honorific use throughout the entire conversation despite that implicit interactional feedback directed at the target honorific items was provided. Example 13. HL Learner and NS Interlocutor in the Implicit Feedback Condition 125 NS: I These salamtuli may hayyo? people-S what do-V What do these people do? HL learner: Swuhuy-ka halmeni-hanthey… Ike Swuhuy -S grandmother-IO mwe yeyyo? this-S what Swuhuy to the grandmother… be-V (pointing out the photo) What is this? Oleyncicwusu? cweyo khepey Orange juice orange juice-DO give-V cup-LOC Orange juice? NS: A, Oleyncicwusu Giving orange juice Swuhuy-ka pour-V pouring in the cup. halmeni-kkey oleyncicwusu-lul grandmother-IO(H) Ah-EX Swuhuy-S tamase. orange juice-DO tulyeyo.  recasts give-V(Hum) Oh, Swuhuy gives orange juice to the grandmother. HL learner: (Looking at the photo carefully) A, ike mwul Ah-EX this-S water iyeyyo? Policha be-V barley tea yeyyo? be-V Oh, is this water? Is this barley tea? Wuli Our-POSS halmeni cip-ey policha isseyo. grandmother home-LOC barley tea-S be-V 126 There is barley tea at my grandmother’s home. NS: Kulsseyyo. Well. HL learner: Swuhuy-ka halmeni-hanthey mwul-ul cwunapwayo. Swuhuy -S grandmother-IO water-DO appear to give-V Swuhuy appears to give water to the grandmother. In contrast, the NHL learners appeared to be very attentive to their own output as well as to linguistic details available in input and consider communicative tasks with the NS interlocutor as a language exercise as they would normally do with their teachers in the classrooms. The following example (14) shows how interactive conversation typically took place between the NHL learners and the NS interlocutor in this study. The feedback episode below suggests the NHL learner’s form-oriented approach to language learning. Based on the NHL learner’s comments made during her turns, it seems that the NHL learner consciously focused on linguistic form, attempted to retrieve the appropriate form from her memory system, requested confirmation to the NS interlocutor about the proper use of honorifics, and repeated the NS interlocutor’s full, correct honorific sentence in order to practice. During her turns, the NHL learner self-corrected her output from the non-honorific indirect object marker form -hanthey to the honorific indirect object marker form -kkey. She appeared to be uncertain about the use of appropriate form between the honorific form cwuseyyo and the humble form tulyeyo in the given context. After receiving a recast, she incorporated the feedback and modified her output 127 successfully. This learning processing strategy is anticipated to assist the NHL learners’ honorific development and thusly may have resulted in excellent performance for these learners in this study. Example 14. NHL Learner and NS Interlocutor in the Implicit Feedback Condition NS: I These salamtuli may hayyo? people-S what do-V What do these people do? NHL learner: Swuhuy-ka hala, halmeni-hanthey… Swuhuy-S kkey, kkey, right? gramdmother-IO mwul-ul cwe, Water-DO IO(H), IO(H) cwe… no… I have learned this, but I forgot… sorry... give-V give-V cwuseyyo? No, it’s not right… Umm… Give-(H)* 1 Swuhuy gives water to the grandmother. NS: A, Ah-EX Swuhuy-ka Swuhuy-S halmeni-kkey mwul-ul grandmother-IO(H) 1 water-DO(H) tulyeyo.  recasts give-V(Hum) * indicates that an inappropriate form was used in the given context. In this example, the humble form tulyeyo should be used instead of the honorific form cwuseyyo. 128 Oh, Swuhuy gives water to the grandmother. NHL learner: Oh, yeah, tulyeyo, to the older person… give-V(Hum) Swuhuy-ka halmeni-kkey mwul-ul tu, tu, tulyeyo. Swuhuy-S grandmother-IO(H) water-DO give-V(Hum) Oh, Swuhuy gives water to the grandmother. In addition to the meaning-focused language processing, the HL learners are subject to the top-down approach (Celce-Murcia & Olshtain, 2000; Z. Han, 2004; Kagan, 2005). This adds further explanations of why the HL learners in this study did not appear to attend to details of the honorific targets as long as they succeeded in communicating with the interlocutor. The literature on HL learners indicated that the HL learners tend to acquire the HL in an implicit way and rely on their intuition as L1-acquiring children (Carreira & Potowski, 2011; Montrul, 2010a). If the adult HL learners still adopt this intuition-based, implicit language learning process, their relative insensitivity to honorific forms and pragmatic rules and principles, as shown in this study, can be better explained. In conjunction with the form-focused orientation, the NHL learners are known to utilize the bottom-up approach to language learning (Celce-Murcia & Olshtain, 2000; Kagan, 2005). As implied in this study, this group of learners showed a tendency to attend to each of the honorific features to create a context-appropriate honorific sentence, and this approach reflects such linguistic behavior. Also, as a way of learning a language explicitly, instructed adult L2 learners 129 generally resort to metalinguistic knowledge as well as analytical learning methods (Carreira & Potowski, 2011). The combination of these characteristics of classic adult L2 learners may have enabled the NHL learners to focus on honorific forms, attend to metalinguistic properties and explanations of honorifics in input, analyze honorific rules, and retrieve their metalinguistic knowledge from their working memory. Through this process, it is inevitable that the NHL learners made a remarkable progress in the development of honorifics in this study. Along with these dissimilar language learning processing strategies, the HL learners’ underperformance in Korean honorifics demonstrated in this study can be explained in light of the characteristics of the HL learner population. Most of all, the contrasting performance between the HL and the NHL learners is ascribed to their prior target language learning experience, perhaps the most prominent difference between these two types of learners. Although the Korean HL learners have an advantage of being exposed to the Korean language input at home as a child, they are usually accustomed to using the language restricted to home routines. This includes varying degrees of honorific use, typically, from the complete absence of honorific use to the limited use of honorifics (Wang, 1995). Recall that Korean honorifics may not be always used between children and parents or grandparents within home environments. In a close relationship, intimacy may override a power relationship. In close-knit families, the young may habitually address the old with non-honorifics, and this practice is often regarded as a demonstration of affection and a bond between family members, not necessarily as lack of deference (Kim-Renaud, 1990). Since the Korean HL learners in the United States suffer from impoverished Korean language input and limited opportunities to use the language outside home, their use of honorifics is unlikely to be fully developed contrary to Korean native-speaking monolingual children who develop an age-appropriate use of honorifics by experiencing a 130 variety of situations that involve honorifics. The HL learners’ childhood language experience as an overhearer, not as an active user (e.g., speaker), may not have helped their honorific development in this study. Au and Romo (1997) found that relearners, including overhearers, might have an advantage in phonology and morphosyntax relative to first-time learners of the language and concluded that even incomplete or brief childhood language learning experience might help adults’ phonological and morphosyntactic development – typically most challenging areas for late language learners to master (Fledge, 1987; Johnson & Newport, 1989; Newport, 1990, 1991; Oyama, 1976). However, the current study suggests that the benefits of child language experience, such as overhearing, may not be applicable to the development of pragmatics, at least Korean honorifics. Also, literature indicated that childhood overhearing experience may suffice for the phonological development (perception), but childhood speaking experience seems crucial for the development of both phonology (production) and grammar (Au et al., 2002; Au et al., 2008; Knightly et al., 2003; Oh et al., 2003). The present study adds further empirical evidence that learners’ consistent output, not just mere overhearing, seems essential to produce more targetlike form in pragmatics in addition to phonology (production) and grammar. As addressed in Chapter 2, literature indicated that fossilization is one of the central characteristics of HL learning (Montrul, 2010a), and thusly one might argue that fossilization is likely to take place for the HL learners whose honorifics are partially or incompletely acquired. However, a careful investigation of the definitions of fossilization demonstrates that the HL learners’ underperformance in honorific development may not be the result of fossilization. Bley-Vroman (1989a) defined the term fossilization as follows: …a stage short of success—and that learners then permanently stabilize at this stage. 131 Development ceases, and even serious conscious efforts to change are often fruitless. Brief changes are sometimes observed, but they do not take. The learner backslides to the stable state [emphasis added]. (pp. 46-47) Similarly, other scholars also characterized fossilization as an ultimate stage or an end state of L2 learning and seem to be generally in agreement with the position that fossilization does not assume learners’ potential for improvement (Z. Han, 2012; Selinker & Lamendella, 1978; Tarone, 1994). Yet, although the HL learners may have failed to reject their old habit, either nonhonorific use or inappropriate honorific use in spite of two intensive feedback treatment sessions in this study, it is still possible that the HL learners are able to advance their use of honorifics in the future with the help of external influence, including explicit classroom instructions and feedback as well as study abroad or residence in Korea. Simply put, learning may continue. It may not be the case that the HL learners’ honorific development is permanently ceased, and thus it is implausible for them to improve their use of honorifics even with a considerable amount of effort. For this reason, it is speculated that fossilization may not best explain the HL learners’ underperformance in the learning of honorifics in this study. Instead, frequency effects may better explicate the HL learners’ inferior performance to the NHL learner counterparts in honorific development. N. Ellis (2002) argued that language learners naturally count on input frequency, and thusly frequency effects may play a large part in explaining language processing, language use, and language acquisition. There is a general tendency that the more times learners experience a certain linguistic form, the stronger their memory for it, and the more fluently it is accessed (N. Ellis, 2012). This claim can be also applied to the HL learners’ lack of ability to use context-appropriate honorifics. That is, the absence of honorifics or the infrequent use of proper honorifics, to which the HL learners are 132 accustomed at home environments during childhood, may have been stored more strongly in their working memory and thus activated more easily in their speech. Gass and Mackey’s (2002) claims that the frequency-based explanation may be compatible with interactionist positions, which comprise the areas of input, interaction, and output, also lend further support to the potential role of frequency effects addressed in this study. Both simplification and overgeneralization may also account for the HL learners’ inferior performance in this study. As addressed earlier, research as well as anecdotal evidence showed that Korean HL learners are known to struggle with using honorifics properly (K. Jeon, 2004, 2007). Consequently, the HL learners in this study may have simplified the complex honorific use to make the language easier and may have been dependent on the use of only a limited range of honorifics or non-honorifics, which reflects the process of language simplification. Also, it is possible that the HL learners in this study may have applied the small number of honorifics they were familiar with or non-honorifics onto many other contexts inappropriately, which is a case of overgeneralization (Brown, 2011). One might argue that the poor performance of the HL learners demonstrated in this study is due to language attrition. This argument may be based on the following assumption that the HL learners used context-appropriate honorifics in childhood, but they went through the process of language attrition upon entering school. However, this claim is questionable considering that although Korean monolingual children start to acquire honorifics around the age of two, acquisition of a full spectrum of honorifics may not be realized until they are in university (L. Brown, personal communication, January 15, 2013). The Korean HL learners already transition to English-dominant environments around the age of five; therefore, a full mastery of the Korean honorific system at this school age is implausible. This supports Keating et al.’s (2011) claims 133 that HL learners are not advantaged compared to NHL learners regarding grammatical features that are acquired later in life or that require a sufficient amount of uninterrupted input, which may be the case for Korean honorifics. It is also consistent with Montrul’s (2011b) findings that HL learners are more targetlike than NHL learners in early-acquired properties, but not in lateacquired grammatical areas. Therefore, since honorifics are late-acquired features, and thusly the Korean HL learners are unlikely to have acquired a complete honorific system, the HL learners’ underperformance in this study is more likely to be the results of incomplete acquisition than language attrition. Therefore, it is reasonable to speculate that as a result of incomplete acquisition in childhood, the HL learners continue to struggle with the honorific use throughout adulthood. However, this needs empirical validation to make further arguments. In contrast, the NHL learners have no exposure to the target language at home in childhood, and their target language learning typically starts in the language classrooms. They rarely have childhood overhearing experience of the target language. As an adult L2 learner, language learning is generally built upon accumulation of rules through formal classroom instruction. Due to the absence of prior target language use in childhood, language interference is less likely to occur and impede the honorific development. This may explain the NHL learner’s excellent performance in the learning of Korean honorifics. Surprisingly, L1 sociopragmatic or pragmalinguistic transfer did not seem to be present for these learners in this study. Although their primary language, English, does not have an honorific system, the majority of the NHL learners appeared to be conscious of the importance of honorifics in Korean, make an effort to attend to honorific features, and incorporate the NS interlocutor’s feedback in order to use honorifics correctly. Rather, dissimilarities between English and Korean, such as existence of the honorific system, may have helped them pay additional attention to the use of Korean honorifics 134 during conversation. The different motivational orientations between these groups of learners may provide a partial explanation for the different learning outcomes in the development of honorifics. In recent years, a dramatically increasing number of NHL students volunteer to take Korean courses only to enhance understanding of Korean language and culture. As observed in this study, this group of learners appeared to be highly motivated to learn Korean, and those students who planned to study abroad in Korea seemed to be particularly eager to learn. However, while pursuing ethnic roots has always attracted Korean HL learners to learn Korean, there was a tendency that a great portion of the Korean HL learners took Korean courses primarily to meet a foreign language requirement for graduation, as indicated in the language background questionnaires in this study. In consideration of this situation, it may not be unreasonable to assume that, in some sense, the NHL learners may have been more intrinsically motivated to learn Korean than the HL learners. As a result, this may have led to successful learning of honorifics for the NHL learners. However, again, more empirical evidence is necessary to confirm this speculation. Example (15), drawn from the data in this study, shows one of the typical linguistic patterns of honorific use that distinguishes the HL learners from the NHL learners. It should be recalled that the proper use of the Korean honorific system requires honorific-specific morphological knowledge as well as pragmatic knowledge of what is socially acceptable and polite in a given context. In this example, the HL learner made no morphological errors, but failed to employ any honorific features in the sentence. In contrast, although the NHL learner committed an inflection error in the predicate, as in pwa-sey-yo, she inserted the honorific subject case marker –kkeyse as well as the honorific verbal morpheme –si, which is appropriate in the given situation. It should be reminded that since this dissertation focuses on the use of the 135 2 target honorific features, no penalty was given for predicate inflection errors. Similar to K. Jeon’s (2004) claim, this example indicates that the HL learners’ underperformance in the honorific development may have been more likely to be from their lack of sociopragmatic knowledge than morphological knowledge. On the contrary, the NHL learners may have taken into consideration both honorific morphology and sociopragmatic rules in their use of honorifics. Example 15. HL Learner and NHL Learner HL learner: Halmeni-ka Grandmother-S kewul-ul mirror-DO pwayo. look at-V The grandmother looks at the mirror. NHL learner: Halmeni-kkeyse Grandmother-S(H) kewul-ul mirror-DO *pwa-sey-yo. 3 look at-V(H) The grandmother looks at the mirror. There were unexpected, but interesting findings in this study that the HL learners 2 One might argue that the NHL learners’ occasional commitment of predicate inflection errors is inconsistent with their relatively strong ability in morphosyntax indicated in the language background questionnaire (see Appendix A). While the NHL learners are known to have solid morphosyntactic knowledge, they often produced predicates with wrong inflection during conversation due to the difficulties in online processing and production. In contrast, although morphosyntax is a weaker area for the HL learners, they generally produced predicates with correct inflection quickly and automatically without referring to the predicate inflection rules. Because only predicates that are commonly used in daily basis were included in the communicative tasks for elicitation in this study, the HL learners may have been already accustomed to producing these predicates as a child at home and thusly may not have had difficulties in making predicate inflection during online production. 3 * indicates a verb inflection error. 136 appeared to learn object honorifics better than subject honorifics; however, a reverse pattern was observed for the NHL learners. The relative easiness the HL learners may have experienced to learn object honorifics can be, in part, attributable to the humble form -(e)tulida. This humble form in object honorifics may be one of the most frequently and productively used verbs to express politeness (Brown, 2011). It is possible that the HL learners may not have possessed solid knowledge with respect to rules and principles that govern object honorifics; however, this humble form -(e)tulida may have been already stored as a lexical item in their working memory through their childhood experience with the HL and retrieved quickly, easily, or automatically through treatments in this study. This is in line with the claims that more frequently used words are accessed and perhaps learned faster than less frequent ones (A. Ellis & Morrison, 1998; Jescheniak & Levelt, 1994; Oldfield & Wingfield, 1965), and that frequency effects influence language acquisition (Bybee & Hopper, 2001; N. Ellis, 2002). Wang (1995) indicated that Korean parents in the United States usually neither emphasize the importance of honorific use nor afford corrective feedback on their children’s wrong honorific usage. However, because this humble form -(e)tulida is widely used in Korean, it is not uncommon to observe that parents correct their children’s improper utterances involving this particular humble form by providing either implicit or explicit feedback during child-adult interactions in the United States as well as in Korea. Thus, although many HL learners did not use this humble verb on the pretest, they may have been able to retrieve it relatively easily and fast during the treatment sessions and continued to use it without much difficulty throughout the conversation. Conversely, the NHL learners appeared to show better performance in learning subject honorifics than object honorifics perhaps due to high input frequency of subject honorifics in Korean. Contrary to subject honorification, however, Korean object honorification is generally 137 described as nonproductive (Yun, 1993) with the exception of the humble form -(e)tulida and employs lexically restricted substitutions (Brown, 2011; Y. Kim, 1997). Moreover, it is possible that the NHL learners may have experienced confusion between subject and object honorification. Specifically, although the comparison should be made between the subject and the object of the sentence to produce object honorifics, the NHL learners in this study seemed to tend to include the speaker in this reference construction and compare the object against the speaker similar to subject honorification (I. Lee & Ramsey, 2000). For instance, the NHL learners occasionally provided the honorific form -cwuseyyo incorrectly where the humble form (e)tulida was required. Alternatively, since the cognitive complexity is required to process or produce object honorifics, it may have been challenging that the learners had to elevate the object while downgrading the subject (Y. Kim, 1997). Though one of the characteristics of HL learners involves their heterogeneity (Kagan & Dillon, 2012; Kanno et al., 2008; Kondo-Brown, 2005; Valdés, 1995), it should be reminded that the HL learners in this dissertation had relatively homogeneous background, and caution is necessary in interpreting the results. For instance, since all the HL learner participants were second-generation Korean-Americans whose both parents were Korean NSs, the learners were exposed to the Korean language and culture and interacted with their parents in Korean at home during childhood and thus generally reported stronger confidence in spoken language in Korean. Although a great number of the HL learners attended weekend Korean schools, their exposure to written language in Korean seemed limited, and they reported weaker confidence in this area (see Appendix C, D, and E). It appeared that most HL learner participants came from welleducated, middle-class families as Korean immigrants in the United States are generally collegeeducated and tend to use the standard Korean language. If a more heterogeneous group of the HL 138 learners had participated, it could have brought different results. Hence, further research is required to generalize the findings obtained in this study. Implicit vs. Explicit Feedback This dissertation generally showed stronger effects of explicit feedback than implicit feedback in learning Korean honorifics. In consort with Sheen’s (2007) claim, the superiority of metalinguistic correction over recasts in the learning of Korean honorifics accounts for Schmidt’s Noticing Hypothesis (1998, 2001), comprising two levels of awareness: noticing and understanding. It should be recalled that while noticing as a lower level of awareness only leads to intake, understanding as a higher level of awareness includes an ability to analyze, compare, and test hypothesis, and thus leads to learning. Recasts assist learners in attending to specific forms available in language input and notice the gap between their erroneous form and the interlocutor’s correct form. In this sense, recasts are limited to the development of lower levels of awareness. In contrast, metalinguistic correction helps learners comprehend underlying rules and principles, and this may be more efficient than recasts in storing knowledge in the working memory. Thus, metalinguistic correction can be more inclusive and effective than recasts given that metalinguistic correction helps develop both levels of awareness. In contrast to recasts, the utilization of metalinguistic correction requires a longer break from conversation to provide metalinguistic information. R. Ellis et al. (2006) indicated that “the metalinguistic time-outs from communicating afforded by explicit correction constitute a perfect context for melding the conscious and unconscious processes involved in learning” (p. 343). The metalinguistic time-outs allow learners to have more time to understand rules and to notice the gap as well as the corrective intention of the feedback than recasts. Therefore, metalinguistic correction may have been more effective than recasts in the present study because the former 139 may have led to deeper learning and longer-lasting memory than the latter. The superiority of metalinguistic correction over recasts can also be explained in terms of types of evidence. Metalinguistic correction provides both positive evidence (i.e., provision of a correct form) and negative evidence (i.e., clear indication of error occurrence). Likewise, recasts also provide learners with both types of evidence (Gass, 2003). However, the corrective intention of recasts may not be noticed by learners due to the ambiguous nature. Hence, it is more likely for the learners to recognize both positive and negative evidence with metalinguistic correction than with recasts. As a result, metalinguistic correction may have supplied more opportunities for learning and thus was more effective than recasts in this study. Unique qualities of honorifics as a linguistic target may be another factor that resulted in the weaker effectiveness of recasts than metalinguistic correction in the present study. As Korean honorifics have a low communicative value and lack salience, they rarely cause communication breakdowns (K. Jeon, 2007). This means that the presence or absence of referent honorifics is unlikely to influence the actual referential meaning of the message during conversation (Brown, 2011). In addition, because recasts are implicit in nature, they do not interrupt the natural flow of interaction (Long, 2006). Thus, it is reasonable to think that when recasts were provided on the non-salient linguistic target in this study, the corrective intent of recasts may have been less likely to be noticed by the learners, and this may have resulted in the weaker effects of recasts. Consequently, more explicit feedback may be needed to draw learners’ attention to linguistic mismatches between the erroneous form and the correct form, especially when the linguistic target, such as honorifics, is not salient and does not cause a communication breakdown. This may be also related to Tateyama’s (2001) suggestion that explicit instruction may be more beneficial for learners in learning pragmatic areas that require higher formality as well as routine 140 formulas associated with age, social status, and in-group and out-groupness. This study is in line with Sheen’s (2007, 2010b) findings in that metalinguistic correction showed stronger effects than recasts in the learning of English articles, namely, non-salient linguistic features. However, the results of this study are also in contrast with those of Sheen’s work. While recasts were indeed effective for the honorific development in this study, recasts involving article errors had no significant positive effects in her study. One might suspect that different research settings may have influenced the differential efficacy of recasts between these studies: While this study was conducted in the laboratory, her study was undertaken in the classroom. Some researchers maintained that recasts in the classroom may not be as noticeable and effective as in the controlled laboratory (e.g., Lyster, 1998a, Lyster & Ranta, 1997; Panova & Lyster, 2002). As recasts are directed at a whole class in the classroom context, it may be easier for students to be distracted and harder for teachers to provide the feedback consistently. Hence, learners in the classroom often perceive recasts as feedback on content, such as either confirmation check or repetition of learner utterance (Lyster, 1997, 1998a). In contrast, recasts can be interpreted as corrective feedback by learners more easily in the laboratory setting because the feedback normally focuses on a single linguistic target and is directed at an individual learner. Also, recasts are provided in a more intensive and consistent manner in less distracted environments (e.g., Carroll & Swain, 1993; Egi, 2007b; Ishida, 2004; Iwashita, 2003; Leeman, 2003; Long et al., 1998; Lyster & Izquierdo, 2009; Mackey & Philp, 1998; McDonough, 2007; Sagarra, 2007). However, in spite of such claims as the above, Gass, Mackey, and RossFeldman (2005, 2011) found no significant differences between classroom and laboratory settings in their investigation of the impact of research settings (i.e., classroom vs. laboratory) on task-based interactions. They further emphasized that the effects of settings on interactional 141 processes should not just be assumed and claimed, but should be empirically demonstrated. As such, it should not be simply assumed and concluded that different research settings are attributed to different results between Sheen’s and this study. Instead, although both of the linguistic targets lack salience, it is possible that English articles in her study may have been more difficult to be noticed or learned by learners during short treatments and thus may need more feedback than Korean honorifics in this study. This is in line with Loewen’s (2012) claim that different linguistic items may require more or less feedback. To date, the mixed results on the effects of implicit and explicit feedback have been reported in interaction research. Yet, this study affords additional empirical evidence to a body of interaction research studies that found a superiority of explicit feedback over implicit feedback (e.g., Nagata, 1993; Rosa & Leow, 2004); more specifically, metalinguistic feedback over recasts (e.g., Carroll, 2001; Carroll & Swain, 1993; R. Ellis et al., 2006; Sheen, 2007, 2010b). Stronger effects of explicit feedback than implicit feedback in the learning of Korean honorifics in this study are consistent with research evidence in instructed SLA (e.g., Norris & Ortega, 2000; Spada & Tomita, 2010) as well as interlanguage pragmatics (e.g., Alcón Soler, 2005; Nguyen et al., 2012; Rose & Ng Kwai-fun, 2001; Takahashi, 2001). The superiority of explicit feedback is also in line with research that learners prefer explicit feedback to implicit feedback (H. Kim & Mathes, 2001; Nagata, 1993). However, this dissertation is in contrast with those studies that did not find a statistically significant difference in the effectiveness of implicit and explicit feedback within the interactionist paradigm. For example, there was no statistical difference between recasts and metalinguistic feedback in H. Kim and Marthes (2001). Because each group included only ten participants, it is assumed that the lack of significant difference is ascribed to the small sample 142 size. Also, Loewen & Nabei (2007) did not find significant differences in the efficacy of recasts, metalinguistic feedback, and clarification requests on the timed grammatical judgment test. There were no effects of feedback on the untimed grammaticality judgment test or the oral production test. They acknowledged that the lack of significant differences may have been as a result of the insufficient treatment time and the number of feedback provided. Along with the present study, a general trend in pragmatic instruction is in support of explicit instruction. However, there are some studies that did not find a statistically significant difference between implicit and explicit instruction in pragmatics research. As Rose (2005) argued, this may be due to methodological flaws in these studies. For instance, Kubota’s (1995) work on implicature comprehension showed that implicit instruction worked better than explicit instruction, but this study raised serious validity issues on treatment and testing measures. Fukuya and Clark (2001) reported no significant difference between implicit and explicit instruction on the target structure, but they admitted that the treatment was too short to find instructional effects. Taken together, the strong effectiveness of explicit feedback demonstrated in the current study provides further evidence of a superiority of explicit feedback over implicit feedback in interaction research. The results also correspond to the supremacy of explicit instruction in instructed SLA and interlanguage pragmatics research. Heritage vs. Non-heritage Language Learners in relation to Implicit vs. Explicit Feedback This study demonstrated that the NHL learners in the explicit feedback condition showed the strongest development in learning honorifics. The strong effects of explicit feedback in tandem with the NHL learners’ inherent explicit approach to language learning shown in this study can possibly explicate the outstanding performance of this group of learners. 143 In spite of the best performance of this NHL/Explicit group and the overall pattern indicating that explicit feedback was more effective than implicit feedback for the NHL learners, there was no statistically significant difference between the two types of feedback for the NHL learners except on the delayed posttest in object honorifics, which found a superior efficacy of explicit feedback. One possibility for the lack of significant difference may be that the NHL learners’ distinct language learning process may have overridden the explicitness of feedback. In other words, the NHL learners’ typical form-focused, bottom-up, analytical, metalinguistic, and explicit learning approaches may have strongly affected the learning of honorifics regardless of the feedback conditions. As discussed earlier, traditional adult L2 learners are trained through formal classroom instruction that subsumes explicit rule presentations and explanations provided by their teachers. Using metalinguistic knowledge and skills they learned in the classroom, these learners are subject to employ the bottom-up approach to language learning. Explicit knowledge about linguistic rules that underlie the target language is essential for them to use the language for a communication purpose until they gain a completely automatic control of the language. As such, given that adult foreign language learners typically rely on form-focused, bottom-up, analytical, metalinguistic, and explicit ways of language learning, the NHL learners in this study may have depended upon the explicit language learning process even in the implicit feedback condition. The NHL learners’ proficiency as opposed to the HL learners’ may also account for this non-significant difference. In general, the NHL learners have heavier cognitive loads for online processing and production and thus are less fluent than the HL learners. Hence, the NHL learners were slower in producing a sentence in this study. Also, since the NHL learners tend to have less vocabulary knowledge than the HL learners (Kagan & Dillon, 2012), the former produced more pauses than the latter to retrieve the vocabulary from their mental lexicon during 144 conversation in this study. All of these characteristics of the NHL learners’ proficiency provided them with additional time irrespective of the feedback conditions, and this may have helped make the effects of both feedback types comparable for these learners. This extra time may have been similar to the metalinguistic time-outs from communicating provided by explicit correction in the explicit feedback condition. As R. Ellis et al. (2006) argued, these time-outs may have served a perfect context for combining conscious and unconscious processes relevant to learning and increased the effects of both feedback types after all. One of the most interesting findings in this study is that the HL learners who received implicit feedback demonstrated the weakest development with the exception of the control group. Also, the developmental pattern of this group was in sharp contrast with the rest of the experimental groups. Possible explanations lie in the HL learners’ weaker performance than the NHL learners in the honorific development in conjunction with the relatively weaker effects of implicit feedback than explicit feedback found in this study. This study showed that explicit feedback was significantly more effective than implicit feedback for the HL learners. Although most of the HL learner participants in this study were both naturalistic and instructed learners, their experience of learning the HL implicitly and intuitively at home environments seemed still dominant, at least in the development of Korean honorifics. Their primary goal during interactive conversation in the experiment appeared to get their message across. They tended not to attend to linguistic form unless it hindered their communication or unless interactional feedback was explicit enough to bring their attention to discrepancy between their erroneous output and the interlocutor’s correct form. Nonetheless, because the omission or the misuse of honorifics generally does not influence the actual referential meaning of the message (Brown, 2011), communication breakdowns rarely occurred 145 in this study. Moreover, the implicit and ambiguous nature of recasts in the implicit feedback condition may have often failed to alert the learners to notice the gap unlike explicit correction. Accordingly, their limited or inappropriate use of honorifics as a result of incomplete acquisition or avoidance of using honorifics still seemed to persist as an adult even after receiving intensive implicit feedback. For all these reasons, the HL learners may have paid little attention to linguistic form in honorifics unless explicit feedback was provided. Therefore, as the findings of this study suggest, explicit feedback is indeed more effective than implicit feedback for the HL learners, at least in the learning of Korean honorifics. The discussion above may be linked to the issue raised in Montrul (2010a). Recall that Montrul pointed out the following questions as the most important investigation in HL acquisition: do adult HL learners in the classroom learn the target language implicitly as children acquiring L1? Or, do they now rely on explicit learning like adult L2 learners? In response to her questions, the findings of this study suggest that while the HL learners learn the HL in an explicit manner in the classroom, they still appear to acquire it implicitly to some extent. However, whether the HL learners acquire the HL more explicitly or implicitly may also depend on several factors, such as linguistic features and task types (Keating et al., 2011; Montrul, 2011b). For instance, the HL learners are more likely to learn implicitly involving early-acquired features and tasks that require minimal metalinguistic knowledge. Conversely, they may learn more explicitly concerning late-acquired features and written tasks that entail learners’ metalinguistic knowledge. In this study, explicit feedback was more effective than implicit feedback for the HL learners. As explicit learning subsumes explicit feedback, this suggests that the HL learners are amenable to an explicit way of learning. Simultaneously, notably weak development of the HL learners in the implicit feedback condition particularly in the learning of subject honorifics shows evidence of 146 implicit learning. In the implicit feedback condition, the NHL learners still appeared to learn the honorific features explicitly. In contrast, it seemed that the HL learners were strongly influenced by their inherent implicit learning orientation, and this may have resulted in the weakest development. It is assumed that the adult HL learners still resorted to the intuition-based language learning processing they had used as a child at home, and thus they were still inclined to learn the HL implicitly without referring to explicit linguistic rules. In addition, implicit feedback may not have been as helpful as explicit feedback in alerting the HL learners to attend to linguistic details and notice the mismatches. As a result, this implicit learning process in combination of the provision of implicit feedback may have made the HL learners pay little attention to detailed linguistic features of subject honorifics and led to little improvement in the learning of honorifics. Taken together, it is possible that the HL learners’ natural and implicit language learning processing is in transition to an analytical and explicit approach to language learning. Yet, this interpretation in relation to Montrul’s inquiry is only intuitive, and caution must be applied when making further claims. More empirical research that directly investigates this issue is strongly needed. Concerning the relative developmental outcomes between subject and object honorifics, the unexpected finding was obtained. While the HL learners in the implicit feedback condition made a statistically significant improvement over time in learning object honorifics, but not in subject honorifics. This suggests that for the HL learners, implicit feedback may be more conducive to the development of object honorifics than subject honorifics. However, further empirical evidence is a must to confirm this finding. The little development of the learners in the HL/Implicit group, in particular, in subject honorifics, echoes K. Jeon’s (2007) investigation in the role of interactional feedback. In her 147 study, Korean subject referent honorifics demonstrated the least gains from the pretest to the posttests in comparison with other linguistic targets under investigation. Interactional feedback, mostly in the form of recasts, was provided for participants who were mainly HL learners. K. Jeon attributed the non-salient feature and low communicative value of Korean honorifics as well as negative pragmatic transfer from the learners’ primary language, namely English, to L2 Korean to the least improvement in pragmatics. Since the research design features and findings in K. Jeon’s work on Korean honorifics are relevant to the HL/Implicit group in this study, similar discussions can be applied to explain the outcomes of this group. In her recent doctoral dissertation, Y. Han (2010) explored the relative effects of implicit feedback in the form of recasts and explicit feedback in the form of metalinguistic feedback on the acquisition of Chinese classifiers by Chinese HL and NHL learners. She found that the HL learners did not show any statistically significant advantage over the NHL learners in the same feedback condition, but the HL learners in the explicit feedback condition perceived feedback more accurately than the NHL learners in the implicit feedback condition. Also, although there was a significant difference between the experimental groups and the control group, no significant difference was found between the implicit and the explicit feedback groups. On the oral imitation test, neither feedback type nor language background influenced the results: Two groups of learners benefited equally from the same type of feedback. On the contrary, on the written cloze test, feedback type significantly affected learners’ development: Explicit feedback was more effective than implicit feedback, but it was only for the NHL learners. Although Y. Han (2010) explored relevant issues to the current study, the results of her study differ from those of this study in many respects. For the most part, no clear difference was established between the two types of feedback and between the two learner populations in her 148 study. One possible explanation lies in different operationalizations of interactional feedback between the two studies. For instance, metalinguistic feedback, which does not involve the provision of a correct form, may not have been sufficient for learners to modify their output in her study. As Li (2009) pointed out, a simple statement, such as “the classifier was wrong,” may have been unlikely to enable learners to correct their inaccurate use of classifiers on their own, unless the correct form already existed in their interlanguage. Because the linguistic focus involved Chinese classifiers in the lexical domain, the provision of a correct form may have been necessary. This may explain why metalinguistic feedback failed to bring differential results from recasts. In contrast, the current study utilized metalinguistic correction that contains not only metalinguistic explanation but also a correct form, and this feedback is generally more explicit than metalinguistic feedback. Therefore, the utilization of metalinguistic correction may have been more effective than metalinguistic feedback and thus enhanced the contrasting effects between implicit and explicit feedback in this study. Different linguistic targets are also attributable to different results in these studies. HL learners are known to possess a good amount of vocabulary in the target language, but it is often limited to home and childhood environments. However, since there is very little research on the HL learners’ lexicon, it is not possible to make any solid claims on their lexical knowledge relative to that of the NHL learners at this point. In Y. Han’s (2010) dissertation, since Chinese classifiers are relatively salient, simple, transparent, and easy to learn even through a short treatment session (Li, 2010b), childhood language input and experience may not have been particularly beneficial to the HL learners relative to the NHL counterparts. In contrast, the present study investigates Korean honorifics, one of the most difficult areas in learning Korean as a second/foreign language even after prolonged exposure to Korean (Brown, 2011). In 149 particular, HL learners are known to experience a great deal of struggles in the use of Korean honorifics (Wang, 1995). This may have yielded the HL learners’ underperformance in the honorific development in the present study. The comparison of these studies demonstrates that different research designs may generate contrasting research findings. It is crucial to design research carefully and use various methodologies, contexts, languages, linguistic targets, and populations because research findings have a great impact on theory establishment and language pedagogy. Pedagogical Implications This study suggests important pedagogical implications. As addressed previously, teaching Korean honorifics in the foreign language context is one of the biggest challenges for language teachers due to the complexity of the Korean honorific system. Even advanced-level learners often show their inability to use context-appropriate honorifics. Although L2 Korean practitioners acknowledge this issue, the proper use of honorifics is not sufficiently emphasized either in the textbooks or in the classrooms (Brown, 2010, 2011). It may be because other aspects of language, such as grammar and vocabulary, also have to be taught within a limited class time. However, it should be reminded that grammatical competence does not guarantee pragmatic competence (Bardovi-Harlig & Hartford, 1990; Olshtain & Blum-Kulka, 1985). Concentration on micro-level grammatical accuracy may be at the expense of macro-level pragmatic appropriateness in the foreign language classrooms (Bardovi-Harlig, 2001; Bardovi-Harlig & Dörnyei, 1998; Niezgoda & Röver, 2001). Research indicated that different pedagogical approaches should be employed to cater to different needs for HL learners and traditional L2 learners (Valdés, Fishman, Chávez, & Pérez, 2006). One of the notable findings in this study was that implicit feedback may not work well for 150 the HL learners’ development of Korean honorifics in part due to their meaning-focused orientation. Many SLA theorists agreed that learners may not notice specific input features in meaning-oriented L2 use. In order to enhance the HL learners’ noticing, input should be made salient through input enhancement, which is likely to raise the learners’ consciousness about the target honorific features (Sharwood Smith, 1993). Input enhancement as well as explicit feedback can be a sound pedagogical option for the HL learner population. The HL learners’ weak performance in the honorific development was observed in this study, and this may be attributable partly to their limited use or no use of honorifics at home and childhood language overhearing experience. Research indicated that parents in Korean-American immigrant families in the United States do not put enough efforts on their children’s acquisition of honorifics (Wang, 1995). It is also reported that the parents do not often correct their KoreanAmerican children’s inappropriate use of honorifics at home, at least not consistently despite that parents play a significant role in the honorific acquisition of their children (M. Lee, 1997; Wang, 1995). Consequently, it is suggested that parents should make contextualized, pragmatically appropriate input available to their children from early stages of acquisition (Bardovi-Harlig, 1996, 2001; Bardovi-Harlig, Hartford, Mahan-Taylor, Morgan, & Reynolds, 1991). It is also suggested that the parents should encourage their children to use honorifics at home and provide correction on the inappropriate use of honorifics in a consistent manner. The outcomes of this study suggest that explicit feedback enhances not only noticing but also understanding and thus is more effective than implicit feedback. Explicit feedback is more useful, particularly in the learning of non-salient linguistic features, such as Korean honorifics, or pragmatics that is difficult to master only through language input and many years of residence in the country where the target language is spoken. This is consistent with Kasper’s (2001) claim 151 that while explicit instruction may not prove to be universally more effective, research evidence strongly suggests that explicit teaching of pragmatics is more conducive to learning for college students in postindustrial societies. Nonetheless, this is not to say that the exclusive use of explicit feedback is recommended for L2 teaching practitioners. Doughty and Williams (1998) suggested that a combination of a variety of feedback is likely to be most useful. Hence, it is suggested that Korean language teachers should teach Korean honorifics (Byon, 2004; M. Lee, 1997), Korean sociolinguistic norms (H.-Y. Kim, 2008), and Korean politeness routines (Sohn, 1999) more explicitly than implicitly, make a conscious effort to draw students’ attention to Korean honorific features, and use innovative instructional methods to increase the salience of honorifics whenever the relevant issues are raised in the classrooms (see Byon, 2000, 2004, for instructional suggestions). It is also suggested that the teachers should encourage the learners’ output production and provide consistent explicit feedback through interaction in order to promote noticing of the discrepancies between the learners’ output and the teachers’ interactional feedback (H. H. Kim, 2008). 152 CHAPTER 6 CONCLUSION An abundance of theoretical and empirical research from the interactionist position has shown that interactional feedback promotes morphosyntactic and lexical development; however, the impact of feedback on pragmatic development has hardly been investigated. Furthermore, although the relative effects of implicit and explicit feedback have been a prolific area of investigation, the results still remain inconclusive. Recently, there have been a couple of attempts to explore how learners of HL background make use of learning opportunities through interactional feedback during conversation in comparison with traditional L2 learners. Yet, research on this topic is still in its incipient stage, and it may be premature to make any solid claims at this time. Therefore, it was considered necessary to conduct empirical research and provide direct evidence for these underexplored research areas and their potential interrelationships. Noteworthy is that this dissertation is the first attempt to investigate the interesting relationships between the hardly-explored areas in interaction research, such as pragmatics and HL learning, and explicitness of feedback. The purpose of this dissertation was to examine factors that may play a role in L2 development through conversational interaction. More specifically, it explored whether and how learners’ HL background and explicitness of interactional feedback influence the learning of L2 pragmatics in Korean, a language typologically different from English. Overall, the results indicated that: (a) interactional feedback was significantly effective in promoting the learning of Korean honorifics; (b) the NHL learners significantly outperformed the HL learners; (c) explicit feedback was significantly more effective than implicit feedback; and (d) explicit feedback was significantly more effective than implicit feedback for the HL learners, whereas a significant difference was not observed between the two feedback types for the NHL learners. In addition, 153 while the NHL learners who received explicit feedback performed the best, the HL learners who received implicit feedback showed the worst performance among the experiment groups. There was no significant difference between the HL learners who received explicit feedback and the NHL learners who received implicit feedback. This study provides empirical evidence in support of the theoretical argument that interactional feedback promotes different language areas. As addressed in Chapter 2, the Interaction Hypothesis (Long, 1996) argued that negative feedback may be facilitative of L2 development, at least of vocabulary and morphosyntax. This study provides further information that negative feedback facilitates L2 development of pragmatics in addition to vocabulary and morphosyntax and thus makes an important contribution to the theoretical foundation in interaction research. The different learning outcomes among the different groups in this study may be explained by various factors and their interaciton. The findings from this study confirmed and expanded previous research in the areas of conversational interaction, HL acquisition, and interlanguage pragmatics research. The underperformance of the HL learners in learning honorifics observed in this study can be explicated in terms of the unique characteristics of HL acquisition, such as incomplete acquisition, childhood language overhearing, simplification, and overgeneralization, as well as their distinctive language learning process, including meaningfocused, top-down, and intuition-based processing. On the contrary, the outstanding performance of the NHL learners is attributable to their form-focused, bottom-up, and analytical processing. Regarding the superiority of explicit feedback, it is speculated that the feedback is more likely to involve the deeper level of awareness, and negative evidence is more likely to be noticed by learners than implicit feedback. The superiority of explicit instruction found in 154 instructed SLA and interlanguage pragmatics research consolidates the findings in this study. The best performance of the NHL learners in the explicit feedback condition can be accountable for a combination of a superioirity of explicit feedback and the NHL learners. However, there was no statistical difference between the two feedback types for the NHL learners. It is assumed that the NHL learners’ explicit language learning process overrode explicitness of feedback. Also, the NHL learners’ lack of fluency and frequent pauses during conversation may have provided them with additional time to process learning a language irrespective of the feedback conditions. Thus, this may have helped make both feedback types equally effective for these learners. The HL learners’ inferior performance in the implicit feedback condition may be also associated with their meaning-focused orientation, distinct properties of Korean honorifics, and ambiguous nature of recasts. Since Korean honorifics are non-salient and have a low communicative value, it is construed that the HL learners who were inclined to be meaningfocused may have been less likely to notice the corrective intention of implicit feedback. Consequently, this may have made it harder for them to abondon their old habit, namely, minimized honorific use at home. Thus, as the findings of this study suggest, explicit feedback is more effective than implicit feedback for the HL learners at least in the learning of Korean honorifics. This dissertation makes an important step toward advancing the current scope of interaction research and SLA. Further research is called for in order to generalize the findings of this study, answer our remaining inquiries about factors affecting the interaction and learning relationship and finally enhance our knowledge and understanding in SLA. Limitations and Suggestions for Future Research 155 Although this dissertation makes an important contribution to expanding the current knowledge of SLA, some of the limitations of this study should be addressed. This is the first attempt to investigate the impact of different interactional feedback on the L2 pragmatic development of learners with different HL background. Therefore, this study must be seen as exploratory, not as confirmatory, and caution must be used to generalize the findings of the study. The discussions provided to explain differential performances of each group of learners are only speculative since learners’ perception was not investigated in the present study. Schmidt (2001) claimed, “The clearest evidence that something has exceeded the subjective threshold and been consciously perceived or noticed is concurrent verbal report” (p. 20). As such, to consolidate the findings of this study and confirm the possible explanations provided to each research question, the utilization of learners’ verbal reports, such as stimulated recall (see Gass & Mackey, 2000, for details), is suggested. The use of introspective methods would measure learners’ noticing of feedback as well as elicit their introspective comments. This would demonstrate if different levels of awareness actually occur in response to different feedback types and thus provide more solid information in interpreting the results. This study was conducted in the controlled laboratory context. Much effort was made to make the context as natural as possible in that the researcher’s real family photos were used, and feedback was not limited to honorific errors. However, since the participants in this study were aware that this investigation was part of the research project, it is possible that the participants may have acted differently from what they would normally do in authentic conversations. Pragmatics, including Korean honorifics, is context-dependent; thus, it is an empirical question if the utilization of authentic setting of the conversation may bring out different results. Similarly, it is uncertain if the results of the current study can be applied to the classroom context. Future 156 studies should be conducted in different contexts in order to see the generalizability of the results of this study. The use of honorifics may differ if a person being respected is present during conversation. In other words, if the honorific person is present, the speaker is more likely to use honorifics. Because it was not feasible to have honorific people in the photos actually participate in this study, some learners might have reduced the use of honorifics. However, as the honorific people in the photos were the researcher’s own older family members or teacher, it was expected for the learners to use honorifics regardless of their presence. Based on the results of the current study, it is unknown whether the HL learners’ underperformance of the honorific development was ascribed to language attrition, erosion, loss or incomplete acquisition. It is speculated that incomplete acquisition may have attributed to their poor performance because honorifics are a late-acquired feature. However, this needs empirical validation. This study was based on learners’ aural/oral abilities to perform communicative tasks and measured their implicit knowledge dimension (Bowles, 2011; R. Ellis et al., 2006). Because the purpose of the study was to see how learners of different HL background make use of honorifics in a context that simulates real life, measuring learners’ explicit knowledge of honorifics was not possible. Different modality (spoken vs. written) may result in different results, and HL learners’ performance may differ according to task types (Montrul, 2011b). Therefore, the honorific development using written tasks or written feedback may be interesting to explore in future research. Finally, it is important to note that only learners’ HL background, explicitness of interactional feedback, and L2 pragmatics as a linguistic focus, were examined as potential 157 mediating factors that may influence the interaction and learning relationship in this study; therefore, the discussions were centered on these factors under investigation. However, as interaction-driven learning is subject to multiple factors, and they may work independently or together to contribute to interaction and learning, it is possible that other relevant variables, addressed in Chapter 2, may have also played a role in this relationship. Future research should explore other potential variables to better understand the interaction and learning mechanism. 158 APPENDICES 159 Appendix A: Language Background Questionnaire 1. Name: _________________________________ 2. E-mail address: ________________________________ 3. Phone number: ________________________________ 4. Gender: □ Female □ Male 5. Age: _______ 6. Affiliated university: ______________________________ 7. Year in college: _____________________________________ □ Freshman □ Sophomore □ Junior □ Senior □ MA □ PhD □ Other__________ 8. Major (if any): _______________________________ Minor (if any): _______________________________ 9. Indicate the course title(s)/course level(s) (example: beginner for HLs, intermediate for NHLs) of the Korean class(es) you are enrolled (or were the most recently enrolled) in the current Korean program and the year/semester offered. Course title(s)/course level(s): _____________________________________ Year/semester offered: ___________________________________ 10. Have you ever studied any other foreign/second language(s) besides Korean? □ Yes □ No If yes, What language(s): ________________________________ At what age: ________________ How long: __________ year(s) _________ month(s) 11. Where were you born? □ US □ Korea 160 □ Other _____________ 12. If you were born in Korea, how old were you when you moved to the US? ______year(s) _____month(s) 13. What is your first language(s)? (e.g., the language(s) your parents spoke to you before your age of five.) □ English □ Korean □ Both 14. What language(s) do you consider to be your dominant language now? □ English □ Korean □ Both 15. Are your parents native speakers of Korean? □ Yes □ No If yes, Who: □ Only mother □ Only father □ Both 16. With whom and how often did you interact in Korean while you were growing up? (check all that apply.)  Mother: □Never □Seldom □Occasionally □Often □Usually □Always  Father: □Never □Seldom □Occasionally □Often □Usually □Always  Sibling(s): □Never □Seldom □Occasionally □Often □Usually □Always  Grandparent(s): □Never □Seldom □Occasionally □Often □Usually □Always  Other relatives: □Never □Seldom □Occasionally □Often □Usually □Always  Friends: □Seldom □Occasionally □Often □Usually □Always  Others (specify): ________________ □Often □Usually □Always □Never □Never □Seldom □Occasionally 161 17. With whom and how often do you currently interact in Korean? (check all that apply.)  Mother: □Never □Seldom □Occasionally □Often □Usually □Always  Father: □Never □Seldom □Occasionally □Often □Usually □Always  Sibling(s): □Never □Seldom □Occasionally □Often □Usually □Always  Grandparent(s): □Never □Seldom □Occasionally □Often □Usually □Always  Other relatives: □Never □Seldom □Occasionally □Often □Usually □Always  Spouse: □Never □Seldom □Occasionally □Often □Usually □Always  Friends: □Never □Seldom □Occasionally □Often □Usually □Always  Others (specify): ________________ □Often □Usually □Always □Never □Seldom □Occasionally 18. Have you ever studied Korean at other non-university institution in the US? (e.g., Saturday school, K-12, etc.) □ Yes □ No If yes, What institution(s): ___________________________________________________ At what age: ________________ How long: ____________ year(s) _________ month(s) Purpose: ___________________________________________________________ 19. Have you ever studied abroad in Korea? □ Yes □ No If yes, What institution(s): ___________________________________________________ At what age: ________________ 162 How long: ____________ year(s) _________ month(s) Purpose: ___________________________________________________________ 20. Have you ever visited or lived in Korea? □ Yes □ No If yes, At what age: ________________ How long: ____________ year(s) _________ month(s) Purpose: ___________________________________________________________ 21. Check your other uses/exposure of Korean (e.g., other than family, friends, taking a Korean course, study abroad, etc.). (Check all that apply.)  I watch Korean TV/movies: □ Never  □ Usually □ Always □ Seldom □ Occasionally □ Often □ Usually □ Always □ Seldom □ Often □ Usually □ Always □ Often □ Usually □ Always □ Often □ Usually □ Always □ Occasionally I write and/or email in Korean: □ Never  □ Often I read books and/or newspapers in Korean: □ Never  □ Occasionally I listen to Korean music: □ Never  □ Seldom □ Seldom □ Occasionally Other: ________________________: □ Never □ Seldom □ Occasionally 22. I am comfortable in the following language skills: 163  Speaking: □ Strongly disagree □ Disagree □ Agree □ Strongly agree  Listening: □ Strongly disagree □ Disagree □ Agree □ Strongly agree  Reading: □ Strongly disagree □ Disagree □ Agree □ Strongly agree  Writing: □ Strongly disagree □ Disagree □ Agree □ Strongly agree  Grammar: □ Strongly disagree □ Disagree □ Agree □ Strongly agree  Vocabulary: □ Strongly disagree □ Disagree □ Agree □ Strongly agree  This is the end of the survey. Thank you very much for your time!  164 Appendix B: Table B-1. Participant Background Information HL status Group Gender Age Korean language course Age of arrival in level assignment Place of birth the US (if born in Korea) 1 HL HL/Implicit F 20 Completed 1st year US 2 HL HL/Implicit F 20 Enrolled in 2nd year US 3 HL HL/Implicit F 21 Enrolled in 2nd year US 4 HL HL/Implicit F 19 Enrolled in 2nd year Korea 5 HL HL/Implicit M 20 Enrolled in 2nd year US 6 HL HL/Implicit F 18 Enrolled in 2nd year US 7 HL HL/Implicit F 20 Enrolled in 2nd year Korea 8 HL HL/Implicit M 18 Enrolled in 2nd year US 9 HL HL/Implicit M 20 Enrolled in 2nd year US 10 HL HL/Implicit M 18 Enrolled in 2nd year Korea 11 HL HL/Implicit F 21 Enrolled in 2nd year US 165 6 months 1.5 months 1 month Table B-1 (cont’d) 12 HL HL/Implicit F 20 Enrolled in 2nd year US 13 HL HL/Implicit F 20 Enrolled in 2nd year US 14 HL HL/Implicit M 21 Enrolled in 2nd year US 15 HL HL/Implicit F 18 Never been enrolled US 16 HL HL/Implicit M 19 Never been enrolled US 17 HL HL/Explicit M 18 Completed 1st year US 18 HL HL/Explicit F 20 Enrolled in 2nd year US 19 HL HL/Explicit M 20 Enrolled in 2nd year US 20 HL HL/Explicit F 22 Enrolled in 2nd year US 21 HL HL/Explicit M 18 Enrolled in 2nd year US 22 HL HL/Explicit M 23 Enrolled in 2nd year US 23 HL HL/Explicit F 20 Enrolled in 2nd year US 24 HL HL/Explicit F 21 Enrolled in 2nd year US 25 HL HL/Explicit M 21 Enrolled in 2nd year US 26 HL HL/Explicit M 19 Enrolled in 2nd year Korea 166 3 months Table B-1 (cont’d) 27 HL HL/Explicit M 19 Enrolled in 2nd year US 28 HL HL/Explicit M 19 Enrolled in 2nd year US 29 HL HL/Explicit M 19 Never been enrolled US 30 HL HL/Explicit M 22 Completed 2nd year US 31 HL HL/Explicit F 19 Never been enrolled US 32 HL Control F 19 Completed 1st year US 33 HL Control M 21 Completed 2nd year US 34 HL Control M 20 Enrolled in 2nd year US 35 HL Control M 20 Enrolled in 2nd year US 36 HL Control F 20 Enrolled in 2nd year Korea 37 HL Control F 19 Enrolled in 2nd year US 38 HL Control F 21 Completed 2nd year US 39 NHL NHL/Implicit M 27 Enrolled in 2nd year US 40 NHL NHL/Implicit F 21 Completed 1st year US 41 NHL NHL/Implicit F 18 Enrolled in 2nd year Other 167 8 months Table B-1 (cont’d) 42 NHL NHL/Implicit F 21 Enrolled in 2nd year Other 43 NHL NHL/Implicit F 19 Enrolled in 2nd year US 44 NHL NHL/Implicit F 20 Enrolled in 2nd year US 45 NHL NHL/Implicit F 20 Enrolled in 2nd year US 46 NHL NHL/Implicit F 19 Enrolled in 2nd year US 47 NHL NHL/Implicit M 20 Enrolled in 2nd year US 48 NHL NHL/Implicit M 26 Enrolled in 2nd year Other 49 NHL NHL/Implicit F 22 Enrolled in 2nd year US 50 NHL NHL/Implicit F 20 Enrolled in 2nd year US 51 NHL NHL/Implicit F 25 Enrolled in 2nd year Other 52 NHL NHL/Implicit F 18 Enrolled in 2nd year Other 53 NHL NHL/Implicit F 19 Enrolled in 2nd year US 54 NHL NHL/Implicit F 20 Enrolled in 2nd year US 55 NHL NHL/Explicit F 19 Enrolled in 2nd year US 56 NHL NHL/Explicit F 22 Enrolled in 2nd year US 168 Table B-1 (cont’d) 57 NHL NHL/Explicit F 25 Enrolled in 2nd year Other 58 NHL NHL/Explicit F 19 Enrolled in 2nd year Other 59 NHL NHL/Explicit F 20 Enrolled in 2nd year US 60 NHL NHL/Explicit F 18 Enrolled in 2nd year US 61 NHL NHL/Explicit F 20 Enrolled in 2nd year US 62 NHL NHL/Explicit F 21 Enrolled in 2nd year Other 63 NHL NHL/Explicit F 21 Enrolled in 2nd year Other 64 NHL NHL/Explicit F 20 Enrolled in 2nd year US 65 NHL NHL/Explicit F 20 Enrolled in 2nd year US 66 NHL NHL/Explicit F 19 Enrolled in 2nd year Other 67 NHL NHL/Explicit F 19 Enrolled in 2nd year US 68 NHL NHL/Explicit F 19 Enrolled in 2nd year US 69 NHL NHL/Explicit F 21 Enrolled in 2nd year Other 70 NHL NHL/Explicit F 20 Enrolled in 2nd year US 71 NHL Control F 21 Completed 2nd year US 169 Table B-1 (cont’d) 72 NHL Control F 18 Enrolled in 2nd year US 73 NHL Control F 25 Enrolled in 2nd year US 74 NHL Control F 20 Enrolled in 2nd year Korea 75 NHL Control M 29 Enrolled in 2nd year US 76 NHL Control F 19 Enrolled in 2nd year US 77 NHL Control F 21 Enrolled in 2nd year US 78 NHL Control M 20 Enrolled in 2nd year US 170 12 months Appendix C: Table C-1. Participants’ Self-rated Comfort with Korean Language Skills (i.e., I am comfortable in the following Korean language skills.) HL learners NHL learners n 1 2.6% 6 15.0% 11 28.9% 28 70.0% 22 57.9% 6 15.0% 4 10.5% 0 .0% Total 38 100.0% 40 100.0% Strongly disagree 1 2.6% 6 15.0% Disagree 0 .0% 16 40.0% Agree 28 73.7% 17 42.5% Strongly agree 9 23.7% 1 2.5% Total 38 100.0% 40 100.0% Strongly disagree 4 10.5% 2 5.0% Disagree 17 44.7% 10 25.0% Agree 16 42.1% 24 60.0% Strongly agree 1 2.6% 4 10.0% Total Writing Strongly disagree Strongly agree Reading Percentage Agree Listening n Disagree Speaking Percentage 38 100.0% 40 100.0% Strongly disagree 4 10.5% 3 7.5% Disagree 23 60.5% 11 27.5% Agree 10 26.3% 25 62.5% Strongly agree 1 2.6% 1 2.5% 171 Table C-1 (cont’d) Total 40 100.0% Strongly disagree 6 15.8% 4 10.0% 23 60.5% 18 45.0% Agree 8 21.1% 17 42.5% Strongly agree 1 2.6% 1 2.5% Total Vocabulary 100.0% Disagree Grammar 38 38 100.0% 40 100.0% Strongly disagree 3 7.9% 3 7.5% Disagree 28 73.7% 20 50.0% Agree 6 15.8% 17 42.5% Strongly agree 1 2.6% 0 .0% Total 38 100.0% 40 100.0% 172 Appendix D: Table D-1. Heritage Language Learners' Childhood and Current Interaction in Korean Childhood interaction Current interaction n Source Frequency n Mother Never 0 .0% 2 5.3% Seldom 1 2.6% 0 .0% Ocassionally 3 7.9% 7 18.4% Often 6 15.8% 5 13.2% Usually 11 28.9% 11 28.9% Always 17 44.7% 13 34.2% Total 38 100.0% 38 100.0% Never 0 .0% 2 5.3% Seldom 5 13.2% 5 13.2% Ocassionally 6 15.8% 6 15.8% Often 5 13.2% 5 13.2% Usually 10 26.3% 8 21.1% Always 12 31.6% 11 28.9% Missing 0 .0% 1 2.6% Total 38 100.0% 38 100.0% Never 16 42.1% 18 47.4% Seldom 10 26.3% 9 23.7% Ocassionally 3 7.9% 7 18.4% Often 4 10.5% 1 2.6% Father Sibling(s) Percentage 173 Percentage Table D-1 (cont’d) Usually 3 7.9% 3 7.9% Always 2 5.3% 0 .0% 38 100.0% 38 100.0% 1 2.6% 4 10.5% Seldom 5 13.2% 8 21.1% Ocassionally 8 21.1% 5 13.2% Often 0 .0% 1 2.6% Usually 7 18.4% 3 7.9% Always 17 44.7% 17 44.7% Total 38 100.0% 38 100.0% Never 1 2.6% 2 5.3% 10 26.3% 11 28.9% Ocassionally 8 21.1% 9 23.7% Often 5 13.2% 3 7.9% Usually 6 15.8% 4 10.5% Always 8 21.1% 7 18.4% Total 38 100.0% 38 100.0% Never 14 36.8% 9 23.7% Seldom 13 34.2% 17 44.7% Ocassionally 5 13.2% 6 15.8% Often 1 2.6% 3 7.9% Usually 1 2.6% 1 2.6% Total Grandparent(s) Never Other relatives Seldom Friends 174 Table D-1 (cont’d) Always 10.5% 2 5.3% Total Others 4 38 100.0% 38 100.0% Never 6 15.8% 8 21.1% 10 26.3% 6 15.8% Ocassionally 1 2.6% 2 5.3% Often 2 5.3% 0 .0% Usually 0 .0% 0 .0% Always 0 .0% 1 2.6% Missing 19 50.0% 21 55.3% Total 19 50.0% 38 100.0% Seldom 175 Appendix E: Table E-1. Participants’ Uses and Exposure of Korean (other than family, friends, taking a Korean course, study abroad, etc.) HL learners NHL learners Source Frequency n TV/movies Never 1 2.6% 1 2.5% Seldom 7 18.4% 4 10.0% 13 34.2% 13 32.5% Often 8 21.1% 14 35.0% Usually 7 18.4% 5 12.5% Always 2 5.3% 3 7.5% Total 38 100.0% 40 100.0% Never 4 10.5% 3 7.5% Seldom 9 23.7% 2 5.0% Occasionally 9 23.7% 8 20.0% Often 4 10.5% 10 25.0% Usually 9 23.7% 6 15.0% Always 3 7.9% 11 27.5% 38 100.0% 40 100.0% 25 65.8% 28 70.0% 10 26.3% 9 22.5% Occasionally 2 5.3% 3 7.5% Often 1 2.6% 0 .0% Usually 0 .0% 0 .0% Occasionally Music Total Books/newspapers Never Seldom Percentage 176 n Percentage Table E-1 (cont’d) Always 0 .0% 38 100.0% 40 100.0% Never 14 36.8% 8 20.0% Seldom 14 36.8% 14 35.0% Occasionally 8 21.1% 14 35.0% Often 1 2.6% 3 7.5% Usually 1 2.6% 1 2.5% Always 0 .0% 0 .0% Total Other .0% Total Writing/email 0 38 100.0% 40 100.0% Never 2 5.3% 4 10.0% Seldom 0 .0% 0 .0% Occasionally 0 .0% 2 5.0% Often 1 2.6% 0 .0% Usually 1 2.6% 1 2.5% Always 0 .0% 0 .0% Missing 34 89.5% 33 82.5% Total 38 100.0% 40 100.0% 177 Appendix F: Figure F-1. Sample Testing Material. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation. 178 Appendix G: Figure G-1. Sample Treatment Material 179 Appendix H: Task Instructions for Photo Description Task Directions Hello, my name is Seonogme Ahn. I am from Korea. Most of my family members and friends live in Korea, so I miss them very much. Recently, they sent me photos of their daily lives. Now I would like to share them with you. I will show you a series of photos. You will see various people at different locations. When you see only one person in the photos, please describe what he or she is doing. When you see two people interacting with each other in the photos, be sure to describe clearly who is doing what action TO whom; for example, “the mother is giving a book TO the niece.” I hope you will enjoy seeing my family in Korea. 180 Appendix I: Task Instructions for Story Sequencing Task Directions Here, you will see photos of my family in Korea again. You will see my family doing various things at different locations. You will be given six photos, and I will have six different photos. We will take turns to describe one photo at a time to each other, alternating through all 12 photos. Then, we will work together to lay out the 12 photos in a story sequence. At the end, you will retell the story we just worked together, but feel free to add any further information you would like. Please remember that when you see a series of photos, tell me what each person is doing in each photo. When you see only one person in the photos, please describe what he or she is doing. When you see two people interacting with each other in the photos, be sure to describe clearly who is doing what action TO whom; for example, “the mother is giving a book TO the niece.” I hope you will enjoy seeing my family in Korea. 181 Appendix J: Sample Treatment Sessions Treatment 1 (New Year’s Day/ One-way Task) NS: 제니는 설날 알아요? 설날은 New Year’s Day예요. Do you happen to know Selnal? Selnam is New Year’s Day. NHL: 네. Yes. NS: 그래서 우리 집에 친척들이 모여요. 설날에 사진을 찍었어요. 여기가 우리 집이예요. 이 사람들은 우리 식구들이예요. 우리 고모, 수희…수희는 오빠의 딸이예요. 그리고 재영이… 재영이는 오빠의 아들이예요. 이 사람은 아이들의 엄마예요. 올케 언니… 오빠의 부인이예요. 우리 식구들이 많아요. 저의 어머니와 수희가 있어요. 이 옷이 뭐예요? So, my relatives get together in our house. We took pictures on New Year’s Day. This is our house. These people are my family members. Our aunt, Soohee… Soohee is my older brother’s daughter. And, Jaeyoung… Jaeyoung is my older brother’s son. This person is the mother of the children. My sister-in-law… (is) my older brother’s wife. We have a lot of family members. (Here) are my mother and Soohee. What is this clothing? NHL: 한복이예요. Hanpok. NS: 한복이지요? 자, 뭐하는 사진이예요? (Is it) Hanpok (Korean traditional costume)? Well, what is this photo about? NHL: 음… 수희는 할머니께…. 음… 이게 뭐예요? 182 Soohee-S grandmother-IO(H) Umm… Soohee… to the grandmother… umm… what is this? NS: 떡이예요. (This is) a rice cake. 할머니께 NHL: 어, 수희는 Soohee-S grandmother-IO(H) 줘요. OBJ 떡을 rice cake-DO give-V Ah, Soohee gives the rice cake to the grandmother. 할머니께 NS: 예. 수희가 Soohee-S 떡을 grandmother-IO(H) rice cake-DO 드려요? recast give-V(Hum) Yeah. Does Soohee give the rice cake to the grandmother? NHL: 네. Yes. NS: 그지요? 이거요. 떡 좋아해요? Is that so? This one. Do you like a rice cake? NHL:… … NS: 제니, 떡 좋아해요? Jenny, do you like a rice cake? NHL: 떡, 떡을 안, 안먹어봤어요. I haven’t had the rice cake. NS: 안먹어봤어요? Haven’t you had one? 183 NHL: 네. No. NS: 아, 그래요? (웃음) 다음엔 먹어봤으면 좋겠어요. 여기가 우리 집이예요. Oh, really? (laughs) I hope (you) will try it next time. Here is our house. NHL: 네. Yeah. NS: 그 다음에는요 할아버지와 수희가 있네요. 무슨 사진이예요? Next, there are the grandfather and Soohee. What is this photo about? NHL: 음… 수희는 Soohee-S 할아버지께 새로운 옷을 grandfather-IO(H) new clothing-DO 보여줘요. OBJ show-V Umm… Soohee shows new clothing to the grandfather. NS: 아, 수희가 Soohee-S 할아버지께 새로운 옷을 grandfather-IO(H) new clothing-DO 보여드려요? recast show-V(Hum) Ah, does Soohee show new clothing? NHL: 네. Yes. NS: 한복 저고리는요 웃도리에요. 이거는 한복조끼예요. The jacket of Hanpok is for (you to wear in) your upper body. This is a vest of Hanpok. NHL: 네. Yes. NS: 그렇지요? 그 다음에 부엌이 있어요. Is it? Next, there is a kitchen. 184 NHL: 네. Yes. NS: 무슨 사진이예요? What is this photo about? NHL: … 할머니는… grandmother-S 밥을 rice-DO 준비해요. SUB prepare-V The grandmother prepares rice. NS: 네. 할머니께서는 grandmother-S(H) 밥을 rice-DO 준비하세요? recast prepare-V(H) Yes, the grandmother prepares rice. NHL: 네. Yes. NS: 밥 좋아해요? (웃음) Do you like rice? (laughs) NHL: 네. (웃음) Yes. (laughs) NS: 그 다음에는요 수희가 있어요. 뭐해요? Next, there is Soohee. What does she do? NHL: 수희는… 애기하고 사진을 같이 줘 줘. Soohee… gives… a photo with the baby. NS: 찍어요? Take (a photo)? 185 NHL: 찍어요. Take (a photo). NS: 이 애기는 저의 딸이예요. 여자 애예요. 그 다음에 여기는요 뭐예요? This baby is my daughter. (She is) a girl. Next, what is this here? NHL: 다른 음식이예요. (This is) different food. NS: 네. 음식이예요. 그 다음에는 저의 어머니하고 수희.. 무슨 사진이예요? Yes, (this is) food. Next, my mother and Soohee… What is this photo about? NHL: 할머니는 수희한테 grandmother-S Soohee-IO 책을 book-DO 읽어세요. SUB read-V(H) The grandmother reads a book to Soohee. NS: 할머니께서는 수희한테 grandmother-S(H) 책을 Soohee-IO book-DO The grandmother reads a book to Soohee. NHL: 읽어주세요. read-V(H) Read. NS: 그 다음에는 무슨 사진이예요? Next, what is this photo about? NHL: 수희는 Soohee-S 할아버지한테 음… 우유? grandfather-IO Soohee is… to the grandfather umm… milk? 186 읽어주세요. recast read-V(H) NS: 네, 우유 같아요. Yes, (it) looks like milk. 줘요. OBJ NHL: 우유, 우유를 milk-DO give-V (Soohee) gives milk. 할아버지께 우유를 grandfather-IO(H) milk-DO NS: 수희가 Soohee-S Soohee gives milk to the grandfather. NHL: 드려요? give-V(Hum) Give? NS: 네. 자, 우유 좋아해요? Yes, well, do you like milk? NHL: 아니요. (웃음) No. (laughs) NS: 저도 안좋아해요. (웃음) 그 다음에는요? I don’t like it, either. (laughs) What is next? NHL: … … NS: 무슨 사진이예요? What is this photo about? NHL: 음… dancing? 187 드려요. recast give-V(Hum) Umm… dancing? NS: 네? Excuse me? NHL: Dancing? (웃음) Dancing (laughs) NS: 춤추다. To dance. NHL: 춤추다. To dance. NS: 네. Yes. NHL: 춤추다… 춤추 춤추는 사진이예요. To dance… This photo is about dancing. NS: 누가요? Who? NHL: 두 여자는 춤을 추는… Two ladies dance… NS: 제니 목소리를 크게 해요. Jenny, would you please speak up? NHL: 어, 네. (웃음) Oh, yes. (laughs) NS: 제가 못들으면 안되니까요. 자, 알겠어요? 188 It wouldn’t be good if I couldn’t hear you. Well, do you understand? NHL: 네. Yes. NS: 자, 여기는 무슨 사진이예요? Well, what is this photo about? NHL: 할머니는 grandmother-S 사과를 준비해요, 준비하세요. SUB apple-DO prepare-V(H) The grandmother prepares an apple, prepares. NS: 네, 할머니께서는 grandmother-S(H) 사과를 apple-DO 준비하세요. 그렇지요? recast prepare-V(H) Yes, the grandmother prepares an apple, right? NHL: 네. Yes. NS: 사과를 척척척… 이렇게 해서. 제니는 사과를 이렇게 해서 먹어요? (Peel, peel, peel) the apple… just like this. Would you eat an apple like this, Jenny? NHL: 아니요. No. NS: 그냥 먹어요? (Do you) eat it just the way it is? NHL: 네. Yes. NS: 네. 미국 사람들은 그냥 먹어요. 네. 189 Yes, Americans eat an apple just the way it is (without peeling), yes. NHL: 네. Yes. NS: 이건 무슨 사진이예요? What is this photo about? 할머니께 책을 grandmother-IO(H) book-DO NHL: 음… 수희는 Soohee-S 읽, 읽어주세, 읽어줘, 줘요. OBJ read-V Umm… Soohee reads a book to the grandmother. 할머니께 책을 grandmother-IO(H) book-DO NS: 수희가 Soohee-S Does Soohee read a book to the grandmother? NHL: 읽어드러요. read-V(Hum) Read. NS: 네. “Top cat”이예요. Yes. “(It is) a top cat.” NHL: 어… Uhhh… NS: (사진에 있는 책을 가리키며) 이거 귀여워요? Is this cute? (pointing out a photo in the book) NHL: 네. (웃음) Yes. (laughs) 190 읽어드려요? recast read-V(Hum) NS: 이것은 아이들이 한복입을 때 하는 hairpin이예요. 귀엽지요? 그 다음에는 저의 친척들이 많이 모였어요. 재영이, 재영이는 수희의 오빠예요. 고모부, 작은 아버지, 형부, 친척 오빠, 고모, 조카, 그 다음 친척 언니, 올케 언니, 작은 아버지, 작은 어머니, 우리 언니예요. 그리고 저는 여기 있어요. 사진예요. 저는 결혼할 때 한복 입고 사진 찍었어요. 여기 이렇게 있어요. 우리 집이예요. 그 다음에는요 무슨 사진이예요? This is a hairpin to use when children wear Hanpok. Is this cute? Next, a lot of my relatives get together. Jaeyoung, Jaeyoung is Soohee’s older brother. My uncle (my father’s brotherin-law), another uncle (my father’s younger brother), my brother-in-law, my male cousin, my aunt (my father’s sister), my nephew, the next person (is) my female cousin, my sisterin-law, my uncle (my father’s younger brother), my aunt (my father’s sister-in-law), (and) my sister. And, I am here. In the photo. I took a photo wearing Hanpok when I got married. Here (I am). This is our house. What is this photo about, next? NHL: 음… Umm… NS: 여기요. (웃음) Here. (laughs) NHL: 수희는 돈을 받아요. Soohee receives money. 할아버지께서는 수희한테 돈을, 돈을 grandfather-S(H) Soohee-IO money-DO The grandfather gives money to Soohee. NS: 네, 설날에 받는 돈은 세뱃돈이예요. 191 주, 주세요, 주세요.  SUB give-V(H) Yes, the money we receive on New Year’s Day is a handsel. NHL: 세뱃돈. A handsel. NS: 이렇게 절 하면 돈을 받아요. 그것이 세뱃돈이예요. 세뱃돈 받고 싶어요? If you bow like this (demonstrating how to bow), you receive money. That is the handsel. Would you like to receive the handsel? NHL: 네. (웃음) Yes. (laughs) NS: 저도 세뱃돈 받고 싶어요. 그 다음에 수희는요? I would like to receive the handsel. What about Soohee, next? NHL: 수희는 Soohee-S 할아버지께… 어… 이게 뭐예요? grandfather-IO(H) Soohee… to the grandfather… uhhh… what is this? NS: 떡국이요. Ttekkwuk. NHL: 떡국? Ttekkwuk? NS: 떡국은 rice cake soup이예요. 설날에 먹어요. Ttekkwuk is a rice cake soup. We eat it on New Year’s Day. NHL: 수희는 Soohee-S 할아버지한테 떡국을 grandfather-IO rice cake-DO Soohee gives ttekkwuk to the grandfather. 192 줘, 줘요.  OBJ give-V 할아버지께 떡국을 grandfather-IO(H) rice cake-DO NS: 수희는 Soohee-S 드려요.  recast give-V(Hum) Soohee gives ttekkwuk the grandfather. NHL: 드려요, 드려요. give-V(Hum) Give, give. NS: 떡국 안먹어봤어요? Haven’t you had ttekkwuk? NHL: 안먹어봤어요. No, I haven’t. NS: 네, 맛있어요. 이것은 뭐예요? 문어예요, 문어. Yeah, it is delicious. What is this? An octopus, octopus. NHL: 문어? An octopus? NS: 이게 일본 말로는 타코래요. 타코 맞아요? It is a “tako” in Japanese. Is a “tako” right? NHL: 어, 네, 네. Uh, yes, yes. NS: 그 다음에는 무슨 사진이예요? Next, what is this photo about? NHL: 할아버지 사진이예요. (It is) the grandfather’s photo. 193 NS: 네, 어떻게 하는 사진이예요? Yes, what is this photo about? 행복하게 웃, 웃어, 웃어요.  SUB happily-ADV smile-V NHL: 어… 할아버지는 grandfather-S Uhh… the grandfather smiles happily. NS: 할아버지께서는 grandfather-S(H) 웃으세요? recast 행복하게 happily-ADV smile-V(H) Does the grandfather smile happily? NHL: 네. 웃어세요. smile-V(H) Yes, (he) smiles. NS: 네, 웃는 얼굴이예요. 아, 이 음식들은 우리가 만들었어요. (음식 사진을 가리키며) 자… 여기에 문어 있어요? Yes, (it is) a smiley face. Ah, we made these (sorts of) foods (pointing out the photos of food). Well, is an octopus here? NHL: 네, 문어. Yes, an octopus. NS: 자, 그리고 떡도 있어요? Well, (is there) a rice cake as well? NHL: 네. Yes. NS: 그리고 술도 있어요. 194 And, there is an alcohol. NHL: 네. (웃음) Yes. (laughs) NS: 그 다음에는요, 자, 무슨 사진이예요? Next, well, what is this photo about? NHL: 음… 수희는 드러요.  OBJ 할, 할머니한테 grandmother-IO Soohee-S 책을 book-DO give-V(Hum) Umm… Soohee gives a book to the grandmother. 할머니께 책을 grandmother-IO(H) book-DO NS: 수희가 Soohee-S Soohee gives a book to the grandmother. NHL: 드려요. give-V(Hum) Gives. NS: 네, 이게 영어책이예요, 영어책이예요. Yes, this is an English book, English book. NHL: 영어책. An English book. NS: 한국 사람들은 영어 공부 많이 해요. Koreans study English a lot. NHL: 아, 네. Oh, I see. 195 드려요. recast give-V(Hum) NS: 그래도 영어를 잘못해요. 저도요. (웃음) However, (they are) not good at English. I am not good at it, either. (laughs) NHL: (웃음) (laughs) NS: 자, 무슨 사진이예요? Well, what is this photo about? NHL: 음… Umm… NS: 애기가 뭐해요? What does the baby do? NHL: 먹어요. (웃음) Eat. NS: 애기가 팥죽 먹어요. The baby eats a red bean soup. NHL: 팥죽. A red bean soup. NS: 네, 그래서 기분이 좋아요. 그렇지요? Yes, so (she is) in a good mood. Is that right? NHL: 네. Yes. NS: 그 다음에요? 이 아이가 또 우리 딸이예요. 여기도 있어요. 똑같은 애기예요. 그 다음에는요? 196 (What’s) next? This child is our daughter again. Here she is again. The same baby. What’s next? NHL: 음… 수희는 할머니… 앞에서… Soohee-S grandmother-IO in front of-LOC Umm… Soohee… in front of the grandmother… NS: 절하다. (demonstrating how to bow.) bow-V To bow. NHL: 절해요? OBJ bow-V Bows? 할머니께 NS: 수희는 Soohee-S grandmother-IO(H) 절드려요. recast bow-V(Hum) Soohee bows to the grandmother. 할머니께 NHL: 수, 수희는 Soohee-S grandmother-IO(H) 절… bow-V Soohee bows to the grandmother… I’ve forgot. NS: 절드려요. recast bow-V(Hum) Bows. NHL: 절, 절드려요. bow-V(Hum) Bows, bows. 197 잊었어요. OBJ NS: 네, 그렇지요? Yes, is that right? NHL: 절, 절. Bow, bow. NS: 그러면 세뱃돈 받아요. 그 다음에는요? Then, (she) receives the handsel. What’s next? NHL: 할아버지는 grandfather-S Soohee-IO 뽀뽀해요. SUB 수희한테 give a kiss-V The grandfather gives a kiss to Soohee. NS: 네, 할아버지께서. grandfather-S(H) Yes, the grandfather. NHL: 께서. S(H) (The grandfather). NS: 수희한테 Soohee-IO 뽀뽀하세요. 그렇지요? recast give a kiss-V(H) (The grandfather) gives a kiss to Soohee. Is that right? NHL: 네. Yes. NS: 자, 이것은요? Well, what about this? 198 NHL: 음… 개는 물을 먹어요. Umm… The dog drinks water. NS: 얘는 토니예요, 토니. (He is) Tony, Tony. NHL: 토니? Tony? NS: 열다섯살이예요. 우리집 강아지예요. 그런데 5월달에 죽었어요. (He is) fifteen years old. (He is) our family dog. However, (he) died in May. NHL: 아… Ah… NS: 그래서 너무 슬퍼요. So, (I am) very sad. NHL: 네. Yes. NS: 너무 너무 똑똑해요. 그 다음에는 사람들이 뭐해요? (He is) really, really smart. Next, what do the people do? NHL: 절 드러요. 절 드러요. Bow, bow. NS: 네, 맞아요. 자, 이것은요 뭐하는 거예요? Yes, that is right. Well, what is this about? NHL: 게임을 해, 해요. Playing a game. 199 NS: 네. 무슨 게임이예요? Yes, what kind of game? NHL: 몰라요. (웃음) I don’t know. (laughs) NS: 이건 윷놀이예요. This is yuchnoli. NHL: 윷놀이. Yuchnoli. NS: 윷놀이는요 설날에 하는 게임이예요. 이렇게 던져서 하는 거예요. 재미있겠지요? Yuchnoli is the game we play on New Year’s Day. (We) do (it) by throwing (it) like this. Would it be fun? NHL: 네. Yes. NS: 안해봤어요? Haven’t you played it before? NHL: 아니요. No. NS: 한국어 수업에서 안해봤어요? Haven’t you played it in your Korean class? NHL: 아니요. No. NS: 유명해요. 유명한 게임이예요. 200 Well-known. (It is) a well-known game. NHL: 네. Yes. NS: 여기에는 할머니 할아버지 이름을 적어서 놓는 곳이예요. (사진을 가리키며) Here is the place where we write down (our) grandmother’s and grandfather’s names (pointing out the photo) NHL: 수희는 Soohee-S 할머니한테… grandmother-IO 절… 이렇게 하는게… 이거, 이거… bow-V Soohee bows to the grandmother… like this… this, this… NS: 아, 이것은 절하다 (demonstrating how to bow). 이것은 인사하다예요 (demonstrating how to greet). 달라요. Ah, this is to bow. This is to greet. (They are) different. NHL: 네. 인사, 인사, 절, 절. 수희는 Soohee-S 할머니한테 인사해요. OBJ grandmother-IO bow-V Yes. Greet, greet, bow, bow. Soohee greets the grandmother. NS: 수희는 Soohee-S 할머니께 grandmother-IO(H) 인사드려요. 그렇지요? recast bow-V(Hum) Soohee greets the grandmother. Is that right? NHL: 할머니께, 네. (웃음) grandmother-IO(H) To the grandmother, yes. (laughs) NS: 용돈 주세요, 이렇게요. (웃음) 그 다음에는요 뭐예요? 201 Give me pocket money, like this. (laughs) What’s next? NHL: 음… Umm… NS: 수희 뭐해요? What does Soohee do? NHL: 음… Umm… NS: 웃어요? 웃어요? (Does she) smile? NHL: … 네, 웃어요. Yes, (she) smiles. NS: 네, 웃어요. 수희는 웃어요. 그리고 누워 있어요. Yes, (she) smiles. Soohee smiles and is lying down. NHL: 누워 있어? (Is she) lying down? NS: 눕다. To die down. NHL: 누워 있어요. (She is) lying down. NS: 뽀로로 알아요? Do you know Pororo? NHL: 아니요. 202 No. NS: 이거 한국에서 유명한 캐릭터예요. 뽀로로예요. 여기 뽀로로 써 있잖아요. This is a famous (cartoon) character. (This is) Pororo. Here it says Pororo. NHL: 네. (웃음) Yes. (laughs) NS: 네, 그 다음에 이것은 무슨 사진이예요? Yes, next, what is this photo about? NHL: 할머니께서는 grandmother-S(H) 수희한테 포로로를 Soohee-IO Pororo-DO The grandmother gives Pororo to Soohee. NS: 네, 맞아요. 수희가 기분이 좋아요? Yes, that is right. Is Soohee in a good mood? NHL: 아, 네. Ah, yes. NS: 웃어요? (Does she) smile? NHL: 네. 크게 웃어요. Yes. (She has) a big smile. NS: 수희는 눈이 작아요. (웃음) She has small eyes. (laughs) NHL: (웃음) (Laughs) 203 주세요. SUB give-V(H) NS: 그 다음에는요 사람들이 뭐해요? What do the people do, next? NHL: 음… 절..을… 드러요? Umm… bow? NS: 네, 맞아요. 제니는 이런 것 본적 있어요? 봤어요? Yes, that is right. Have you seen something like this before, Jenny? Have you? NHL: 아니요. No. NS: 못봤어요? Haven’t you? NHL: 못봤어요. No. NS: 우리 집에서는요 일년에 세번, 네번, 다섯 번 이런거 해요. 여러 번해요. We perform something like this three times, four times, five times per year in our house. (We do it) several times. NHL: 네. Yeah. NS: 여기는 제 방이예요. Here is my room. NHL: 네. Yeah. NS: 냉장고가 있어요. (웃음) 자, 이것은 무슨 사진이예요? 204 (There is) a refrigerator. (laughs) Well, what is this photo about? NHL: 할아버지께서는 써, 써세요, 써세요. SUB 컴퓨터를 grandfather-S(H) computer-DO use-V(H) The grandfather uses a computer, uses. NS: 네, 맞아요. 자, 그 다음에는요? Yes, that is right. Well, what’s next? NHL: 수희는 테레비전을 봐요. Soohee watches TV. NS: 네, 재밌겠어요? 재밌어요? 재미있을까요? Yes, (would it be) fun? (Is it) fun? (Is it going to be) fun? NHL: 네, 재밌어요. Yes, (that is) fun. Treatment 1 (Fun Day/ One-way Task) NS: 저는 서울 사람이예요. 그런데 저희 집은… 롯데월드 알아요? I am from Seoul. By the way, our house is… do you know Lotte World? NHL: 네. (웃음) Yes. (laughs) NS: 어떻게 알아요? How do (you) know (that)? NHL: 드라마, 뮤지컬… 205 Dramas, musicals... NS: 롯데월드에서 저희 집은 가까워요. 그래서 걸어서 가요. 저희 가족들은 저희 가족은 롯데월드에 가요. 그리고 롯데마트에서 시장봐요. 장보러 가요. 그리고 옆에 공원도 있어요. 이게 공원이구요 그리고 석촌호수도 있어요. 우리 가족들은 자주 가요. 자, 여기는 공원이예요. 이것은 뭐예요? Our house is close to Lotte World. So, (we) walk (to get) there. My families, my family go to Lotte World. And, (we) buy groceries at Lotte Mart. (We go) grocery shopping. And, there is a park next to (Lotte World). This is the park, and (there is) the Seokchon Lake as well. My family (members) often go there. Well, here is the park. What is this? NHL: 몰라요. (웃음) I don’t know. (laughs) NS: 이것은요 무덤이예요. This is a tomb. NHL: 무덤? A tomb? NS: 왕의 무덤이예요. 그래서 아주 커요, 아주 아주 커요. A king’s tomb. So, (it is) very big, very, very big. NHL: 네. Yeah. NS: 여기 공원이예요. 사람들이 앉아 있어요. 여기서부터는 제니가 해요. Here is the park. People are sitting. From here, Jenny can go ahead. NHL: 아, 있어요. 206 Ah, there it is. NS: 네. Yes. NHL: 롯데월드이예요. (This is) Lotte World. NS: 네. Yes. NHL: 음… 아버지는 숙희한테 father-S 돈을 Sookhee-IO money-DO 주세요. SUB give-V(H) Um… The father gives money to Sookhee. NS: 할아버지께서는 수희한테.... recast grandfather-S(H) Soohee-IO The grandfather… to Soohee… NHL: 할아버지께서는 수희한테 돈을 주세요. grandfather-S(H) Soohee-IO money-DO give-V(H) The grandfather gives money to Soohee. NS: 네, 그 다음은요? Yeah, what’s next? NHL: 수희는 Soohee-S 할아버지께 표를, 표를 드려요, 드려요. OBJ grandfather-IO(H) ticket-DO give-V(Hum) Soohee gives, gives a ticket, a ticket to the grandfather. NS: 네, 그 다음에는요 공원이예요. 매표소 알아요? 207 Yes, the next one is the park. Do you know a ticket office? NHL: 매표소? A ticket office? NS: 표 파는 곳이예요. 그 다음에는요? A place where tickets are sold. What’s next? NHL: 공원에서 할아버지께서는 park-LOC grandfather-S(H) 음… The grandfather… umm… in the park. NS: 소리지르다. To shout. NHL: 소리, 소리지르세요. SUB shout-V(H) Shout, shout. NS: 네. “수희야” 이렇게요. 그 다음에는요? 저희 어머니 (pointing out her mother). (웃음) Yes. “Soohee” like this. What’s next? My mother (pointing out her mother). (laughs) NHL: 음… Ummm… NS: 수희는 뭐해요? What does Soohee do? NHL: 수희는 Soohee-S 할아버지께… 이게 뭐예요? grandfather-IO(H) Soohee… to the grandfather… what’s this? 208 NS: 핸드백, 핸드백이예요. A purse, a purse. 할아버지께 핸드백을 grandfather-IO(H) purse-DO NHL: 수희는 Soohee-S 드려요. OBJ give-V(Hum) Soohee give a purse to the grandfather. NS: 네. 그 다음에 공원이예요. 무슨 사진이예요? Yes. Next, (this is) the park. What is this photo about? 할머니께 NHL: 수희는 Soohee-S 드려요. OBJ 장갑을 grandmother-IO(H) glove-DO give-V(Hum) Soohee gives gloves to the grandmother. NS: 그렇죠? 추워요. 겨울이예요. 그 다음에, 어, 공원이예요. 무슨 사진이예요? Is that so? (It is) cold. (It is) a winter. Next, ah, this is the park. What is this photo about? NHL: 할아버지께서는 grandfather-S(H) 신문을 찢, 찢… newspaper-DO tear-V The grandfather tears, tears newspapers. NS: 찢다. To tear. NHL: 신문을 찢, 찢, 찢, 찢으세요. SUB newspaper-DO tear-V(S) Tear, tear, tear, tear newspapers. NS: 이 사람 누구예요? 알아요? Who is this person? Do you know (her)? 209 NHL: 몰라요. I don’t know. NS: 김연아 알아요? Do you know Yen A Kim? NHL: 어, 네. Oh, yes. NS: 네. 유명해요. 피겨 스케이팅 김연아예요. Yes. (She is) famous. Figure skating, Yen A Kim. NHL: 네. Yes. NS: 김연아 싫어하나봐요. (He) seems to dislike Yen A Kim. NHL: 아니요. (웃음) No. (laughs) NS: 그 다음에 무슨 사진이예요? Next, what is this photo about? NHL: 수희는 롯데월드에서 사진을 찍어요. Soohee takes a photo at Lotte World. NS: 네. 이것은요? 풍선이예요. Yes. What is this? A balloon. NHL: 풍선. A balloon. 210 NS: 롯데월드 마스코트이예요. 그 다음에는 수희는 뭐해요? (This is) a mascot of Lotte World. Next, what does Soohee do? 할아버지께 책을 grandfather-IO(H) book-DO NHL: 수희는 Soohee-S 읽어줘요. OBJ read-V Soohee reads a book to the grandfather. 할아버지께 책을 grandfather-IO(H) book-DO NS: 수희는 Soohee-S 읽어드려요.  recast read-V(Hum) Soohee reads a book to the grandfather. NHL: 읽어드려요. read-V(Hum) Reads (a book) to. NS: 이건 옛날 이야기예요. 그 다음 이건 무슨 사진이예요? (This is) an old-time story. Next, what is this photo about? NHL: 음… 한국말로 ball? Umm… a ball in Korean? NS: 공. A ball. NHL: 할아버지께서는 grandfather-S(H) 수희한테 공을.. Soohee-IO ball-DO The grandfather… a ball to Soohee. NS: 던지다. To throw. 211 NHL: 던져, 던져, 던져세요. SUB throw-V(H) Throw, throw, throw. NS: 할아버지께서 grandfather-S(H) 수희한테 공을 Soohee-IO ball-DO 던지세요. recast throw-V(H) The grandfather throws a ball to Soohee. NHL: 네. Yeah. NS: 그 다음에는요? What about next? NHL: 롯데월드이예요. (This is) Lotte World. NS: 여기에 아이스링크가 있어요. 롤러코스터도 있어요. 제니는 이런것 타는것 좋아해요? Here is an ice rink, and (here is) a roller coaster. Do you like riding this sort of things, Jenny? NHL: 타는것? 네, 좋아해요. Riding? Yes, I do. NS: 안무서워요? Isn’t that scary? NHL: 아니요. No. NS: 안무서워요? 저는 무서워요. 자, 이것은요? Isn’t that scary? (It’s) scary for me. Well, what about this? 212 NHL: 음… 수희는 ride를 타요? Um... does Soohee ride a ride (name of attraction)? NS: tea cup을 타요? Ride a tea cup? NHL: tea cup을 타요? Ride a tea cup? NS: 저 이것 이름 몰라요. 제가 그냥 이름을 지었어요. I don’t know its name. I just made it up. NHL: 네. Yeah. NS: 그 다음에 수희는 뭐해요? Next, what does Soohee do? 할머니께 책을 grandmother-IO(H) book-DO NHL: 수희는 Soohee-S 읽, 읽어, 읽어 드, 읽어드려요. OBJ read-V(Hum) Soohee reads, reads, reads, reads a book to the grandmother. NS: 네, 이것 스타벅스 커피 있어요. Yes, this is Starbucks coffee. NHL: 네. (웃음) Yeah. (laughs) NS: 스타벅스 한국에서 비싸요. 미국보다 비싸요. 그 다음에는요 무슨 사진? Starbucks is expensive in Korea. (It is) more expensive than the United States. Next, what photo (is this)? 213 NHL: 할아버지께서는 공을… grandfather-S(H) ball-DO The grandfather… a ball… NS: 차다. To kick. NHL: 차? 차세요. 공을 Kick-V ball-DO 차세요. SUB kick-V(H) Kicks? Kicks. Kicks a ball. NS: 네, 수희 공이예요. 귀엽죠? Yeah, (that is) Soohee’s ball. (Is that) cute? NHL: 네. Yes. NS: 그 다음에요? What about next? NHL: 수희는 Soohee-S 할머니한테… grandmother-IO 그림을 보, 보여, 보여드려요. OBJ picture-DO show-V(Hum) Soohee shows, shows, shows a photo to the grandmother. NS: 수희는 Soohee-S 할머니께 grandmother-IO(H) 사진을 photo-DO 보여드려요. recast show-V(Hum) Soohee shows a photo to the grandmother. NHL: 수희는 Soohee-S 할머니께 사진을 보여드려요. grandmother-IO(H) photo-DO show-V(Hum) 214 Soohee shows a photo to the grandmother. NS: 네, 여기는요? Yes, what about here? NHL: 음… 어, 숙희와 할머니와 할아버지는… 수희는 엄마, 할머니와 할아버지하고 사진을 찍어요, 찍어요. Um… Ah, Sookhee, the grandmother, and the grandfather… Soohee takes, takes a photo with the mother, the grandmother and the grandfather. NS: 네, 저희 올케 언니예요, 오빠의 부인. Yes, (she is) my sister-in-law, my older brother’s wife. NHL: 네. Yeah. NS: 그 다음에는요? What’s next? NHL: 음… 할아버지께서는 grandfather-S(H) 수희한테… 핸드백을 주어요, 주, 주세요. SUB Soohee-IO purse-DO Um… the grandfather gives, gives, gives a purse to Soohee. NS: 네. 이거 핸드백 예뻐요? (웃음) Yes. Is this purse cute? (laughs) NHL: 네. (웃음) Yes. (laughs) NS: 이것 갖고 싶어요? Do you want to have this? 215 give-V(H) NHL: 음… Um… NS: 제니, 이것 갖고 싶어요? (웃음) Jenny, do you want to have this? (laughs) NHL: 아니요. (웃음) No. (laughs) NS: 애기꺼 같아요? Does (it) look like for babies’? NHL: 네. (웃음) Yes. (laughs) NS: 네, 맞아요. (웃음) 그 다음에는요… 이거는요? Yes, that’s right. (laughs) Next, what about this? NHL: 음… 수희는 Soohee-S 인사해요. OBJ 할머니한테 grandmother-IO greet-V Um… Soohee greets the grandmother. NS: 네, 수희가 Soohee-S 할머니께 grandmother-IO(H) 인사드려요. recast greet-V(Hum) Yes, Soohee greets the grandmother. NHL: 인사드려요. greet-V(Hum) Greets. NS: 네. 216 Yeah. NHL: 할머니께 인사드려요. grandmother-IO(H) greet-V(Hum) Greets the grandmother. NS: 네. 그 다음에는요? Yes. What’s next? NHL: 할머니께서는 운동을 하세요. SUB grandmother-S(H) exerse-V(H) The grandmother does exercise. NS: 네, 맞아요. 훌라후프 운동을… (웃음) 할줄 알아요? 제니, 할줄 알아요? 못해요? Yes, that is right. Hula hoop exercise… (laughs) (Do you) know how to do this? Jenny, can you do this? Can’t you? NHL: 못해요. (웃음) (I) can’t. (laughs) NS: 그 다음에는요? 이거는 가짜 가방이예요. 이거는 쌀과자, 뻥튀기, 강냉이예요. 먹어봤어요? Next? This is a fake purse. These are rice snacks, fries, and corn snacks. Have you tried these? NHL: 아니요. No. NS: 다이어트에 좋아요. 칼로리가 적어요, 낮아요. (They are) good to lose weight. (They are) low in calories, low. 217 NHL: 네. (웃음) Yeah. (laughs) NS: 그 다음에는요 수희 뭐해요? Next, what does Soohee do? NHL: 수희는 carousel? Soohee… carousel? NS: 네. 회전목마. Yes, carousel. NHL: 수희는 회전목마를 타요. Soohee rides a carousel. NS: 네, 맞아요. 신났어요. 제니는 타봤어요? Yes, that is right. (She is) excited. Have you ridden a carousel, Jenny? NHL: 네, 많이 타봤어요. Yes, (I) have ridden (it) a lot. NS: 지금도 타요? (Do you) still ride it? NHL: 네. (웃음) Yes. (laughs) NS: (웃음) 그 다음은요? What about next? NHL: 수희가 Soohee-S 할아버지께 아이스크림을 grandfather-IO(H) ice cream-DO 218 드려요, 드려요. OBJ give-V(Hum) Soohee gives, gives an ice cream to the grandfather. NS: 네, 한국 사람들은 소프트 아이스크림이라고 해요. Yes, Koreans call this a soft ice cream. NHL: 소프트 아이스크림? A soft ice cream? NS: 여기는요? What about here? NHL: 공원이예요. (This is) a park. NS: 네. 뭐해요? Yes, what does (this person) do? NHL: 여자는 개하고 걸어요. The woman walks a dog. NS: 네, 맞아요. Yes, that is right. NHL: 그리고… 남자는 전화해요. And, the man talks on the phone. NS: 네, 맞아요. 바쁜가봐요. 그 다음에는요? 이건 공원이예요. 밤이예요, 밤. 무슨 사진이예요? Yes, that is right. (They) seem busy. What about next? This is the park. (It’s) a night, night. What is this photo about? NHL: 할머니께서는 뽀뽀 드, 드… SUB 수희한테 219 grandmother-S(H) Soohee-IO give a kiss-V(Hum) The grandmother gives, gives a kiss to Soohee. NS: 할머니께서는 grandmother-S(H) 수희한테 Soohee-IO 뽀뽀하세요. recast give a kiss-V(H) The grandmother gives a kiss to Soohee. NHL: 하세요, 하세요. give (a kiss)-V(H) Gives, gives. NS: 수희는 기분이 좋아요? Is Soohee in a good mood? NHL: 음음... (웃음) 아니요. Umm, umm… (laughs) no. NS: 기분이 안좋아요? 왜 그래요? Isn’t (she) in a good mood? Why not? NHL: 음… Umm… NS: 수희가 뭐해요? What does Soohee do? NHL: 수희는 핸드백을 사고 싶어요. Soohee wants to buy a purse. NS: 네, 아까는 핸드백을 샀어요. 그런데 조그만 핸드백을 샀어요. 그런데 지금 큰 핸드백을 찾았어요. 그래서 기분이 안좋아요. 220 Yes, (she) bought a purse before. However, (she) bought a small purse. However, (she) found a big purse. So, (she is) not in a good mood. Treatment 2 (Birthday Day/ Two-way Task) NS: 수희가 칼을 들어요. Soohee holds a knife. NHL: 할머니께서는 주세요. SUB 수희에게 Soohee-IO grandmother-S(H) 손을 hand-DO give-V(H) The grandmother gives a hand to Soohee. NS: 수희 생일 케잌이예요. (This is) Soohee’s birthday cake. NHL: 수희는 생일 케잌을, 수희는 할머니께 생일 케잌을 드, 드러요. OBJ Soohee-S grandmother-IO(H) birthday cake-DO give-V(Hum) Soohee… the birthday cake, Soohee gives the birthday cake to the grandmother. NS: 수희가 생일 케잌을 찾았어요. Soohee found the birthday cake. NHL: 수희는 Soohee-S 할, 할머니 사진을 grandmother-IO picture-DO 찍어요. OBJ take-V Soohee takes a photo of the grandmother. NS: 수희가 Soohee-S 할머니께 사진을 grandmother-IO(H) picture-DO Soohee takes a photo of the grandmother. 221 찍어드려요. recast take-V(Hum) NHL: 찍어드려요. take-V(Hum) Takes (a photo of). NS: 수희가 접시를 들어요. Soohee holds a plate. NHL: 할머니께서는 grandmother-S(H) 생일 케잌을… birthday cake-DO The grandmother… the birthday cake… NS: 자르다. To cut. NHL: 자르, 자르, 자르세요. SUB cut-V(H) Cuts, cuts, cuts. NS: 수희가 케잌을 먹고, 맛있게 먹어요. Soohee eats, eats the delicious cake. NHL: 할머니께서는 grandmother-S(H) 생일 케이크를 아… (웃음) birthday cake-DO The grandmother… the birthday cake… ah… (laughs) NS: (웃음) 놓다? 접시? (Laughs) To put? A plate? NHL: 접시, 접시에 놓, 놓. A plate, puts… on the plate. 222 NS: 놓다. To put. NHL: 놓, 놓, 놓으세요. SUB put-V(H) Puts, puts, puts. NS: 수희가 웃어요. Soohee smiles. NHL: 수희가 할머니께 Soohee-S grandmother-IO(H) 생일 케잌을 드, 드셔, 드려요. OBJ birthday cake-DO Soohee gi, give, gives the birthday cake to the grandmother. NS: 먹여주다. To feed. NHL: 먹, 먹, 먹여드려요? feed-V(Hum) Feeds, feeds, feeds? NS: Do you want to lay out the twelve photos? NHL: In order? NS: 준비됐어요? Are you ready? NHL: 네. 수희는… Yes, Soohee… NS: 칼? 223 give-V(Hum) A knife? NHL: 칼. A knife. NS: 칼. A knife. NHL: Oh, it looks like a television remote. (웃음) NS: 네. Yeah. NHL: Like over here. (웃음) NS: 아, 아니요. 칼이예요. (웃음) No, no, (it is) a knife. (laughs) NHL: 아… Ahh… NS: 소리 크게 해요. 괜찮아요. 소리 크게 해도 되요. Could you please speak up? That is ok. You can speak loudly. NHL: 네. 음… 수희는 선물을 찾, 찾아요. 그래서 할머니께서는 grandmother-S(H) 수희에게 선물을 Soohee-IO gift-DO 주세요. SUB give-V(H) Yes. Umm… Soohee looks, looks for a gift. So, the grandmother gives the gift to Soohee. NS: 네, 저는 여기 있어요, 여기요. 가족 사진이예요. Yes, here I am, here. (This is my) family photo. NHL: 얘는 생일 케잌을 이예요. 224 This is the birthday cake. 수희는 할머니께… Soohee… the grandmother… 할머니께 수희가 Soohee-S grandmother-IO(H) 사진을 찍, 찍어요, 찍어드려요.  OBJ photo-DO take-V(Hum) Soohee ta, take, takes a photo of the grandmother. 할머니께 수희는 Soohee-S grandmother-IO(H) 드러요.  OBJ 케잌을 cake-DO give-V(Hum) Soohee gives the cake to the grandmother. 음… 수희는 칼… 칼. Umm… Soohee a knife… knife. NS: 들다. To hold. NHL: 들어요. (웃음) 할머니께서는 grandmother-S(H) 생일 케잌을… birthday cake-DO To hold. (laughs) The grandmother… the birthday cake… NS: 자르다. To cut. NHL: 자, 자르세요. 자르세요? SUB cut-V(H) Cuts, cuts. Cuts? NHL: 수희는… 잡시. 225 Soohee… a plate NS: 접시. A plate. NHL: 접시를 들, 들어요. 그리고 할머니께서는 케이크를… grandmother-S(H) cake-DO (Soohee) holds a plate. And, the grandmother… a cake… NS: 접시에 놓다. To put (it) on the plate. 놓, 놓, 놓으세요.  SUB NHL: 접시에 plate-LOC put-V(H) Puts (it) on the plate. 수희가 케이크를 먹어요. Soohee eats the cake. NS: 네, 맛있게 먹어요. Yes, (she) eats the delicious cake. NHL: 네. 음… 수희는 웃어요, (웃음) 웃어요. Yes. Um… Soohee smiles, (laughs) smiles. NS: 네, 맞아요. 기분이 좋은가봐요. Yes, that is right. (She) seems to be in a good mood. NHL: 네. 음… 수희가 Soohee-S 할머니께 케이크를 grandmother-IO(H) cake-DO Yes. Um… Soohee feeds the grandmother the cake. 226 먹어드려요. OBJ feed-V(Hum) 할머니께 케이크를 먹여드려요? recast grandmother-IO(H) cake-DO feed-V(Hum) NS: 수희가 Soohee-S Does Soohee feed the grandmother the cake? NHL: 먹여, 먹여드려요. feed-V(Hum) Feeds, feeds. Treatment 2 (Department Store/ Two-way task) NS: 수희 생일 파티 후에 수희 생일 선물을 사러 백화점에 갔어요. 할아버지하고 백화점에 갔어요. 롯데 백화점이예요. 롯데 백화점 알아요? After Soohee’s birthday party, (Soohee) went to a department store to buy a birthday gift. (She) went to the department store with the grandfather. (This is) the Lotte department store. Do you know the Lotte department store? NHL: 네. Yes. NS: 롯데 백화점 앞문이예요. (This is) a front door of the Lotte department store. NHL: 음… 할아버지께서는 grandfather-S(H) 롯데 백화점에 Lotte department store-LOC Umm… the grandfather comes to the Lotte department store. NS: 수희도 롯데 백화점에 와요. 227 오세요. SUB come-V(H) Soohee comes to the Lotte department store as well. NHL: 할아버지께서는 롯데 백화점에서… grandfather-S(H) Lotte department store-LOC 앉고 있으세요. SUB sit-V(H) The grandfather is sitting at the Lotte department store. NS: 할아버지께서는 grandfather-S(H) 롯데 백화점에서… Lotte department store-LOC 앉으세요? recast sit-V(H) The grandfather sits at the Lotte department store? NHL: 앉으세요. sit-V(H) Sits. NS: 네. 자, 수희가 손을 흔들어요. 인사하나봐요. 인사하는 것 같아요. Yes. Well, Soohee waives a (her) hand. (She) seems to greet. (She) seems to greet. 할아버지께 NHL: 음… 수희가 Soohee-S 인사를 해요. OBJ grandfather-IO(H) greet-V Umm… Soohee greets the grandfather. 할아버지께 NS: 수희가 Soohee-S grandfather-IO(H) 인사를 드려요. recast greet-V(Hum) Soohee greets the grandfather. NHL: 인사가, 인사를 드려요. greet-V(Hum) Greets, greets. NS: 자, 아이들 옷이 있어요. 아이들 옷이 있어요. 228 Well, (there is) children’s clothing. (There is) children’s clothing. NHL: 아이들, 아, 아이들. (웃음) Children, chi, children. (laughs) NS: (웃음) (laughs) NHL: 수희가… 할아버지께서는 수희한테 돈을 주세요. SUB grandfather-S(H) Soohee-IO money-DO give-V(H) Soohee… The grandfather gives money to Soohee. NS: 네. 수희가 거울을 봐요. Yes. Soohee looks at (herself) in the mirror. 할아버지께 드레스를 grandfather-IO(H) dress-DO NHL: 음… 수희가 Soohee-S 사고 싶어, 싶, 싶다고 해요. OBJ tell-V Umm… Soohee tells the grandfather that she wants, wants, wants to buy a dress. NS: 드레스를 보여주다. To show a dress. NHL: 보여, 보여주, 보여줘요. Shows, shows, shows. 할아버지께 드레스를 보여드려요. recast grandfather-IO(H) dress-DO show-V(Hum) NS: 수희가 Soohee-S Soohee shows a dress to the grandfather. NHL: 네. 수희가 Soohee-S 할아버지께 드레스를 보여드려요. grandfather-IO(H) dress-DO show-V(Hum) 229 Yes. Soohee shows a dress to the grandfather. NS: 수희가 돈을 내요. Soohee pays money. NHL: 아… 수희가 Soohee-S Ah… Soohee… 할아버지께 … 음… 가방. grandfather-IO(H) bag-DO to the grandfather… umm… a bag. NS: 쇼핑백? A shopping bag? 드러요. OBJ NHL: 쇼핑백을 shopping bag-DO give-V(Hum) (Soohee) gives a shopping bag. NS: 자, lay out해볼까요? 자, 할까요? Okay, do you want to lay out (the photos)? Well, shall we? NHL: 아, 롯데백화점이예요. Ah, (this is) the Lotte department store. 할아버지께서는 롯데백화점에 grandfather-S(H) Lotte department store-LOC 오세요. SUB come-V(H) The grandfather comes to the Lotte department store. 그리고 수희가 롯데 백화점에도 와요. (웃음) 음… 수희가 인사해요. And, Soohee comes to the Lotte department store as well. (laughs) Umm… Soohee greets. 할아버지께서는 앉으세요. SUB grandfather-S(H) sit-V(H) 230 The grandfather sits. 할아버지께 음… 수희가 Soohee-S 인사를 드, 드러요. OBJ grandfather-IO(H) greet-V(Hum) Umm… Soohee greets, greets the grandfather. 아이들, 옷. Children, clothing. NS: 옷. Clothing. NHL: 옷, 옷이예요. (This is) clothing, clothing. NS: 네. Yup. NHL: 수희가… 거울, 거울을 봐요. Soohee… looks at (herself) in the mirror, mirror. 수희가 Soohee-S 할아버지께 grandfather-IO(H) 드레스를 dress-DO 보여드려요. OBJ show-V(Hum) Soohee shows the dress to the grandfather. 할아버지께서는 수희에게 grandfather-S(H) Soohee-IO 돈을 money-DO The grandfather gives money to Soohee. 수희가 돈을 내, 내요. Soohee pays money. 231 주세요. SUB give-V(H) 수희가 Soohee-S 할아버지께 쇼핑백을 grandfather-IO(H) shopping bag-DO 드려요. OBJ give-V(Hum) Soohee gives a shopping bag to the grandfather. Note. Italicized parts indicate the linguistic targets (i.e., honorific or humble form) in this study. 232 Appendix K: Table K-1. Average Time for Each Experimental Group in Each Task Group Pretest Treatment1 Treatment2 Photo Story Photo Story description sequencing description Immediate Delayed posttest posttest sequencing HL/Implicit (n = 16) 13 min. 31 min. 22 min. 35 min. 17 min. 15 min. 16 min. HL/Explicit (n = 15) 13 min. 33 min. 21 min. 31 min. 17 min. 15 min. 16 min. NHL/Implicit (n = 16) 15 min. 35 min. 19 min. 33 min. 21 min. 15 min. 15 min. NHL/Explicit (n = 16) 15 min. 39 min. 28 min. 35 min. 19 min. 14 min. 15 min. 233 Appendix L: Coding and Scoring Guidelines Subject Referent Honorifics Honorific subject case marker -kkeyse(nun) and honorific verbal morpheme -si 1. Identify the obligatory contexts that involve subject honorific use and underline the sentences with an honorific subject (someone who is older or higher in social status than the speaker). Example. Halmeni-kkeyse meli-lul piseyo. Grandmother-S(H) hair-DO brush-V The grandmother brushes her hair. Example. Sentences with an omitted honorific subject (Halmeni-ka) (Grandmother-S) twupwu-lul yolihayyo. tofu-DO cook-V (The grandmother) cooks tofu. 2. If the sentence involves subject honorifics, code the whole sentence as SUB. Example. Sentences that include a subject honorifics Sensayngnim-kkeyse moca-lul Teacher-S(H) hat-DO ssu-sey-yo. SUB wear-V(H). The teacher wears a hat. 3. Write S below the subject and V below the verb. Example. Sentences with an honorific subject Halmeni-ka S kewul-ul pwayo. SUB V 234 Grandmother-S mirror-DO look-V The grandmother looks at the mirror. Example. Halapeci-kkeyse tampay-lul phi-sey-yo. S Grandfather-S(H) V cigarette-DO smoke-V(H) The grandfather smokes a cigarette. 4. If the learner produces a complex sentence with an overt and omitted honorific subject, code each verb separately. Example. Sensayngnim-i ancase S V Teacher-S sit-V chayk-ul ilku-sey-yo. V book-DO read-V(H) The teacher sits and reads a book. 5. If the learner uses auxiliary + main verb construction in the predicate, code the whole verb as one V. Example. Halmeni-kkeyse hwacho-ey mwul-ul cwu-si-ko isseyo. V Grandmother-S(H) plant-LOC water-DO The grandmother is watering the plant. Example. Halapeci-kkeyse chengsohako kyeyseyyo. 235 give-V(H)-PROG V Grandfather-S(H) clean-V(SH)-PROG The grandfather is cleaning. Example. Halapeci-kkeyse wuntongha-si-ko kyeyseyyo. V Grandfather-S(H) exercise-V(H)(SH)-PROG The grandfather is doing an exercise. 6. If the subject includes an honorific subject and a non-honorific subject, disregard the sentence. Example. Halapeci-hako hanna-hako chayk-ul ilkeyo. Grandfather-S Hanna-S book-DO read-V The grandfather and Hanna read a book. 7. If the learner self-corrects or self-repeats the target forms, underline only the final forms. Example. Sensayngnim-kkeyse chayk-ul ilk-sey-yo, ilku-sey-yo. V Teacher-S(H) book-DO read-V(H) read-V(H) The teacher reads, reads a book. 8. If the honorific subject case marker -kkeyse(nun) or the honorific verbal morpheme -si is missing or inaccurately used, 0 is awarded. (1) 0 point is awarded for the missing honorific subject case marker -kkeyse(nun). Example. 236 Sensayngnim-i khothu-lul coat-DO Teacher-S ipu-sey-yo. wear-V(H) The teacher wears a coat. Example. Halapeci Grandfather-S tampay phyeyo. tobacco-DO smoke-V The grandfather smokes a cigarette. (2) 0 point is awarded for the missing honorific verbal morpheme -si. Example. Halapeci-kkeyse Grandfather-S(H) hanna-eykey Hanna-IO ppoppohayyo. kiss-V The grandfather kisses Hanna. 9. 1 point is awarded for the use of the honorific subject case marker -kkeyse(nun) or the honorific verbal morpheme -si. (1) 1 point is awarded for the use of the honorific subject case marker -kkeyse(nun). Example. Sensayngnim-kkeyse chayk-ul ilku-sey-yo. Teacher-S(H) book-DO read-V(H) The teacher reads a book. Example. Halmeni-kkeysenun hanna-eykey pap-ul meki-sey-yo. Grandmother-S(H) Hanna-IO rice-DO feed-V(H) The grandmother feeds Hanna rice. 237 (2) 1 point is awarded for the use of the honorific verbal morpheme –si and/or the special polite verb kyeyseyyo. Example. Honorific verbal morpheme –si in the single verb constructions in the predicate Halapeci-kkeyse chengso-lul ha-sey-yo. Grandfather-S(H) cleaning-DO do-V(H) The grandfather cleans. Example. Honorific verbal morpheme –si, but without the special polite verb kyeyseyyo in the auxiliary + main verb constructions in the predicate Halmeni-kkeyse hwacho-ey mwul-ul cwu-si-ko isseyo. V Grandmother-S(H) plant-LOC water-DO give-V(H) The grandmother is watering the plant. Example. Honorific verbal morpheme –si and the special polite verb kyeyseyyo in the auxiliary + main verb constructions in the predicate Halapeci-kkeyse wuntongha-si-ko kyeyseyyo. V Grandfather-S(H) exercise-V(H) (SH) The grandfather is doing an exercise. Example. Special polite verb kyeyseyyo, but without the honorific verbal morpheme –si in the auxiliary + main verb constructions in the predicate Halapeci-kkeyse chengsohako kyeyseyyo. V Grandfather-S(H) clean-V(SH) 238 The grandfather is cleaning. (3) No penalty is given for the verbal inflection error. Example. Halmeni-kkeyse meli-lul pis-sey-yo. Grandmother-S(H) hair-DO brush-V(H) The grandmother brushes her hair. (4) No penalty is given for the inaccurate pronunciation the learner produced. Example. Sensayngnim-kkeysennun chayk-ul ilku-sey-yo. Teacher-S(H) book-DO read-V(H) The teacher reads a book. (5) No penalty is given for the inaccurate use of verb tense. Example. Sensayngnim-kkeysenun Teacher-S(H) cayyengi-hanthey kong-ul cha-syess-eyo. Cayyeng-IO ball-DO kick-V(H)-PAST The teacher kicked a ball to Cayyeng. (6) No penalty is given for the wrong choice of vocabulary as long as the honorific subject case marker -kkeyse(nun) or the honorific verbal morpheme -si is accurately used. Example. Halmeni-kkeysenun swuhyu-hanthey ayki ches sayngil sacin-ul pokoka-sey-yo. Grandmother-S(H) Swuhyu-IO baby first birthday picture-DO see and go-V(H) The grandmother shows Swuhyu a baby’s first birthday picture. 239 Object Referent Honorifics Honorific indirect object case marker -kkey and humble verbal form -(e)tulita 1. Identify the obligatory contexts that involve object honorific use and underline the sentences with an honorific object (someone who is older or higher in social status than the speaker). Example. Hanna-ka halapeci-kkey sakwa chayk-ul Hanna-S grandfather-IO(H) apple book-DO tulyeyo. give-V(Hum) Hanna gives an apple book to the grandfather. 2. If the sentence involves object honorifics, code the whole sentence as OBJ. Example. Sentences that include an object honorifics Swuhyu-ka Swuhyu-S halapeci-kkey oppauy tol sacin-ul poye-tulyeyo. OBJ grandfather-IO(H) brother’s first birthday picture-DO show-V(Hum) Swuhyu shows her brother’s first birthday picture to the grandfather. 3. Write O below the object and V below the verb. Example. Sentences with an honorific object Hanna-ka halmeni-kkey insahayyo. OBJ O Hanna-S V grandmother-IO(H) greet-V Hanna greets the grandmother. Example. Hyengcwungi-ka halmeni-hanthey chayk O Hyengcwung-S grandmother-IO ilke-tulyeyo. OBJ V book-DO 240 read-V(Hum) Hyengcwung reads a book to the grandmother. 4. If the learner produces a complex sentence with an honorific object, code each verb separately. Example. Cayyengi-ka sensayngnim-kkey chayk-ul cwumyense insahayyo. V V O Cayyeng-S teacher-IO(H) book-DO give-V greet-V Cayyeng gives a book to the teacher and greets her. 5. If the sentence indicates that the object is involved in the activity with the subject, disregard the sentence. Example. Cwunghaksayng-i Middle school student-S sensayngnim-hako chayk-ul ilkeyo. teacher-ADV Book-DO read-V The middle school student reads a book with the teacher. 6. If the learner self-corrects or self-repeats the target forms, underline only the final forms. Example. Cwunghaksayng-i sensayngnim-hanthey senmwul-ul cweyo, tulyeyo. V Middle school student-S teacher-IO gift-DO give-V(Hum) The middle school student gives a gift to the teacher. 7. If the honorific indirect object case marker -kkey or humble verbal form -(e)tulita is missing or inaccurately used, 0 is awarded. (1) 0 point is awarded for the missing honorific subject case marker -kkey Example. 241 Hanna-ka halmeni-hanthey ppololo chinkwu eyti inhyeng-ul tulyeyo. Hanna-S grandmother-IO Ppololo’s friend Eyti doll-DO give-V(Hum) Hanna gives Ppololo’s friend, Eyti, doll to the grandmother. Example. Swuhyu-nun Swuhyu-S halapeci moca cweyo. grandfather-IO hat-DO give-V Swuhyu gives a hat to the grandfather. (2) 0 point is awarded for the missing humble verbal form -(e)tulita. Example. Hanna-ka halmeni-hanthey Hanna-S grandmother-IO sewul wuyu Sewul milk-DO cweyo. give-V Hanna gives Sewul milk to the grandmother. Example. Yecwungsayng-i sensayngnim-kkey chayk-ul ilke-cweyo. teacher-IO(H) book-DO read-V Female middle school student-S The female middle school student reads a book to the teacher. 8. 1 point is awarded for the use of the honorific indirect object case marker -kkey or humble verbal form -(e)tulita. (1) 1 point is awarded for the use of an honorific indirect object case marker -kkey. Example. Cayyengi-ka Cayyeng-S sensayngnim-kkey chayk-ul ilke-tulyeyo. teacher-IO(H) book-DO read-V(Hum) Cayyeng reads a book to the teacher. 242 (2) 1 point is awarded for the use of humble verbal form -(e)tulita. Example. Cayyengi-ka sensayngnim-eykey tulyeyo. teacher-IO Cayyeng-S khonchip-ul cornchip-DO give-V(Hum) Cayyeng gives corn chips to the teacher. Example. Hanna-nun Hanna-S halmeni-kkey appa-uy colep sacin-ul poye-tulyeyo. grandmother-IO(H) father’s graduation picture-DO show-V(Hum) Hanna shows the father’s graduation picture to the grandmother. (3) No penalty is given for the verbal inflection error. Example. Hyengcwungi-ka Hyengcwung-S halmeni-kkey iyaki chayk ilk-tulyeyo. grandmother-IO(H) story book-DO read-V(Hum) Hyengcwung reads a story book to the grandmother. (4) No penalty is given for the inaccurate pronunciation the learner produced. Example. Swuhyu-ka halmeni-kkey cha han can-ul tuleyo. Swuhyu-S grandmother-IO(H) a cup of tea-DO give-V(Hum) Swuhyu gives a cup of tea to the grandmother. (5) No penalty is given for the inaccurate use of verb tense. Example. Hanna-ka halmeni-hanthey insa-lul tulyesseyo. Hanna-S grandmother-IO greet-DO 243 give-V(Hum)-PAST Hanna greeted the grandmother. (6) No penalty is given for the wrong choice of vocabulary as long as the honorific indirect object case marker -kkey or humble verbal form -(e)tulita is accurately used. Example. Swuhyu-ka Swuhyu-S halapeci-kkey grandfather-IO(H) os-lul tulyeyo. clothes-DO give-V(Hum) Swuhyu gives clothes to the grandfather. 244 REFERENCES 245 REFERENCES Adams, R. (2007). Do second language learners benefit from interacting with each other? In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 29-51). Oxford: Oxford University Press. Ahn, S. (2012). The relationships between grammatical sensitivity, noticing of recasts, and learning of Korean object relative clauses through conversational interaction. Applied Language Learning, 22, 47-68. Albirini, A., Benmamoun, E., & Saadah, E. (2011). Grammatical features of Egyptian and Palestinian Arabic heritage speakers’ oral production. Studies in Second Language Acquisition, 33, 273-303. Alcón Soler, E. (2005). Does instruction work for learning pragmatics in the EFL context? System, 33, 417-435. Ammar, A., & Spada, N. (2006). One size fits all? Recasts, prompts, and L2 learning. Studies in Second Language Acquisition, 28, 543-574. Au, T., Knightly, L., Jun, S., & Oh, J. (2002). Overhearing a language during childhood. Psychological Science, 13, 238-243. Au, T., Oh, J., Knightly, L., Jun, S., & Romo, L. (2008). Salvaging a childhood language. Journal of Memory and Language, 58, 998-1011. Au, T., & Romo, L. (1997). Does childhood language experience help adult learners? In H. Chen (Ed.), The cognitive processing of Chinese and related Asian languages (pp. 417-441). Hong Kong: Chinese University Press. Bardovi-Harlig, K. (1987). Markedness and salience in second language acquisition. Language Learning, 37, 385-407. Bardovi-Harlig, K. (1999). Exploring the interlanguage of interlanguage pragmatics: A research agenda for acquisitional pragmatics. Language Learning, 49, 677-713. Bardovi-Harlig, K. (2001). Evaluating the empirical evidence: Grounds for instruction in pragmatics? In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 13-32). Cambridge: Cambridge University Press. Bardovi-Harlig, K., & Dörnyei, Z. (1998). Do language learners recognize pragmatic violations? Pragmatic versus grammatical awareness in instructed L2 learning. TESOL Quarterly, 32, 233-262. 246 Bardovi-Harlig, K., & Hartford, B. (1990). Congruence in native and nonnative conversations: Status balance in the academic advising session. Language Learning, 40, 467-501. Bardovi-Harlig, K., & Hartford, B. (1993). Learning the rules of academic talk: A longitudinal study of pragmatic development. Studies in Second Language Acquisition, 15, 279-304. Bardovi-Harlig, K., Hartford, B., Mahan-Taylor, R., Morgan, M., & Reynolds, D. (1991). Developing pragmatic awareness: Closing the conversation. ELT Journal, 45, 4-15. Billmyer, K. (1990). I really like your lifestyle: ESL learners learning how to compliment. Penn Working Papers in Educational Linguistics, 6, 31-48. Bley-Vroman, R. (1989a). The logical problem of second language learning. In S. Gass & J. Schachter (Eds.), Linguistic perspectives on second language acquisition (pp. 41-68). Cambridge: Cambridge University Press. Bley-Vroman, R. (1989b). The evolving context of the fundamental difference hypothesis. Studies in Second Language Acquisition, 31, 175-198. Bley-Vroman, R. (1990). The logical problem of foreign language learning. Linguistic Analysis, 20, 3-49. Bouton, L. (1994). Can NNS skill in interpreting implicature in American English be improved through explicit instruction? A pilot study. In L. Bouton & Y. Kachru (Eds.), Pragmatics and language learning [Monograph]. (pp. 88-109). Urbana-Champaign, Illinois: University of Illinois at Urbana-Champaign. Bouton, L. (1996). Pragmatics and language learning [Monograph]. Urbana-Champaign, Illinois: University of Illinois at Urbana-Champaign. Bowles, M. (2010). Concurrent verbal reports in second language acquisition research. Annual Review of Applied Linguistics, 30, 111-127. Bowles, M. (2011). Measuring implicit and explicit linguistic knowledge. Studies in Second Language Acquisition, 33, 247-271. Brinton, D., Kagan, O., & Bauckus, S. (2008). Heritage language education: A new field emerging. New York, NY: Routledge. Brown, L. (2010). Questions of appropriateness and authenticity in the representation of Korean honorifics in textbooks for second language learners. Language, Culture, and Curriculum, 23, 35-50. Brown, L. (2011). Korean honorifics and politeness in second language learning. Amsterdam: John Benjamins. 247 Bruhn de Garavito, J. (2002). Verb-raising in Spanish: A comparison of early and late bilinguals. th In B. Skarabela, S. Fish, & A. H.-J. D. (Eds.), Proceedings of the 26 Annual Boston University Conference on Language Development (pp. 84-94). Sommerville, MA: Cascadilla Press. Bybee, J., & Hopper, P. (2001). Frequency and the emergence of linguistic structure. Amsterdam: John Benjamins. Byon, A. (2000). Teaching Korean honorifics. The Korean Language in America, 5, 275-289. Byon, A. (2004). Learning linguistic politeness. Applied Language Learning, 14, 37-62. Campbell, R., & Rosenthal, J. (2000). Heritage languages. In J. Rosenthal (Ed.), Handbook of undergraduate heritage language education (pp. 165-184). Mahwah, NJ: Erlbaum. Campbell, R., & Wales, R. (1970). The study of language acquisition. In J. Lyons (Ed.), New horizons in linguistics (pp. 242-260). Baltimore, MD: Penguin Books. Carpenter, H., Jeon, K., MacGregor, D., & Mackey, A. (2006). Learners’ interpretations of recasts. Studies in Second Language Acquisition, 28, 209-236. Carreira, M. (2003). Profiles of SNS students in the twenty-first century: Pedagogical implications of the changing demographic and social status of U.S. Hispanics. In A. Roca & M. C. Colombi (Eds.), Mi lengua: Spanish as a heritage language in the United States (pp. 51-77). Washington, DC: Georgetown University Press. Carreira, M. (2007). Teaching Spanish in the U.S.: Beyond the one-size fits all paradigm. In K. Potowski & R. Cameron (Eds.), Spanish in contact: Policy, social and linguistic inquiries (pp. 61-79). Amsterdam: John Benjamins. Carreira, M., & Potowski, K. (2011). Commentary: Pedagogical implications of experimental SNS research. Heritage Language Journal, 8, 134-151. Carroll, S., & Swain, M. (1993). Explicit and implicit negative feedback: An empirical study of the learning of linguistic generalizations. Studies in Second Language Acquisition, 15, 357386. Celce-Murcia, M., & Olshtain, E. (2000). Discourse and context in language teaching. Cambridge: Cambridge University Press. Chomsky, N. (1986). Knowledge of language. New York: Praeger. Choo, M. (1999). Teaching language styles of Korean. The Korean language in America, 3, 7795. Choo, M. (2006). The structure and use of Korean honorifics. In H. Sohn (Ed.), Korean language 248 in culture and society (pp. 132-145). Honolulu, HI: University of Hawaii Press. Cohen, J. (1988). Statistical power analysis for the behavior sciences. New York, NY: Routledge. Cook, H. (2001). Why can’t learners of JFL distinguish polite from impolite speech style? In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 80-102). Cambridge: Cambridge University Press. Cook, V. (2002). Background to the L2 user. In V. Cook (Ed.), Portraits of the L2 user (pp. 128). Buffalo, NY: Multilingual Matters. de Bot, K., & Weltens, B. (1991). Recapitulation, regression, and language loss. In H.W. Seliger & R.M. Vago (Eds.), First language attrition (pp. 31-51). New York: Cambridge University Press. DeKeyser, R. (1993). The effect of error correction on L2 grammar knowledge and oral proficiency. The Modern Language Journal, 77, 501-514. DeKeyser, R. (2003). Implicit and explicit learning. In C. Doughty & M. Long (Eds.), Handbook of second language acquisition (pp. 313-348). Malden, MA: Blackwell. Domínguez, L. (2009). Charting the route of bilingual development: Contributions from heritage speakers’ early acquisition. International Journal of Bilingualism, 13, 271-287. Doughty, C. (2001). Cognitive underpinnings of focus on form. In P. Robinson (Ed.), Cognition and second language instruction (pp. 206-257). Cambridge: Cambridge University Press. Doughty, C., & Varela, E. (1998). Communicative focus on form. In C. Doughty & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp. 114-138). New York: Cambridge University Press. Duff, P. (1986). Another look at interlanguage talk: Taking task to task. In R. R. Day (Ed.), Talking to learn: Conversation in second language acquisition (pp. 147-181). Rowley, MA: Newbury House. Egi, T. (2007a). Interpreting recasts as linguistic evidence: The roles of linguistic target, length, and degree of change. Studies in Second Language Acquisition, 29, 511-537. Egi, T. (2007b). Recasts, learners’ interpretations, and L2 development. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 249-267). Oxford: Oxford University Press. Ellis, A. & Morrison, C. (1998). Real age-of-acquisition effects in lexical retrieval. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24, 515-523. 249 Ellis, N. (2002). Frequency effects in language processing. Studies in Second Language Acquisition, 24, 143-188. Ellis, N. (2012). Frequency effects. In P. Robinson (Ed.), Routledge encyclopedia of second language acquisition (pp. 260-265). Florence, KY: Routledge. Ellis, R. (1994). A theory of instructed second language acquisition. In N. Ellis (Ed.), Implicit and explicit learning of languages (pp. 79-114). San Diego, CA: Academic Press. Ellis, R. (2006). Researching the effects of form-focused instruction on L2 acquisition. AILA Review, 19, 18-41. Ellis, R. (2007). The differential effects of corrective feedback on two grammatical structures. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 339-360). Oxford: Oxford University Press. Ellis, R. (2008). The study of second language acquisition. Oxford: Oxford University Press. Ellis, R. (2010). Epilog: A framework for investigating oral and written corrective feedback. Studies in Second Language Acquisition, 32, 335-349. Ellis, R., Basturkmen, H., & Loewen, S. (2001). Learner uptake in communicative ESL lessons. Language Learning, 51, 281-318. Ellis, R., Loewen, S., & Erlam, R. (2006). Implicit and explicit corrective feedback and the acquisition of L2 grammar. Studies in Second Language Acquisition, 28, 339-368. Ellis, R., & Sheen, Y. (2006). Reexamining the role of recasts in second language acquisition. Studies in Second Language Acquisition, 28, 575-600. Fishman, J. (2001). 300-plus years of heritage language education in the United States. In J. Peyton, D. Ranard, & S. McGinnis (Eds.), Heritage languages in America: Preserving a national resource (pp. 81-89). Washington, DC: Center for Applied Linguistics. Flege, J. (1987). A critical period for learning to pronounce foreign language? Applied Linguistics, 8, 162-177. Fujii, A., & Mackey, A. (2009). Interactional feedback in learner-learner interactions in a taskbased EFL classroom. International Review of Applied Linguistics, 47, 267-301. Fukuya, Y., & Clark, M. (2001). A comparison of input enhancement and explicit instruction of mitigators. In L. Bouton (Ed.), Pragmatics and language learning [Monograph]. (pp. 111-130). Urbana-Champaign, Illinois: University of Illinois at Urbana-Champaign. Fukuya, Y., & Hill, Y. (2005). The effects of recasting on the production of pragmalinguistic conventions of request by Chinese learners of English. Issues in Applied Linguistics, 15, 250 59-91. Fukuya, Y., & Martínez-Flor, A. (2008). The interactive effects of pragmatic-eliciting tasks and pragmatic instruction. Foreign Language Annals, 41, 478-500. Gambhir, S. (2001). Truly less commonly taught languages and heritage language learners in the United States. In J. Peyton, D. Ranard, & S. McGinnis (Eds.), Heritage languages in America: Preserving a national resource (pp. 207-228). Washington, DC: Center for Applied Linguistics. Gass, S. (1988). Integrating research areas: A framework for second language studies. Applied linguistics, 9, 198-217. Gass, S. (1997). Input, interaction and the second language learner. Mahwah, NJ: Erlbaum. Gass, S. (2003). Input and interaction. In C. Doughty & M. Long (Eds.), Handbook of second language acquisition (pp. 224-255). Malden, MA: Blackwell. Gass, S., Behney, J., & Plonsky, L. (2013). Second Language Acquisition: An Introductory Course. New York, NY: Routledge. Gass, Behney, & Uzum. (in press). Inhibitory control, working memory, and L2 interaction gains. In K. Droździał-Szelest & M. Pawlak (Eds.), Psycholinguistic and sociolinguistic perspectives on second language learning and teaching: Studies in honor of Waldemar Marton. Heidelberg-New York: Springer. Gass, S., & Lewis, K. (2007). Perceptions about interactional feedback: differences between heritage language learners and non-heritage language learners. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp.77-99). Oxford: Oxford University Press. Gass, S., & Mackey, A. (2000). Stimulated recall. Mahwah, NJ: Erlbaum. Gass, S., & Mackey, A. (2002). Frequency effects and second language acquisition. Studies in Second Language Acquisition, 24, 249-260. Gass, S., & Mackey, A. (2006). Input, interaction and output: An overview. AILA Review, 19, 317. Gass, S., & Mackey, A. (2007a). Data elicitation for second and foreign language research. Mahwah, NJ: Erlbaum. Gass, S., & Mackey, A. (2007b). Input, interaction and output in second language acquisition. In B. VanPatten & J. Williams (Eds.), Theories in second language acquisition (pp. 175200). Mahwah, NJ: Erlbaum. 251 Gass, G., Mackey, A., & Ross-Feldman, L. (2005). Task-based interactions in classroom and laboratory settings. Language Learning, 55, 575-611. Gass, G., Mackey, A., & Ross-Feldman, L. (2011). Task-based interactions in classroom and laboratory settings. Language Learning, Supplement 1, 61, 189-220. Gass, S., & Selinker, L. (2008). Second language acquisition: An introductory course. New York, NY: Routledge. Gass, S., & Varonis, E. (1994). Input, interaction and second language production. Studies in Second Language Acquisition, 16, 283-302. Goo, J. (2012). Corrective feedback and working memory capacity in interaction-driven L2 learning. Studies in Second Language Acquisition, 34, 445-474. Halenko, N., & Jones, C. (2011). Teaching pragmatic awareness of spoken requests to Chinese EAP learners in the UK: Is explicit instruction effective? System, 39, 240-250. Han, Z. (2002). A study of the impact of recasts on tense consistency in L2 output. TESOL Quarterly, 36, 543-572. Han, Z. (2004). Fossilization in adult second language acquisition. Clevedon, UK: Multilingual Matters. Han, Z. (2012). Fossilization-A classic concern of SLA research. In S. Gass & A. Mackey (Eds.), The routledge handbook of second language acquisition (pp. 476-490). New York, NY: Routledge. Han, Y. (2010). The effect of implicit and explicit feedback: A study on the acquisition of mandarin classifiers by Chinese heritage and non-heritage language learners (Unpublished doctoral dissertation). University of Florida, Gainesville, FL. Hornberger, N., & Wang, S. (2008). Who are out heritage language learners? In D. Brinton, O. Kagan, & S. Bauckus (Eds.), Heritage language education: A new field emerging (pp. 335). London: Taylor & Francis Group. House, J. (1996). Developing pragmatic fluency in English as a foreign language: Routines and metapragmatic awareness. Studies in Second Language Acquisition, 18, 225-252. Hymes, D. (1972). On communicative competence. In J. Pride & J. Holmes (Eds.), Sociolinguistics: Selected readings (pp. 269-293). Harmondsworth, England: Penguin. Ishida, M. (2004). Effects of recasts on the acquisition of the aspectual form –te i-(ru) by learners of Japanese as a foreign language. Language Learning, 54, 311-394. Iwashita, N. (2003). Positive and negative input in task-based interaction: Differential effects on 252 L2 development. Studies in Second Language Acquisition, 25, 1-36. Jeon, K. (2004). Interaction-driven learning: Characterizing linguistic development (Unpublished doctoral dissertation). Georgetown University, Washington, DC. Jeon, K. (2007). Interaction-driven L2 learning: Characterizing linguistic development. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 379-403). Oxford: Oxford University Press. Jeon, E., & Kaya, T. (2006). Effects of L2 instruction on interlanguage pragmatic development: A meta-analysis. In J. Norris & L. Ortega (Eds.), Synthesizing research on language learning and teaching (pp. 165-211). Amsterdam: John Benjamins. Jescheniak, J., & Levelt, W. (1994). Word frequency effects in speech production: Retrieval of syntactic information and phonological form. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20, 824-843. Johnson, J., & Newport, E. (1989). Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language. Cognitive Psychology, 21, 60-99. Kagan, O. (2005). In support of a proficiency-based definition of heritage language learners: The case of Russian. The international journal of bilingual education and bilingualism, 8, 213221. Kagan, O., & Dillon, K. (2012). Heritage languages and L2 learning. In S. Gass & A. Mackey (Eds.), The routledge handbook of second language acquisition (pp. 491-505). New York: NY, Routledge. Kang, H. (2009). The relative efficacy of explicit and implicit feedback in the learning of a lesscommonly-taught foreign language. International Review of Applied Linguistics, 47, 303324. Kang, H. (2010a). Korean as a heritage language in the U.S. university classroom. Journal of the National Council of Less Commonly Taught Languages, 8, 141-167. Kang, H. (2010b). Negative evidence and its explicitness and positioning in the learning of Korean as a heritage language. The Modern Language Journal, 94, 582-599. Kanno, K., Hasegawa, T., Ikeda, K., Ito, Y., & Long, M. (2008). Relationships between prior language-learning experience and variation in the linguistic profiles of advanced Englishspeaking learners of Japanese. In D. Brinton, O. Kagan, & S. Bauckus (Eds.), Heritage language education (pp. 165-180). New York, NY: Routledge. Kasper, G. (2001). Classroom research on interlanguage pragmatics. In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 33-60). Cambridge: Cambridge University 253 Press. Kasper, G., & Dahl, M. (1991). Research methods in interlanguage pragmatics. Studies in Second Language Acquisition, 13, 215-247. Kasper, G., & Rose, K. (2001). Pragmatics in language teaching. In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 1-9). Cambridge: Cambridge University Press. Kasper, G., & Rose, K. (2002). Pragmatic development in a second language. Malden, MA: Blackwell Publishing. Ke, C. (1998). Effects of language background on the learning of Chinese characters among foreign language students. Foreign Language Annals, 31, 91-100. Keating, G., VanPatten, B., & Jegerski, J. (2011). Who was walking on the beach? Anaphora resolution in Spanish heritage speakers and adult second language learners. Studies in Second Language Acquisition, 33, 193-221. Kim, H.-Y. (2008). Commentary. Heritage Language Journal, 6, 94-104. Kim, H. H. (2001). Issues of heritage learners in Korean language classes. The Korean Language in America, 6, 257-274. Kim, H. H. (2002). The language backgrounds, motivations, and attitudes of heritage learners in KFL classes at University of Hawaii at Manoa. The Korean language in America, 7, 205221. Kim, H. H. (2005). Processing strategies and transfer of heritage and non-heritage learners of Korean (Unpublished doctoral dissertation). University of Hawaii, Honolulu, HI. Kim, H. H. (2008). Heritage and nonheritage learners of Korean: Sentence processing differences and their pedagogical implications. In K. Kondo-Brown & J. Brown (Eds.), Teaching Chinese, Japanese, and Korean heritage language students: Curriculum needs, materials, and assessment (pp. 99-134). New York, NY: Erlbaum. Kim, J. (2001). The degree of L1 interference among heritage and non-heritage learners of Korean: Do heritage students have advantages over non-heritage students? The Korean Language in America, 6, 285-296. Kim, H., & Mathes, G. (2001). Explicit vs. implicit corrective feedback. The Korean TESOL Journal, 1, 57-72. Kim, Y. (1997). The acquisition of Korean. In D. Slobin (Ed.), The cross-linguistic study of language acquisition (pp. 335-443). Hillsdale, NJ: Erlbaum. Kim-Renaud, Y. (1990). On Panmal in Korean. In E. Baek (Ed.), ICKL 7: Paper from the 254 seventh international conference on Korean linguistics (pp. 232-255). King, R. (1998). Korean as a heritage language vs. Korean as a foreign language in North America and the former USSR. Acta Koreana, 1, 24-40. Knightly, L., Jun, S., Oh, J., & Au, T. (2003). Production benefits of childhood overhearing. Journal of the Acoustic Society of America, 114, 465-474. Koike, D. (1989). Pragmatic competence and adult L2 acquisition: Speech acts and interlanguage. The Modern Language Journal, 73, 279-289. Koike, D., & Pearson, L. (2005). The effect of instruction and feedback in the development of pragmatic competence. System, 33, 481-501. Kondo-Brown, K. (2005). Differences in language skills: Heritage language learner subgroups and foreign language learners. The Modern Language Journal, 5, 563-581. Krashen, S. (1982). Principles and practice in second language acquisition. Oxford: Pergamon Press. Krashen, S. (1998). Language shyness and heritage language development. In S. Krashen, L. Tse, & J. McQuillan (Eds.), Heritage language development (pp. 41-49). Culver City, CA: Language Education Associates. Kubota, M. (1995). Teachability of conversational implicature to Japanese EFL learners. IRLT Bulletin, 9, 35-67. Lee, C. (1993). The acquisition of mood indicators in Korean. In S. Kuno, J. Whitman, Y.-S. Kang, I.-H. Lee, J. Maling, & Y. Kim (Eds.), Harvard studies in Korean linguistics (pp. 41-61). Seoul: Hanshin. Lee, I., & Ramsey, S. R. (2000). The Korean Language. Albany, NY: State University of New York Press. Lee, J. (2005). Through the learners’ eyes: Reconceptualizing the heritage and non-heritage learner of the less commonly taught languages. Foreign Language Annals, 38, 554-563. Lee, J., & Kim, H.-Y. (2008). Heritage language learners’ attitudes, motivations, and instructional needs: The case of postsecondary Korean language learners. In K. KondoBrown & J. Brown (Eds.), Teaching Chinese, Japanese, and Korean heritage language students: Curriculum needs, materials, and assessment (pp. 159-185). New York, NY: Erlbaum. Lee, J., & Shin, S. (2008). Korean heritage language education in the United States: The current state, opportunities, and possibilities. Heritage Language Journal, 6, 1-20. 255 Lee, K. (1997). Yuauy enecek thukcingkwa kyoswupep: Contaypep suptuk hyenhwang cosa [The first language acquisition of Korean honorification and its teaching]. Eneyenkwu, 2, 23-36. Lee, M. (1997). Acquisition of Korean referent honorifics by adult learners of Korean as a second language. The Korean Language in America, 2, 99-110. Lee, W. (1991). Honorifics and politeness in Korean (Unpublished doctoral dissertation). University of Wisconsin, Madison, WI. Leeman, J. (2003). Recasts and second language development: Beyond negative evidence. Studies in Second Language Acquisition, 25, 37-63. Leow, R., & Bowles, M. (2005). Attention and awareness in SLA. In C. Sanz (Ed.), Adult second language acquisition (pp. 179-203). Washington, DC: Georgetown University Press. Li, S. (2009). The differential effects of implicit and explicit feedback on second language (L2) learners at different proficiency levels. Applied Language Learning, 19, 53-79. Li, S. (2010a). The effectiveness of corrective feedback in SLA: A meta-analysis. Language Learning, 60, 309-365. Li, S. (2010b). Corrective feedback in perspective: The interface between feedback type, proficiency, the choice of target structure, and learners’ individual differences in working memory and language analytic ability (Unpublished doctoral dissertation). Michigan State University, East Lansing, MI. Liddicoat, A., & Crozet, C. (2001). Acquiring French interactional norms through instruction. In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 125-144). Cambridge: Cambridge University Press. Lindman, H. R. (1974). Analysis of variance in complex experimental designs. San Francisco, CA: W. H. Freeman & Co. Loewen, S. (2012). The role of feedback. In S. Gass & A. Mackey (Eds.), The routledge handbook of second language acquisition (pp. 24-40). New York, NY: Routledge. Loewen, S., & Nabei, T. (2007). Measuring the effects of oral corrective feedback on L2 knowledge. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 361-377). Oxford: Oxford University Press. Loewen, S., & Philp, J. (2006). Recasts in the adult English L2 classroom: characteristics, explicitness, and effectiveness. The Modern Language Journal, 90, 536-556. Long, M. (1981). Input, interaction, and second language acquisition. Annals of the New York Academy of Sciences, 379, 259-278. 256 Long, M. (1983). Native speaker/non-native speaker conversation and the negotiation of comprehensible input. Applied Linguistics, 4, 126-141. Long, M. (1996). The role of the linguistic environment in second language acquisition. In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of second language acquisition (pp. 413-468). New York, NY: Academic Press. Long, M. (2006). Problems in SLA. Mahwah, NJ: Erlbaum. Long, M., Inagaki, S., & Ortega, L. (1998). The role of negative feedback in SLA: Models and recasts in Japanese and Spanish. The Modern Language Journal, 82, 357-371. Lörscher, W., & Schulze, R. (1988). On polite speaking and foreign language classroom discourse. IRAL, 26, 183-199. Lyon, J. (1970). New horizons in linguistics. Harmondsworth, UK: Penguine Books. Niezgoda, K., & Röver, C. (2001). Pragmatic and grammatical awareness: A function of the learning environment? In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 63-79). Cambridge: Cambridge University Press. Lynch, A. (2003). The relationship between second and heritage language acquisition: Notes on research and theory building. Heritage Language Journal, 1, 1-18. Lyster, R. (1994). The effect of functional-analytic teaching on aspects of French immersion students’ sociolinguistic competence. Applied Linguistics, 15, 263-287. Lyster, R. (1998a). Recasts, repetition, and ambiguity in L2 classroom discourse. Studies in Second Language Acquisition, 20, 51-81. Lyster, R. (1998b). Negotiation of form, recasts, and explicit correction in relation to error types and learner repair in immersion classrooms. Language Learning, 48, 183-218. Lyster, R. (2004). Different effects of prompts and recasts in form-focused instruction. Studies in Second Language Acquisition, 26, 399-432. Lyster, R., & Izquierdo, J. (2009). Prompts versus recasts in dyadic interaction. Studies in Second Language Acquisition, 59, 453-498. Lyster, R., & Mori, H. (2006). Interactional feedback and instructional counterbalance. Studies in Second Language Acquisition, 28, 269-300. Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake: Negotiation of form in communicative classrooms. Studies in Second Language Acquisition, 19, 37-66. Lyster, R., & Saito, K. (2010). Oral feedback in classroom SLA. Studies in Second Language 257 Acquisition, 32, 265-302. Mackey, A. (1999). Input, interaction, and second language development. Studies in Second Language Acquisition, 21, 557-587. Mackey, A. (2007). Introduction: The role of conversational interaction in second language acquisition. In A. Mackey (Ed.), Conversational interaction in SLA: A collection of empirical studies (pp. 1-26). New York: Oxford University Press. Mackey, A., & Gass, S. (2006). Introduction. Studies in Second Language Acquisition, 28, 169178. Mackey, A., Gass, S., & McDonough, K. (2000). How do learners perceive interactional feedback? Studies in Second Language Acquisition, 22, 471-497. Mackey, A., & Goo, J. (2007). Interaction research in SLA: A meta-analysis and research synthesis. In A. Mackey (Ed.), Conversational interaction in SLA: A collection of empirical studies (pp. 408-452). Oxford: Oxford University Press. Mackey, A., & Oliver, R. (2002). Interactional feedback and children’s L2 development. System, 30, 459-477. Mackey, A., & Philp, J. (1998). Conversational interaction and second language development: Recasts, responses, and red herrings? The Modern Language Journal, 82, 338-356. Mackey, A., Philp, J., Egi, T., Fujii, A., & Tatsumi, T. (2002). Individual differences in working memory, noticing of interactional feedback and L2 development. In P. Robinson (Ed.), Individual differences and instructed language learning (pp. 181-209). Philadelphia: John Benjamins. Mackey, A., & Sachs, R. (2012). Older learners in SLA research: A first look at working memory, feedback, and L2 development. Language Learning, 62, 704-740. Maher, J. (1991). A crosslinguistic study of language contact and language attrition. In H. W. Seliger & R. M. Vago (Eds.), First language attrition (pp. 67-84). New York: Cambridge University Press. Martínez-Flor, A. (2006). The effectiveness of explicit and implicit treatments on EFL learners’ confidence in recognizing appropriate suggestions. In K. Bardovi-Harlig, C. FélixBrasdefer, & A. Omar (Eds.), Pragmatics language learning (pp. 119-225). Honolulu, HI: University of Hawaii Press. McDonough, K. (2007). Interactional feedback and the emergence of simple past activity verbs in L2 English. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 323-338). Oxford: Oxford University Press. 258 Montrul, S. (2002). Incomplete acquisition and attrition of Spanish tense/aspect distinctions in adult bilinguals. Bilingualism: Language and Cognition, 5, 39-69. Montrul, S. (2004). Subject and object expression in Spanish heritage speakers: A case of morphosyntactic convergence. Bilingualism: Language and Cognition, 7, 125-142. Montrul, S. (2005). Second language acquisition and first language loss in adult early bilinguals: Exploring some differences and similarities. Second Language Research, 21, 199-249. Montrul, S. (2006). Incomplete acquisition as a feature of L2 and bilingual grammars. In R. Slabakava, S. Montrul, & P. Prévost (Eds.), Inquiries in language development: Studies in honor of Lydia White (pp. 335-359). Amsterdam: John Benjamins. Montrul, S. (2008a). Incomplete acquisition in bilingualism: Re-examining the age factor. Amsterdam: John Benjamins. Montrul, S. (2008b). Second language acquisition welcomes the heritage language learner: Opportunities of a new field. Second Language Research, 24, 487-506. Montrul, S. (2010a). Current issues in heritage language acquisition. Annual Review of Applied Linguistics, 30, 3-23. Montrul, S. (2010b). How similar are adult second language learners and Spanish heritage speakers? Spanish clitics and word order. Applied Psycholinguistics, 31, 167-207. Montrul, S. (2010c). Dominant language transfer in adult second language learners and heritage speakers. Second Language Research, 26, 293-327. Montrul, S. (2011a). Introduction: The linguistic competence of heritage speakers. Studies in Second Language Acquisition, 33, 155-161. Montrul, S. (2011b). Morphological errors in Spanish second language learners and heritage speakers. Studies in Second Language Acquisition, 33, 163-192. Montrul, S., & Bowles, M. (2010). Is grammar instruction beneficial for heritage language learners? Dative case marking in Spanish. Heritage Language Journal, 7, 47-73. Montrul, S., Foote, R., & Perpiñán, S. (2008). Gender agreement in adult second language learners and Spanish heritage speakers: The effects of age and context of acquisition. Language Learning, 58, 503-53. Montrul, S., & Perpiñán, S. (2011). Assessing differences and similarities between instructed heritage language learners and L2 learners in their knowledge of Spanish tense-aspect and mood (TAM). Heritage Language Journal, 8, 90-133. Moon, S. (2010). Yuauy contay phyohyen paltaley kwanhan congcek yenkwu: 23kaywel~36 259 kaywel atongul taysangulo [A longitudinal study of language acquisition of proper honorific development: from 23 to 36 months children] (Unpublished master’s thesis). Hanyang University, Seoul. Nagata, N. (1993). Intelligent computer feedback for second language instruction. The Modern Language Journal, 77, 330-339. Nassaji, H. (2009). Effects of recasts and elicitations in dyadic interaction and the role of feedback explicitness. Language Learning, 59, 411-452. Newport, E. (1990). Maturational constraints on language learning. Cognitive Science, 14, 11-28. Newport, E. (1991). Contrasting conceptions of the critical period of language. In S. Carey & R. Gelman (Eds.), The epigenesist of mind: Essays on biology and cognition (pp. 111-130). Hillsdale, NJ: Erlbaum. Nicholas, H., Lightbown, P., & Spada, N. (2001). Recasts as feedback to language learners. Language Learning, 51, 719-758. Nguyen, T., Pham, T., & Pham, M. (2012). The relative effects of explicit and implicit formfocused instruction on the development of L2 pragmatic competence. Journal of Pragmatics, 44, 416-434. Norris, J., & Ortega, L. (2000). Effectiveness of L2 instruction: A research synthesis and quantitative meta-analysis. Language Learning, 50, 417-528. O’Grady, W., Kwak, H., Lee, O., & Lee, M. (2011). An emergentist perspective on heritage language acquisition. Studies in Second Language Acquisition, 33, 223-245. O’Grady, W., Lee, M., & Choo, M. (2001). The acquisition of relative clauses by heritage and non-heritage learners of Korean as a second language: A comparative study. Journal of Korean Language Education, 12, 283-294. Oh, J., Jun, S., Knightly, L., & Au, T. (2003). Holding on to childhood language memory. Cognition, 86, B53-B64. Oldfield, R. C., & Wingfield, A. (1965). Response latencies in naming objects. Quarterly Journal of Experimental Psychology, 17, 273-281. Oliver, R. (1995). Negative feedback in child NS/NNS conversation. Studies in Second Language Acquisition, 17, 459-481. Oliver, R. (2002). Age differences in negotiation and feedback in classroom and pair work. Language Learning, 50, 119-151. Olshtain, E., & Barzilay, M. (1991). Lexical retrieval difficulties in adult language attrition. In H. 260 W. Seliger & R. M. Vago (Eds.), First language attrition (pp. 139-150). New York: Cambridge University Press. Olshtain, E., & Blum-Kulka, S. (1985). Degree of approximation: Nonnative reactions to native speech act behavior. In S. Gass & C. Madden (Eds.), Input in second language acquisition (pp. 303-325). New York: Newbury House. Olshtain, E., & Cohen, D. (1990). The learning of complex speech act behavior. TESL Canada Journal, 7, 45-65. Oyama, S. (1976). A sensitive period for the acquisition of a nonnative phonological system. Journal of Psycholinguistic Research, 5, 261-283. Panova, I., & Lyster, R. (2002). Patterns of corrective feedback and uptake in an adult ESL classroom. TESOL Quarterly, 36, 573-595. Park, J., & Kim, M. (2010). Man 3·4·5sey yuauy contaysmal suptukey kwanhan yenkwu [A study to the acquisition of honorific markers by three-, four-, and five-year-old young children]. Atonghakhoyci, 31, 153-166. Pavlenko, A., & Jarvis, S. (2002). Bidirectional transfer. Applied Linguistics, 23, 190-214. Philp, J. (2003). Constraints on “noticing the gap”: Nonnative speakers’ noticing of recasts in NS-NNS interaction. Studies in Second Language Acquisition, 25, 99-126. Pica, T., Holliday, L., Lewis, N., & Morgenthaler, L. (1989). Comprehensible output as an outcome of linguistic demands on the learner. Studies in Second Language Acquisition, 11, 63-90. Pica, T., Kanagy, R., & Falodun, J. (1993). Choosing and using communication tasks for second language instruction and research. In G. Crookes & S. Gass (Eds.), Tasks and language learning: Integrating theory and practice (pp. 9-34). Philadelphia, PA: Multilingual Matters. Polinsky, M. (1995). Cross-linguistic parallels in language loss. Southwest Journal of Linguistics, 14, 88-123. Polinsky, M. (2000). A composite linguistic profile of a speaker of Russian in the USA. In O. Kagan & B. Rifkin (Eds.), The Learning and Teaching of Slavic Languages and Cultures (pp. 437-466). Bloomington, IN: Slavica. Polinsky, M. (2007). Incomplete acquisition: American Russian. Journal of Slavic linguistics,14, 191-262. Polinsky, M. (2008). Heritage language narratives. In D. Brinton, O. Kagan, & S. Bauckus (Eds.), Heritage language education: A new field emerging (pp. 149-164). London: Taylor & 261 Francis Group. Polinsky, M. (2011). Reanalysis in adult heritage language: New evidence in support of attrition. Studies in Second Language Acquisition, 33, 305-328. Polinsky, M., & Kagan, O. (2007). Heritage languages: In the ‘wild’ and in the classroom. Language and linguistics compass, 1, 368-395. Potowski, K. (2002). Experiences of Spanish heritage speakers in university foreign language courses and implications for teacher training. ADFL Bulletin, 33, 35-42. Potowski, K., Jegerski, J., & Morgan-Short, K. (2009). The effects of instruction on linguistic development in Spanish heritage language speakers. Language Learning, 59, 537-579. Révész, A., & Han, Z. (2006). Task content familiarity, task type, and efficacy of recasts. Language Awareness,15, 160-179. Robinson, P. (1995). Attention, memory, and the “noticing” hypothesis. Language Learning, 45, 283-331. Robinson, P. (1996). Consciousness, rules, and instructed second language acquisition. New York, NY: Peter Lang. Robinson, P. (2002) Learning conditions, aptitude complexes and SLA: A framework for research and pedagogy. In P. Robinson (Ed.), Individual differences and instructed language learning (pp. 113-136). Oxford: Oxford University Press. Robinson, P., Mackey, A., Gass, S., & Schmidt, R. (2012). Attention and awareness in second language acquisition. In S. Gass & A. Mackey (Eds.), The routledge handbook of second language acquisition (pp. 247-267). New York, NY: Routledge. Rosa, E., & Leow, R. (2004). Awareness, different learning conditions, and second language development. Applied Psycholinguistics, 25, 269-292. Rose, K. (2005). On the effects of instruction in second language pragmatics. System, 33, 385399. Rose, K., & Kasper, G. (2001). Pragmatics in language teaching. Cambridge: Cambridge University Press. Rose, K., & Ng Kwai-fun, C. (2001). Inductive and deductive teaching of compliments and compliment responses. In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 145-170). Cambridge: Cambridge University Press. Ross-Feldman, L. (2007). Interaction in the L2 classroom: does gender influence learning opportunities? In A. Mackey (Ed.), Conversational interaction in second language 262 acquisition (pp. 53-77). Oxford: Oxford University Press. Russell, J., & Spada, N. (2006). The effectiveness of corrective feedback for second language acquisition: A meta-analysis of the research. In J. Norris & L. Ortega (Eds.), Synthesizing research on language learning and teaching (pp. 131-164). Amsterdam: John Benjamins. Sagarra, N. (2007). From CALL to face-to-face interaction: the effect of computer-delivered recasts and working memory on L2 development. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 229-248). Oxford: Oxford University Press. Sanz, C. (2003). Computer delivered implicit vs. explicit feedback in processing instruction. In B. VanPatten (Ed.), Processing instruction: Theory, research, and commentary (pp. 241-255). Mahwah, NJ: Erlbaum. Sato, M., & Lyster, R. (2007). Modified output of Japanese EFL learners: variable effects of interlocutor versus feedback types. In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 123-142). Oxford: Oxford University Press. Seliger, H. W., & Vago, R. M. (1991). The study of first language attrition. In H. W. Seliger & R. M. Vago (Eds.), First language attrition (pp. 3-15). New York: Cambridge University Press. Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, 11, 129-158. Schmidt, R. (1993). Consciousness, learning and interlanguage pragmatics. In G. Kasper & S. Blum-Kulka (Eds.), Interlanguage Pragmatics (pp. 21-42). New York: Oxford University Press. Schmidt, R. (1994). Implicit learning and the cognitive unconsciousness: Of artificial grammars and SLA. In N. Ellis (Ed.), Implicit and Explicit Learning of Languages (pp. 165-209). London: Academic Press. Schmidt, R. (1998). Attention. University of Hawaii Working Papers in ESL, 15, 1-34. Schmidt, R. (2001). Attention. In P. Robinson (Ed.), Cognition and Second Language Instruction (pp. 3-32). Cambridge: Cambridge University Press. Schwartz, B. (1993). On explicit and negative data effecting and affecting competence and linguistic behavior. Studies in Second Language Acquisition, 15, 143-145. Selinker, L. & Lamendella, J. (1978). Two perspectives on fossilization in interlanguage learning. Interlanguage Studies Bulletin, 3, 143-191. Sharwood Smith, M. (1993). Input enhancement in instructed SLA: Theoretical bases. Studies in Second Language Acquisition, 15, 165-179. 263 Sheen, Y. (2004). Corrective feedback and learner uptake in communicative classrooms across instructional settings. Language Teaching Research, 8, 263-300. Sheen, Y. (2006). Exploring the relationship between characteristics of recasts and learner uptake. Language Teaching Research, 10, 361-392. Sheen, Y. (2007). The effects of corrective feedback, language aptitude, and learner attitudes on the acquisition of English articles. In A. Mackey (Ed.), Conversational interaction in SLA: A collection of empirical studies (pp. 301-322). Oxford: Oxford University Press. Sheen, Y. (2008). Recasts, language anxiety, modified output, and L2 learning. Language Learning, 58, 835-874. Sheen, Y. (2010a). The role of oral and written corrective feedback in SLA: Introduction. Studies in Second Language Acquisition, 32, 169-179. Sheen, Y. (2010b). Differential effects of oral and written corrective feedback in the ESL classroom. Studies in Second Language Acquisition, 32, 203-234. Simard, D., & Wong, W. (2001). Alertness, orientation, and detection: The conceptualization of attentional functions in SLA. Studies in Second Language Acquisition, 23, 103-124. Sohn, H. (1999). The Korean language. Cambridge: Cambridge University Press. Sohn, H. (2001). Teaching politeness routines in Korean. The Korean Language in America, 6, 25-35. Soler, E. (2002). Relationship between teacher-led versus learners’ interaction and the development of pragmatics in the EFL classroom. International Journal of Educational Research, 37, 359-377. Spada, N., & Tomita, Y. (2010). Interactions between type of instruction and type of language feature: A meta-analysis. Language Learning, 60, 263-308. Strauss, S., & Eun, J. (2005). Indexicality and honorific speech level choice in Korean. Linguistics, 43, 611-651. Stutterheim, C., & Nüse, R. (2003). Processes of conceptualization in language production: language-specific perspectives and event construal. Linguistics, 41, 851-881. Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. Gass & C. Madden (Eds.), Input in second language acquisition (pp. 235-252). Rowley, MA: Newbury House. Swain, M. (1993). The output hypothesis: Just speaking and writing aren’t enough. The Canadian Modern Language Review, 50, 158-164. 264 Swain, M. (1995). Three functions of output in second language learning. In G. Cook & B. Seidelhofer (Eds.), Principle and practice in applied linguistics: Studies in honour of H. G. Widdowson (pp. 125-144). Oxford: Oxford University Press. Swain, M. (1998). Focus on form through conscious reflection. In C. Doughty & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp. 64-81). Cambridge: Cambridge University Press. Swain, M. (2005). The output hypothesis: Theory and research. In E. Hinkel (Ed.), Handbook of research in second language teaching and learning (pp. 471-483). Mahwah, NJ: Erlbaum. Taguchi, N. (2011). Teaching pragmatics: Trends and issues. Annual Review of Applied Linguistics, 31, 289-310. Takahashi, S. (2001). The role of input enhancement in developing pragmatic competence. In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 171-199). Cambridge: Cambridge University Press. Tarone, E. (1994). Interlanguage. In R. Asher (Ed.), The encyclopedia of language and linguistics (pp. 1715-1719). Oxford: Pergamon. Tateyama, Y. (2001). Explicit and implicit teaching of pragmatic routines: Japanese sumimasen. In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 200-222). Cambridge: Cambridge University Press. Tateyama, Y., Kasper, G., Mui, L., Tay, H., & Thananart, O. (1997). Explicit and implicit teaching of pragmatics routines. In L. Bouton (Ed.), Pragmatics and language learning [Monograph]. (pp. 163-177). Urbana-Champaign, Illinois: University of Illinois at UrbanaChampaign. Thomas, J. (1983). Cross-cultural pragmatic failure. Applied Linguistics, 4, 91-112. Tomlin, R. S., & Villa, V. (1994). Attention in cognitive science and second language acquisition. Studies in Second Language Acquisition, 16, 183-203. Turnbull, W. (2001). An appraisal of pragmatic elicitation techniques for the social psychological study of talk: The case of request refusals. Pragmatics, 11, 31-61. Trofimovich, P., Ammar, A., & Gatbonton, E. (2007). The role of attention, memory, and analytical ability. In A. Mackey (Ed.), Conversational interaction in second language acquisition: A collection of empirical studies (pp. 171-195). Oxford: Oxford University Press. Truscott, J. (1999). What’s wrong with oral grammar correction. The Canadian Modern Language Review, 55, 437-455. 265 Valdés, G. (1995). The teaching of minority languages as academic subjects: Pedagogical and theoretical challenges. The Modern Language Journal, 79, 299-328. Valdés, G. (1997). The teaching of Spanish to bilingual Spanish-speaking students: Outstanding issues and unanswered questions. In M. C. Columbi & F. X. Alarcón (Eds.), La ensenanza del español a hispanohablantes: Praxis y teoria [Teaching Spanish to Spanish speakers: Practice and theory] (pp. 93-101). Boston, MA: Houghton Mifflin. Valdés, G. (2001a). Heritage language students: Profiles and possibilities. In J. Peyton, D. Ranard, & S. McGinnis (Eds.), Heritage languages in America: Preserving a national resource (pp. 37-80). Washington, DC: Center for Applied Linguistics. Valdés, G. (2001b). Learning and not learning English: Latino students in American schools. New York, NY: Teachers College Press. Valdés, G. (2001c). The teaching of heritage languages: An introduction for Slavic-teaching professionals. In O. Kagan & B. Rifkin (Eds.), The learning and teaching of Slavic languages and cultures (pp. 375-403). Bloomington, IN: Slavica. Valdés, G. (2005). Bilingualism, heritage language learners and SLA research: Opportunities lost or seized? The Modern Language Journal, 89, 410-426. Valdés, G., Fishman, J., Chávez, R., & Pérez, W. (2006). Developing minority language resources: The case of Spanish in California. Cleveland, UK: Multilingual Matters. Valdés, G., & Geoffrion-Vinci, M. (1998). Chicaco Spanish: The role of the “underdeveloped” code in bilingual repertoires. The Modern Language Journal, 82, 473-501. Van Deusen-Scholl, N. (2003). Toward a definition of heritage language: Sociopolitical and pedagogical considerations. Journal of Language, Identity, and Education, 2, 211-230. Yang, Y., & Lyster, R. (2010). Effects of form-focused practice and feedback on Chinese EFL learners’ acquisition of regular and irregular past tense forms. Studies in Second Language Acquisition, 32, 235-263. Yoshida, R. (2010). How do teachers and learners perceive corrective feedback in the Japanese language classroom? The Modern Language Journal, 94, 293-314. Yoshimi, D. (2001). Explicit instruction and JFL learner’s use of interactional discourse markers. In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 223-245). Cambridge: Cambridge University Press. Yun, S. (1993). Honorific agreement (Unpublished doctoral dissertation). University of Hawaii, Honolulu, HI. Wang, H. (1995). The impact of family background on the acquisition of Korean honorifics. The 266 Korean Language in America, 1, 197-211. Wiley, T. (2001). On defining heritage languages and their speakers. In J. K. Peyton, D. A. Ranard, & S. McGinnis (Eds.), Heritage languages in America: Preserving a national resource (pp. 29-36). Washington, DC: Center for Applied Linguistics. Wiley, T. (2005). The reemergence of heritage and community language policy in the U.S. national spotlight. The Modern Language Journal, 89, 594-601. 267