University of Southern California Dissertations and Theses

(USC Thesis Other)

Register and style variation in speakers of Spanish as a heritage and as a second language

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content REGISTER AND STYLE VARIATION IN SPEAKERS OF SPANISH AS A HERITAGE AND AS A SECOND LANGUAGE by Ana Sánchez Muñoz ________________________________________________________________________ A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (HISPANIC LINGUISTICS) May 2007 Copyright 2007 Ana Sánchez Muñoz Dedication To my parents: Jesús Sánchez González and María Salud Muñoz Arriero. Sin vuestro apoyo nunca hubiera logrado llegar hasta aquí. ii Acknowledgments The writing of this dissertation would not have been possible without the personal and practical support of numerous people. Thus, my sincere gratitude goes to my professors and mentors, my colleagues and peers in the Spanish and Linguistics departments at USC, my family, friends, and my husband for their love, support, and patience over the last few years. I am appreciative and thankful for the support of my faculty advisor and dissertation chair, Dr. Carmen Silva-Corvalán. Carmen has given me so much important feedback on my research and has provided me with invaluable assistance for my professional development all these years as a graduate student at USC. I would also like to thank the other members of my committee: Dr. Edward Finegan, Dr. Mario Saltarelli, and Dr. Robert Rueda for their valuable, constructive criticism. In addition, I want to thank all the students who participated in this project. They took time from their busy schedules to meet with me for interviews and other data collection situations. This project is about them and for them, and it would not have been possible without their candor. Thank you so much! Special thanks go to the faculty of the Department of Spanish at USC for helping me out with the recruitment of speakers for this dissertation. Many of the faculty and staff have become friends with whom I have shared enjoyable moments in and outside the Spanish department. In particular, I would like to thank Dr. Gayle Vierma for her valuable advice, support, and encouragement, both at a professional and personal level, throughout these years. iii All my fellow students in Linguistics at USC have been very important to me, providing support and companionship all along the way. Thanks to the students in my year: Jelena Krivokapi ć, Rebeka Campos, Eunjeong Oh, and Fetiye Karabay. Going beyond my year, I would like to thank Fleur van der Houwen, Simona Montanari, Isabelle Roy, and Michal Temkin Martínez among others. Many thanks are due to the Hispanic Linguistics students, past and present, for their helpful feedback regarding my research, and for their friendship and support: Álvaro Cerrón- Palomino, Asier Álcazar, Emily H. Nava, Michael Rushforth, Mónica Cabrera, Omar Beas, Rebeka Campos, and Roberto Mayoral. A todos, ¡gracias!. I would like to specially thank my good friends Jelena and Fleur for the good times we have spent together, for innumerable interesting discussions on all sorts of topics, for their encouragement and their trust in me, and for sharing with me so much good food and good wine. Rebeka Campos has also been a great friend and a constant source of support, always generous with her time and willing to help and listen, gracias. Finally, I must thank my family and friends who have supported me so much emotionally over the years: My parents, Marisa y Jesús; my grandparents, Mara y Pura; my brother Carlos, my sister Beatriz, and my beautiful nephew César (and Bruno, who is on the way!); my friend Conny, for always being there even if it takes traveling around the world to meet me; and last, but definitely not least, my husband, Dusty, thank you for all your patience and for helping me so much in every possible way. All of you are special people. Thanks for believing in me. iv Table of Contents Dedication ii Acknowledgements iii List of Tables vii List of Figures viii Abstract ix Chapter 1. Register variation in heritage and second languages 1 1. Introduction 1 1.1. Objectives 1 1.2. Background 1 1.3. Working assumptions: Register and style 4 2. Research questions 8 3. Literature review and conceptual framework 8 4. Linguistic features correlated with register variation in heritage languages 17 4.1. Discourse particles 20 4.2. Contractions 23 4.3. Lexical choices 24 5. Methodology 27 5.1. Sampling methods and characteristics of the speakers 28 5.2. Situations of data collection 30 5.3. Data collection procedures and materials 33 5.4. Statistical analysis 34 Chapter 2. Discourse particles and contractions 36 1. Introduction 36 2. Discourse particles 39 2.1. Como (‘as’/‘like’) 40 2.2. So, así que, entonces 48 2.3. Punctors 52 3. Contractions 54 3.1. Pues (‘so’/‘well’) and para (‘for’/‘so that’) 55 3.2. Estar (‘to be’) 56 4. Hypotheses 56 5. Results and discussion 58 5.1. Como 59 5.2. So, así que, and entonces 65 5.3. Punctors 70 v 5.4. Contractions 72 5.5. Discussion 75 6. Conclusions 86 Chapter 3. Lexical Choices 88 1. Introduction 88 2. Lexical features analyzed 89 2.1. Informal, formal, and technical vocabulary 89 2.2. Lexical transfer 105 2.3. Attributive adjectives 108 2.4. Type/token ratio 111 3. Hypotheses 112 4. Results and discussion 114 4.1. Informal, formal, and technical vocabulary 114 4.2. Lexical transfer 116 4.3. Attributive adjectives 120 4.4. Type/token ratio 121 4.5. Discussion 122 5. Conclusions 133 Chapter 4. Conclusions 136 1. Summary of the findings 139 1.1. Linguistic features in the speech of HLS across registers 139 1.2. Comparison between HLS and SLS 141 2. Pedagogical implications 144 3. Direction of future research and concluding remarks 149 Bibliography 153 Appendices Appendix A. Speaker information and coding 166 Appendix B. HLS Biographical and language background questionnaire 168 Appendix C. SLS Biographical and language background questionnaire 170 Appendix D. Spanish courses taken by HLS and SLS in college 171 vi List of Tables Table 1. Framework for the study of register variation 16 Table 2. Linguistic features included in this dissertation 20 Table 3. Characteristics of the registers analyzed in this dissertation 31 Table 4. Discourse particles studied 40 Table 5. Rate of use of como produced by HLS across registers 59 Table 6. Frequency of the three introducers of direct quotations used by HLS in each register 61 Table 7. Rate of use of como produced by SLS across registers 61 Table 8. Rate of use of like produced by SLS across registers 63 Table 9. Rate of use of punctors by HLS and SLS across registers 70 Table 10. Mean values for the type/token ratios of HLS and SLS across registers 122 Table 11. Rate of use of colloquialisms, slang, and idioms across situations by HLS 123 vii List of Figures Figure 1. Register and Style relationship 7 Figure 2. Percentage of use of como across registers by HLS 60 Figure 3. Percentage of SLS use of como and like across registers 64 Figure 4. Percentage of use of so, así que, and entonces by HLS across registers 66 Figure 5. Use of so, así que, and entonces by individual HLS in the three registers 67 Figure 6. Percentage and rate of use of so, así que, and entonces by SLS across registers 68 Figure 7. Percentage of pues versus pos across registers by HLS 72 Figure 8. Percentage of para versus pa across registers by HLS 73 Figure 9. Rate of use and percentage of estar-contractions across registers by HLS 74 Figure 10. Lexical transfer categories 107 Figure 11. HLS use of informal, formal, and technical lexicon 115 Figure 12. SLS use of informal, formal, and technical lexicon 115 Figure 13. HLS rate of lexical transfers across registers 117 Figure 14. SLS rate of lexical transfer across registers 117 Figure 15. HLS percentages of lexical features by register 118 Figure 16. SLS percentages of lexical features by register 119 Figure 17. Rate of use of lexical features by group of speakers 120 Figure 18. Percentage of attributive adjectives in different registers 121 viii Abstract One of the fundamental principles of sociolinguistics is that language is not homogeneous and that no single person speaks in the same way all the time. Numerous studies have provided evidence of linguistic variation across situations of use in English (e.g. Bell 1984; Biber 1988; Biber and Finegan 1994). However, under special conditions when a language is restricted to very familiar situations, speakers might not show register variation (Dressler 1982). For most heritage speakers of Spanish in the U.S., English is the dominant language while Spanish is largely restricted to home and family interactions. This dissertation explores the hypothesis of variation across registers in Spanish as a heritage language. Additionally, it examines speakers of Spanish as a second language since Spanish is also their non-dominant language. For the purpose of studying register and style variation, this dissertation focuses on several linguistic features that are expected to vary in relation to the type of register: discourse particles, contractions, and various lexical choices. The data analyzed come from recorded spoken samples produced in Spanish by heritage and second language speakers and collected in three situations of use: conversations, interviews, and presentations, ranging on a scale from less to more formal. The results indicate that both heritage and second language speakers show linguistic variation in their Spanish across registers. The results also reveal some quantitative as well as qualitative differences between the production of heritage and ix second language speakers across registers. These contrasts are the result of the different input to which the speakers have been exposed to during acquisition. This dissertation contributes to further our understanding of bilingualism by examining Spanish as a heritage and as a second language across different registers, which has not been previously investigated. It provides evidence of variation in a relatively small range of registers in the speakers’ non-dominant language. This is an important finding since it shows that even when the use of the language is largely restricted to a particular domain (home and family interactions for heritage speakers and classroom interactions for second language learners), we can still find register variation. 1 Chapter 1. Register variation in heritage and second languages 1. Introduction 1.1. Objectives There has been little research on register and style variation in the language spoken by heritage language speakers of Spanish. The purpose of this dissertation is to help fill in this gap by studying the oral production of bilingual heritage speakers of Spanish in different situations of use in order to examine whether there is variation in register and style reflected in the speech of these speakers. This dissertation focuses on the oral production of college students. I examine the Spanish spoken by heritage language speakers (henceforth HLS) and the Spanish produced by English speakers who are studying Spanish as a second language (henceforth SLS). Since Spanish is also their non-dominant language, the purpose of including the oral production of SLS is to examine whether they show variation and investigate how their oral production compares to that of HLS in terms of linguistic features across styles and registers. 1.2. Background Linguistic variation is one of the fundamental principles of sociolinguistic investigation. Language is used differently depending on the situation and no speaker talks in the same manner in all occasions (Labov 1970; Hymes 1984). The study of register variation has provided ample evidence of different use of linguistic features depending on the register (e.g. Biber 1988). However, minority languages have been reported to lose stylistic range as a result of various processes of language attrition, 2 such as simplification and loss (e.g. Dorian 1981; Silva-Corvalán 1994). In some extreme cases of language attrition, speakers may become monostylistic since their language is restricted to very casual situations at home and with familiar interlocutors (Dressler 1982, 1988). In the case of Spanish in the U.S., although HLS usually start acquiring Spanish at home, they become gradually dominant in English, among other factors, as a result of formal education in English (Campbell and Christian 2003; Silva- Corvalán 1994: 9); the higher prestige of English in most bilingual communities in the U.S. (Silva-Corvalán 1994: 9; Valdés and Geoffrion-Vinci 1998: 476), contact with English-speaking children on the playground or in the streets (Sankoff et al. 1997); and exposure to English through television and other media. Since knowledge of register is learned from interaction with members of one's speech community, a narrowing of the stylistic range might be expected. At the same time, as the linguistic repertoire in English expands to include an increasing number of domains, the functional use of the home-based language is restricted to fewer domains, until it is ultimately limited, in many cases, to the home and family domain. A family, homebound language is characterized by a casual, conversational speech style, used with familiar interlocutors to a restricted set of topics focused on everyday life. For most HLS, Spanish is mainly used with family and close friends as an oral informal variety. On the other hand, English is the language for most social interactions as well as the language in which they have been accustomed to think and learn in school (Lipski 1996; Silva-Corvalán 1991, 1994; Valdés 1992, 1995). 3 In view of this subordination of Spanish in relation to English and the subsequent restriction of usage to the home and family domain, it may be expected that the language spoken by HLS has become monostylistic (Dressler 1982, 1991). However, research has also shown that even in languages facing foreseeable extinction, the fundamental principle of intra-speaker variation still applies (Dorian 1994). Because there has been little research on the register and style variation of Spanish as a heritage language in the U.S., it is not known whether there is register variation or what the extent of that variation may be. The main goal of this dissertation is to determine whether Spanish as a non-dominant language spoken by HLS is consistent with the fundamental notion that there “are no single-style speakers” (Labov 1970: 19), or whether a lack of stylistic variation across registers occurs which would justify the term ‘monostylistic’ for these speakers. It has been observed that some of the phenomena characteristic of bilingualism and multilingualism, such as simplification, overgeneralization, and transfer are actually attested across different situations of linguistic stress (Silva- Corvalán 1991, 1994). This observation has motivated the proposal of a number of corresponding theories (theory of transfer, theory of simplification, etc.) valid for a variety of language situations including second language acquisition. However, even though there is a vast body of literature tackling many aspects of second language acquisition, there are few studies on register and style variation in learners of Spanish as a second language. This dissertation also examines SLS oral production 4 across different registers because Spanish is their non-dominant language, as it is for HLS. Although for both groups of speakers Spanish is non-dominant in relation to English, SLS and HLS have acquired Spanish under different circumstances. Unlike HLS, SLS have acquired Spanish almost entirely through formal exposure to the language in the classroom. Therefore, their stylistic range may be narrowed to the academic register characteristic of classroom interactions. A lack of fluency may be expected in casual registers and conversational speech styles; thus, their register domain in Spanish may be very different from HLS since SLS do not normally use the language with family and close friends as an oral informal variety. Thus, the purpose of including SLS data in this dissertation is twofold: (1) determine whether Spanish as a non-dominant language in second language learners shows variation across registers and (2) compare potential patterns of variation between SLS and HLS to examine the impact that different acquisition circumstances may have on the register range of these two groups. 1.3. Working assumptions: Register and style It is not easy to define such concepts as register or style. Biber (1995), for example, has pointed out the conceptual confusion of the term register because of the variety of ways that it has been used by different researchers from various perspectives. Furthermore, the terms register and style have sometimes been used interchangeably by different researchers to refer to any variety of language 5 associated to situational uses. Therefore, it is important to make explicit the working assumptions that will be relevant for my study of register and style variation. The term register has been used as a cover term to refer to language varieties associated with particular social situations. For example, conversations, political speeches, or university lectures are all instances of spoken registers (Finegan 2004: 19). Register is not defined in linguistic terms but rather in terms of situational parameters. Following this usage of the term, in my dissertation, register will refer to varieties associated to particular situations of use (Biber 1995; Ferguson 1994; Finegan 2004). Specifically, three spoken registers will be analyzed: conversations, interviews, and class presentations. Although registers are defined in nonlinguistic terms, usually registers have characteristic linguistic features. For example, private conversations are characterized by the use of colloquial items and punctors, among other features (Brown and Yule 1983: 15-16). In this type of involved production, interlocutors are able to rely more on the situational context and their shared background and often use what Biber (1988) and Finegan and Biber (1994, 2001) call ‘economy’ variants. On the other hand, academic presentations are informational in purpose and are typically addressed to non-interactive audiences; they are characterized by ‘elaboration’ features rather than ‘economy’ features (Biber 1988; Finegan and Biber 1994). Some of the features that are typically found in academic presentations are the use of formal and technical vocabulary, high lexical diversity, and more complex grammatical structures (Walters 1994). 6 In general, all languages can be viewed as a continuum of registers including oral and written forms and ranging from formal to more informal or casual varieties. The Dictionary of Language Teaching and Applied Linguistics (Richards et al. 1997: 144) defines formal speech as “the type of speech used in situations when the speaker is very careful about pronunciation and choice of words and sentence structure. This type of speech may be used, for example, at official functions, and in debates and ceremonies”. On the other hand, a prototypical informal speech would be produced in a relaxed conversation among close friends or family members. In my study, the three registers that will be analyzed, namely oral class presentations, interviews, and conversations range in a scale from more to less formal. That is, a class presentation is generally a more formal type of situation than a relaxed interview, whereas conversations among friends are typically very informal 1 . The term style refers to intra-speaker variation. Stylistic variation results from the fact that different people express themselves in different ways, and that the same person may express the same idea quite differently when addressing different audiences, using different modalities, or tackling different tasks (Bell 1984, 1997; Rickford and McNair-Knox 1994). The term style has been used by some researchers as a cover term similar to register (e.g. Bell 1984). However, in this dissertation, I focus on one aspect or dimension of style: formality. Accordingly, I will use the terms register and style distinctly in my dissertation: The term register is used as a broader term to refer to 1 The methodology section (section 5 of chapter 1) describes in detail the characteristics of each of these three registers. 7 the language variety associated with a given situation of use (i.e. conversation, interview, presentation), whereas the term style is used in a narrower sense to refer to intra-speaker variation associated with varying levels of formality within each register. Despite this terminological distinction, it is nevertheless difficult to separate clearly register and style in my study as style relates to the varying level of formality of each of the different registers. Therefore, on a scale of formality, the three registers analyzed in this dissertation correspond to three different styles, which can be visualized as follows: Registers: Presentation Interview Conversation Styles: + formal/ -informal +/- formal -formal/+informal Figure 1. Register and style relationship Therefore, register and style in this dissertation become almost synonymous as presentations are formal, interviews can be considered semi-formal, and conversations are informal. Even though from the point of view of the methodology of this dissertation, it is difficult to separate register and style, there are conceptual motivations for doing so; namely, whereas register is a broader concept that refers to the language used in specific situations, style is used in a narrower sense to refer to the level of formality that may exist not only across registers but also within specific registers. 8 2. Research questions The main question that my dissertation addresses is whether the Spanish spoken by HLS shows variation according to the situation of use. In order to examine this central question, I address a number of related questions explicitly as follows: 2.1. Linguistic features across registers: How do HLS and SLS use specific linguistic features in each of the social situations studied? Do these linguistic features form patterns shared by the speakers in the sample? 2.2. Comparison between groups of speakers: How does the range of register and style variation of HLS compare to that of SLS? If SLS and HLS differ in their range of registers and styles, what does this tell us about how non-dominant languages are acquired? 3. Literature Review and Conceptual Framework Register and style variation have been explored from a variety of perspectives including linguistics, sociology, and anthropology. There have been studies concentrating on single registers (e.g. Bell 1990; Ferguson 1964, 1983), as well as comparing across registers (e.g. Biber 1989, Halliday and Hasan 1989). While the majority of studies in this area focus on English-language registers, there have also been some cross-cultural studies comparing different registers (Biber 1995). However, relatively few studies have explored register variation in bilingual varieties or varieties in situations of language contact. This section offers a brief review of 9 some key studies from both register variation and language contact, which are relevant in order to contextualize this study of register and style variation in bilingual HLS as well as frame and substantiate the theoretical framework within which my study will be developed. Interest in register variation grew from studies, especially the social dialectology studies in the 1960s, which examined linguistic variation across groups of speakers as wells as across different communicative and social situations (e.g. Ferguson 1964, Gumperz 1964, Hymes 1964). The study of linguistic and social variation has been very productive from the 1960s and 1970s onwards; however, as Eckert and Rickford point out (2001: 2), research on intra-speaker variation has been less continuous. Yet many researchers have underscored the importance of stylistic variation to the understanding of language change. Among these, Labov’s influential work on intra-speaker variation has focused on the influence of attention paid to speech in style shifting. Labov’s (1966) New York City study demonstrates that stylistic variation is linked to the socioeconomic status of speakers. Labov talks about the importance of prestige and stigma and places each individual’s stylistic range in relation to the continuum between the ‘prestigious’ end and the ‘stigmatized’ end. According to Labov (1972), attention paid to speech is essential in stylistic variation; he views the prestigious end of the speaker’s range closer to a careful, more formal speech and, on the opposite end, the more careless or casual speech. However, for him, the central problem in stylistic analysis is separating the casual and careful speech within the 10 context of the sociolinguistic interview. Accordingly, Labov designed the sociolinguistic interview to elicit as wide a range of a speaker’s style as possible, manipulating the topics on the assumption that some topics will focus interviewees’ attention on their speech while others will divert them from it. Labov’s (1984) work on sociolinguistic methods presents ten principles governing sociolinguistic interviews. A classic example of his interview design is the danger of death question where Labov (1966) asked his speakers whether they had ever feared for their lives. The aim of this question is to elicit excited or emotional responses because people tend to speak more naturally if they become emotionally involved in what they are saying. This question created great emotional involvement and led speakers to the narration of personal experiences, which diminished the effects of audio-monitoring. In Labov’s work, the organization of contextual styles along the axis of attention paid to speech is a way to organize and use intra-speaker variation that may occur within the limits of the sociolinguistic interview. While recognizing the major influence of Labov’s work for the study of stylistic variation, many researchers have also criticized the focus of his work on attention paid to speech as the sole correlative of style shift (Bell 1984: 149-150; Milroy 1987: 172-183). For example, Eckert (2001: 122) points out that attention paid to speech is related to many other issues, among others, the genre, the topic, and the interlocutors. It is not clear how the emergence of the vernacular (i.e. the most natural speech) is due to lack of attention and not to other aspects of the interaction. With these considerations in mind, I believe that attention paid to speech is one of 11 many factors that may affect one’s speech. I consider attention paid to speech in this dissertation as an aspect that may vary depending on the characteristics of the situation; for example, in academic presentations, speakers are expected to have a greater degree of self-monitoring than in a casual conversation with friends; however, the potential variation between these two situations will respond not only to attention but also to other factors, including the audience. Bell’s 1984 seminal paper launched the focus on the audience to explain intra-speaker variation. Bell’s research has particularly focused on stylistic variation as a product of audience design. Bell (1984, 2001) defined style as “what an individual speaker does with a language in relation to other people” (2001: 141) 2 . He argued that intra-speaker variation reflects inter-speaker variation and that style-shift can be explained as the accommodation of speakers’ speech to that of their audience. Bell considered that the audience is not only the actual addressee but also third persons such as auditors and overhearers. Bell’s (1984) model of style drew attention not only to the responsive dimension of style but also to an initiative dimension, that is, style shift may itself initiate a shift in the situation rather than being derived from such a change. In this context, Bell proposes that initiative style shifts reflect the linguistic features associated with a reference group. Therefore, referees are a key component of Bell’s audience design even though not physically present as audience, since they are present in the mind of the speaker and thus affect his/her speech. 2 Bell (1984) uses the term ‘style’ as a cover term in the same way I use the term ‘register’ to refer to any variety associated with specific situations. My use of the term ‘style’ differs from Bell’s in that I use this term to refer to degrees of formality. 12 Bell (1984: 178-181) also recognizes that certain topics or settings prompt speakers to style-shift in a certain direction. However, for Bell, ‘non-audience’ factors that may affect a speaker’s style, such as topic and setting, still derive from audience design (1984: 181). Bell’s explanation is that speakers associate classes of topics or settings with classes of persons, and therefore, style shift occurs as the speaker accommodates her/his speech to that of the audience whom they associate with the topic or setting. Bell’s (1984) model of audience design has influenced many subsequent studies of register and style variation. However, just as Labov’s focus on attention paid to speech has been criticized, the major role given by Bell to the audience to explain intra-speaker variation has also been questioned. Undoubtedly, the audience has an important effect on one’s speech; however, many other factors may play an important role too and it is not clear to me why all other aspects (such as topic, setting, purpose, etc.) should derive from audience design as Bell claimed. Consequently, I regard audience in my dissertation as an important but not the sole factor that causes speakers to style shift in a given direction. The registers that I analyze in this study differ as far as the configuration and participation of the audience, yet other parameters are also considered to affect speaker’s production, such as the purpose of the interaction and the formality of the situation. Finegan and Biber (1994) claim that it is situational factors that structure register variation. Finegan and Biber’s model focuses on the broader ‘situation’, by which they refer to a variety of aspects that may influence the linguistic expression 13 of the speakers; among other factors, situations include setting, relative status of interlocutors, shared background, setting, topic, purpose (Finegan and Biber 2001: 239-240). In their view, situations are fundamental and register variation depends upon the “communicative constraints inherent in particular situations” (Finegan and Biber 1994: 339). Finegan and Biber (1994) argued as well that situations that are more literate call for more elaborate language, whereas less literate situations are characterized by language that is more economical. Biber (1988, 1989) analyzed a great number of linguistic features across various registers in order to arrive at a characterization of these registers. In particular, his research aimed at providing a comprehensive analysis of register variation for which he developed a multi-dimensional approach, originally for English but also for comparison across languages (e.g. Biber 1995). In his multi- dimensional model, Biber analyzed linguistic features that tend to co-occur or appear together in a text in terms of dimensions of variation. Dimensions are viewed as continuous scales of variation, defined by characteristic groupings of linguistic features and interpreted in terms of the communicative functions shared by each dimension. Five major dimensions are identified in Biber (1988) and each of them comprises a set of linguistic features. For instance, dimension 1 is labeled ‘informational versus involved production’ and it contains features which correlate with informational texts on the one hand (e.g. type/token ratio, attributive adjectives, word length), and involved texts on the other hand (e.g. contractions, discourse particles, hedges). Registers that co-occur with ‘informational’ features on dimension 14 1 include academic prose and prepared speeches. Conversely, face-to-face conversations have high frequencies of ‘involved’ features. Biber (1988) offered a comprehensive analysis of register variation including many different texts in English, both spoken and written, and considering a wide range of linguistic features. His model is particularly useful for this study of register and style variation in HLS since this dissertation examines different linguistic features that have been shown to vary in English and other languages (see Biber 1995 for a crosslinguistic perspective on register variation) and, therefore, may also show register variation in HLS’ production. However, unlike Biber, my focus is on a small number of spoken registers. I also take into account the situational parameters of planning and purpose (Finegan and Biber 1994, 2001) in the analysis of the registers that I include in this dissertation. Apart from taking into consideration the main issues examined in the studies of register and style variation discussed so far, I also need to take into account contact phenomena in order to answer the main question that I address in my dissertation, namely whether HLS show variation across registers. In particular, the research on transfer of features from one language into another is important for my study of HLS variation since it has been shown that transfer is a common strategy used by bilinguals to cope with the task of using two different linguistic systems (Weinreich 1968; Silva-Corvalán 1998; Thomason and Kaufman 1988). Transfer is generally defined as the influence resulting from similarities and differences between two (or more) languages in contact (Odlin 1989: 27). Some 15 scholars consider that transfer may have occurred whenever one or more of the following phenomena are present in the data (Silva-Corvalán 1994; Weinreich 1968): (a) Borrowing: The replacement of a form in a language A from a language B, or the incorporation from language B into language A of a form previously absent in A. (b) Extension or reduction of function: The incorporation of the meaning of a form R from language B, which may be part of the meaning of a form P in language S, into another form, structurally similar to R, in system B. (c) Higher frequency of use of a form in language A, determined on the basis of a comparison with a more conservative internal community norm, in contexts where a partially corresponding form in language B is used either categorically or preferentially (Klein-Andreu 1986; Silva-Corvalán 1994). (d) Neglect or elimination: The loss of a form from language A that does not exist in the system of B. I include the analysis of a number of linguistic features, which may show influence from English in the direction of any of the four transfer effects mentioned above. The comparison of these features across registers has the purpose of examining whether there is variation in their use depending on the situation. The contact between English and Spanish in the U.S. accounts for many instances of transfer in the Spanish of heritage speakers such as borrowings and code-switches. Many of these linguistic features are not considered appropriate for situations in which the prestige or standard variety is required or expected; for instance, formal events, business interactions, or academic settings. The standard variety of Spanish taught in high schools and universities is typically a standard monolingual variety (for example, from Mexico or Spain). Therefore, 16 variation in the production and frequency of linguistic transfer from English may be expected depending on the register if heritage speakers of Spanish are not monostylistic. In order to investigate register and style variation in speakers of Spanish as a heritage language, as well as the comparison with Spanish as a second language, I use a framework that combines the main issues of the register and style variation models reviewed here; specifically, attention paid to speech (Labov 1972), audience design (Bell 1984), situational parameters (Finegan and Biber 1994), and the co- occurrence of linguistic features with specific registers (Biber 1988). Furthermore, my framework includes the insights from research on contact situations and examines the issue of transfer between Spanish and English in HLS and SLS. Table 1 illustrates the aspects included in the framework used in this dissertation for the study of variation in Spanish as a non-dominant language. PREVIOUS STUDIES THIS DISSERTATION Labov 1972 Attention paid to speech Bell 1984 Audience Finegan & Biber 1994 Situational parameters Attention paid to speech Factors that affect Audience intra-speaker Situational parameters variation (planning, purpose, topic, degree of formality) Biber 1988 Co-occurrence of linguistic features with specific registers Otheguy 1993 Silva-Corvalán 1994 Contact phenomena Variation reflected If the above in the frequency of factors change occurrence of (as in different linguistic features registers) and transfer depending on the register Table 1. Framework for the study of register variation 17 The top shaded area of table 1 shows the factors included in this framework that affect intra-speaker variation, as previous research has shown. A change in these factors will cause speakers to vary, provided that they are not monostylistic. In order to examine whether HLS and SLS show variation, these factors are manipulated so that the situations of data collection (the three registers analyzed in this dissertation: conversations, interviews, presentations) differ in the expected degree of attention paid to speech by the speakers, the composition of the audience, the possibility of careful production, the degree of formality, the purpose, and the topic. 3 4. Linguistic features correlated with register variation in heritage languages The main hypothesis of the present dissertation is that HLS will show variation across registers and this variation will be reflected in the use of particular linguistic features. SLS may also show register and style variation, which may differ in some ways from HLS’ variation since they have acquired the language under different circumstances. Taking into account the linguistic features analyzed in previous studies of register variation and the literature of Spanish in contact, I have identified a number of features relevant for the examination of register variation in HLS and SLS. The selection of linguistic features for this study is primarily based upon a subset of features that define Biber’s dimension 1 from his multidimensional model of register variation (1988, 1989). In Biber’s study, dimension 1 represents ‘involved 3 The methodology section (section 5) explains in more detail the characteristic parameters of the three registers analyzed in this dissertation in relation to the framework identified above. 18 versus informational production’ and it is characterized by two groups of linguistic features that co-occur with texts that are highly involved, affective and interactive on the one end (involved production) and highly informational and non-involved on the other end (informational production). The motivation for focusing on this particular dimension and for including some of its characteristic features is that the registers analyzed in this dissertation can also be interpreted as a continuous between the highly involved (a conversation between friends) on the one end of the scale, then the moderately informational and somewhat involved (an interview between speaker and researcher), and the highly informational and non-involved at the opposite end (an academic presentation). Among the features that co-occur with involved production, Biber (1988) finds discourse particles and contractions. Regarding informational production, some of the characteristic features include type/token ratio and attributive adjectives. These features, according to Finegan and Biber (1994, 2001) represent ‘economy’ and ‘elaboration’ respectively. For example, phonological contractions are an example of expressive economy, whereas type/token ratio is a measure of lexical elaboration (Finegan and Biber 2001: 246). These features (i.e. discourse particles, contractions, type/token ratio, attributive adjectives) are selected for the present study for two reasons: (1) They correspond to linguistic features that can be examined in Spanish, whereas other features from Biber’s dimension 1 (e.g. “that” deletion, “do” as pro-verb, pronoun “it”, “WH” questions and clauses) are specific to the English language and do not correspond to Spanish uses. (2) These linguistic 19 features have also been studied in the Spanish literature (e.g. Cortés 1991; Martín Zorraquino and Montolío 1998) and are of interest regarding Spanish as a heritage language in the U.S. For example, Sánchez (1983) talks about the production of contractions in casual registers by Spanish speakers in the U.S. and Valdés and Geoffrion Vinci (1998) discuss the use of some discourse particles and attributive adjectives in academic presentations by HLS. Apart from discourse particles, contractions, type/token ratio and attributive adjectives, which Biber (1988) analyzes in his study of variation, this dissertation also includes a set of linguistic features that relate to the issue of Spanish in contact with English in the U.S., as it was explained in the framework presented above. Various studies have pointed out the occurrence of lexical transfer from English in Spanish varieties in the U.S. (Otheguy 1989; Lipski 1994, 2004; Silva-Corvalán 1994). This dissertation analyzes a number of lexical features that have been shown to characterize informal registers in the speech of HLS. These include the use of borrowings and code-switches (Otheguy 1989, Silva-Corvalán 1994). Additionally, the use of colloquialisms and other informal items has been reported in the speech of HLS in formal situations in which these uses are not expected (Achugar 2003; Hidalgo 1987; Sánchez 1993; Valdés 1995; Valdés and Geoffrion-Vinci 1998). For the purpose of analyzing whether there is variation in the choice of lexical items, this dissertation also includes the study of formal, informal, and technical vocabulary. Table 2 summarizes the linguistic features that are analyzed in this dissertation: 20 Table 2. Linguistic features included in this dissertation Related to the expected outcomes of the analysis of all the linguistic features examined, I propose several sub-hypotheses that are tested in subsequent chapters of this dissertation with the purpose of analyzing variation in the production of HLS and SLS across registers. The following paragraphs describe in more detail each of the linguistic features and justify their inclusion in this dissertation. 4.1. Discourse particles Chapter 2 contains the analysis of a number of discourse particles including connectors (e.g. como ‘as’/‘like’; so, así que ‘so’, and entonces ‘so’/‘then’) and punctors (e.g. I mean) across situations. The use of discourse particles has been shown to vary across registers and styles (e.g. Biber 1988, 1995; Cortés 1991; Martín Zorraquino and Montolío 1998. These studies are discussed in more detail in chapter 2). Therefore, I include the study of discourse particles produced by HLS in different situations of use with various degrees of formality in order to examine register variation in the use of discourse particles in Spanish as a heritage language. Furthermore, the analysis of the production of SLS can show whether language learners use these discourse particles in the same way as HLS across registers. LINGUISTIC FEATURES DESCRIPTION Discourse particles Contractions Linguistic features correlated with “involved production” (Economy features) Type/token Ratio Attributive adjectives Linguistic features correlated with “informational production” (Elaboration features) Lexical transfer Lexical choices Linguistic features of Spanish in contact with English that are expected to vary across registers 21 In chapter 2, I include a number of particles that I consider especially relevant for the study of register variation in the context of Spanish in contact with English. Specifically, the connectors como ‘as’/‘like’; so, así que ‘so’; and entonces ‘so’/‘then’; and the fillers or punctors “I mean”, “you know”, “I don’t know” in English and este ‘uh’/‘um’, sabes ‘you know’, no sé ‘I don’t know’ in Spanish. In the case of the word como ‘as’/‘like’, I hypothesize that in the Spanish of HLS it has some uses not found in monolingual varieties, namely as quotative marker and punctor (Sánchez-Muñoz, forthcoming). These uses are similar to the functions of quotative and punctor ‘like’ in American English. Furthermore, como may have a wide variety of functions in Spanish and it may occur in both formal and informal oral registers. In particular, the use of como as a marker of ‘looseness’ or ‘approximation’ is considered characteristic of informal, colloquial registers (Moreno Ayora 1991; Santos Río 2003). I compare the use of approximative como by HLS and SLS across situations in order to test the claims put forth in the Spanish literature. The second set of particles (so, entonces ‘then’/‘so’ and así que ‘so’) is also interesting from the perspective of the study of Spanish as a heritage language. Several studies have shown that bilingual varieties in contact with English incorporate English-origin so in their oral discourse, adapting it to Spanish phonology (it is pronounced [so], without gliding) (Aaron 2004; Lipski 1994; Sankoff et al. 1997; Silva-Corvalán 1994). English so has two possible equivalents in Spanish: así que ‘so’ and entonces ‘so’/‘then’ and some studies have investigated 22 whether so is replacing the Spanish forms así que or entonces (e.g. Aaron 2004). I investigate the use of these three discourse particles by HLS to see whether there is variation depending on register. Additionally, I analyze the production of SLS in order to find out whether SLS produce so along with así que and entonces within Spanish discourse and how their variation compares to HLS. The third set of particles that are included for their particular interest in the investigation of the Spanish of HLS and SLS consists of fillers or punctors. The reason for the inclusion of punctors is two-fold: (1) since punctors are characteristic of informal, casual production (Brown and Yule 1983; Vincent and Sankoff 1992), I ask whether HLS use punctors more frequently in conversations with friends than in more formal situations and also whether SLS produce punctors at all; and (2) whether the use of English-origin punctors in Spanish varies across registers. With respect to discourse particles, I hypothesize that HLS and SLS will vary their use of discourse particles and that variation will be reflected across registers in the following specific ways: 1. More informal situations will show an increase of the non-canonical uses of como ‘as’/‘like’ in Spanish, namely como as a quotative and como as a punctor. Furthermore, since como used with the value of approximation is more frequent in casual registers (Moreno Ayora 1991; Santos Río 2003), I expect the approximative use of como to be higher the less formal the situation is. 2. The use of English-origin so in Spanish will vary across registers, being less frequent the higher the level of formality. I expect that since así que and 23 entonces are not borrowed from English, these will be used more frequently in the presentations and interviews, and occurrences of so will be higher in the conversations. 3. Speakers are expected to produce an overall lower number of punctors in the presentations. Furthermore, I hypothesize that Spanish-origin punctors este, sabes, o sea will be used more often in the interview situation, while they may resort to English-origin punctors (“you know”, “I don’t know”, “I mean”) more often in the conversation situation. 4. SLS are expected to produce an overall smaller amount of discourse particles, especially significantly fewer punctors when compared to HLS since SLS’ contact with oral informal varieties has been more limited. 4.2. Contractions Also included in chapter 2 of this dissertation is the analysis of the contraction of certain words such as prepositions and conjunctions (para ‘for/so that’ and pues ‘so’) and verbs (estar ‘to be’). Finegan and Biber (1994: 322) state that contractions are an instance of economy in language and they find that this feature usually correlates with situations that are less planned, more affective and less informational, and with shared background between interlocutors. Some researchers have also suggested the loss or lessening of part of phonetic body in high frequency words such as prepositions and conjunctions in informal registers (Vincent 1993; Vincent and Sankoff 1992). With this in mind, I include the study of the pronunciation of the particles pues ‘so’/‘well’ and para ‘so that’/‘for’, which are 24 contracted and realized as pos and pa, as well as the verb estar ‘to be’, also contracted in the various forms of its conjugation (e.g. toy, tá, or tás instead of estoy ‘I am’, está ‘he/she is’ or estás ‘you are’) (Hidalgo 1987; Sánchez 1983; Valdés 1988). The hypotheses tested in relation to the use of contractions are that: 1. HLS will phonetically reduce pues to pos, para to pa, and the verb estar ‘to be’ in the different forms of its conjugation (e.g. tá instead of está ‘he/she is’) more frequently in relaxed casual situations (Hidalgo 1987; Valdés 1988); therefore, I expect that the conversation situation will show a higher number of contractions. 2. SLS will not produce contractions but if they do, these will be significantly less frequent than in the case of HLS. The reason for this hypothesis is that SLS have been exposed to formal instruction of Spanish and they have not been exposed to informal oral varieties as HLS have. 4.3. Lexical Choices Chapter 3 deals with the analysis of a number of lexical features that I consider relevant for the study of variation in Spanish as a heritage and second language. These include the use of formal, informal, and technical vocabulary and lexical transfer. Additionally, I analyze the use of attributive adjectives and the feature of type/token ratio. 4.3.1. Formal, informal, and technical vocabulary It is expected that in academic registers, such as class presentations, speakers will make use of a formal and technical vocabulary in both English and Spanish 25 (Achugar 2003; Biber 1988; Solomon and Rhodes 1995). Conversely, casual conversations are typically characterized by informal lexicon, such as the use of idioms and slang items (Celce-Murcia and Larsen-Freeman 1983). Despite these expectations, some studies have pointed out the use of colloquial words and expressions not considered appropriate for formal or academic situations in the Spanish produced by HLS (Achugar 2003; Hidalgo 1987; Sánchez 1993; Valdés 1995). According to Valdés and Geoffrion-Vinci (1998), the use of ‘colloquialisms’ in academic registers indicates that speakers have not developed formal registers or that they are still in the process of acquiring the ‘formal’ variety appropriate for academic settings. With these considerations in mind, I include the analysis of formal, technical, and colloquial vocabulary and examine the results from the data of HLS and SLS across registers. The hypothesis tested is that more informal and affective situations, such as conversations with friends, will have a higher frequency of colloquial items than other more formal situations (i.e. interviews and presentations). In addition, in academic registers, HLS and SLS are expected to produce a high frequency of formal and technical vocabulary (Valdés and Geoffrion-Vinci 1998). 4.3.2. Lexical transfer Chapter 3 also examines some cases of lexical transfer in Spanish by HLS and SLS in order to compare these across situations of use. These are interesting from the perspective of language contact since transfer is expected at the lexical level between two languages that co-exist in a given community, as is the case of Spanish 26 as a heritage language in the U.S. (numerous studies have investigated lexical transfer between Spanish and English, for example, Otheguy et al. 1989, 1993; Silva- Corvalán 1994; Valdés and Geoffrion-Vinci 1998). Additionally, transfer has also been shown to occur at the lexical level in situations of adult acquisition of a second language (Odlin 1989), which is the case of SLS in this study. The cases of transfer that are taken into account in this dissertation include borrowings (subcategorized into loans and calques), code-switches to English within Spanish discourse, and ‘lexical creations’ in Spanish based on the form of the English word (e.g. investan ‘[they] invest’ instead of invierten). Thus, in chapter 3, I examine these instances of lexical transfer in order to find out whether speakers vary across situations of use, and whether HLS and SLS produce similar types of lexical transfer. In general, I expect more informal and affective situations to have a higher frequency of borrowed items and lexical creations as well as more code-switching. On the other hand, in academic and more formal registers in Spanish, HLS and SLS may not use English words (Valdés and Geoffrion-Vinci 1998). 4.3.3. Attributive adjectives According to some researchers (e.g. Finegan and Biber 1994, 2001; Hudson 1994), more elaborate and literate registers are characterized by the use of a high frequency of attributive adjectives. Thus, this may be a linguistic feature that may be evidence of variation across the three registers analyzed in this dissertation. 27 Related to the use of attributive adjectives, the hypothesis tested is that speakers will display a quantitave difference across situations in the following manner: a higher frequency of attributive adjectives will correlate with more formal and informational situations. On the other hand, affective and informal situations will have a lower frequency of attributive adjectives. 4.3.4. Type/token ratio Type/token ratio represents lexical diversity and elaboration (Finegan and Biber 2001: 246). This feature is included in the analysis of register variation in order to investigate whether HLS and SLS have different degrees of lexical diversity depending on the situation. Other studies have found a correlation between type/token ratio and different registers; for example, Valdés and Geoffrion-Vinci (1998) found that bilinguals (HLS) displayed greater lexical diversity in academic argumentations than in the informal narration of personal experiences. I will examine type/token ratio of conversations, interviews, and presentations. As in the case of attributive adjectives, the hypothesis related to type/token ratio is that there will be higher lexical diversity (reflected in a higher type/token ratio) in more formal and informational situations, whereas affective and informal situations will have a lower type/token ratio. 5. Methodology In this section, I explain the methodology developed for the study of register variation in the oral discourse of HLS and SLS, including the following aspects: (1) 28 sampling methods and characteristics of the speakers; (2) the situations of use in which data were collected; (3) the data collection procedure and coding process; and (4) the statistical analysis. 5.1. Sampling methods and characteristics of the speakers All the participants in this study are university students who, at the time of data collection, were taking Spanish courses as part of their undergraduate major or minor. The factors that were taken into account in the selection of speakers included: home language (Spanish for HLS and English for SLS); the level of education of the speakers in Spanish (i.e. how many years of formal education before and after high school); (3) the geographical origin of the heritage language (i.e. the country of origin of the speakers’ parents for HLS). With these aspects in mind, I conducted interviews to select the participants. The speakers reported in this dissertation include eight HLS and eight SLS; thus, a total of sixteen college students who were taking Spanish courses as part of their major or minor. Appendix A contains a list of the speakers together with their characteristics and the speaker coding used in this dissertation. Prior to the collection of data, all participants were asked to fill out a biographical and language background questionnaire to ensure the homogeneity of the sample. There was one questionnaire for the HLS group and one for the SLS group, which can be seen in appendices B and C. These questionnaires were designed to obtain personal information from the speakers; information about the amount and type of formal education in Spanish and about their patterns of use of 29 Spanish (e.g. with whom the speaker most frequently interacts in Spanish; see appendices B and C). Only those speakers who had a comparable amount of formal education as well as informal contact with the language were selected for this study; thus, the subjects’ level of Spanish proficiency is fairly homogeneous within the HLS group on the one hand, and the SLS group on the other hand. After selecting a uniform sample of speakers, I set up individual data collecting sessions with each speaker. The following are the characteristics of the speakers from whom I have collected the data: 1- Home language: HLS have Spanish as their home language. All HLS in this study are second generation Mexican-American, that is, their parents were born in Mexico and immigrated to the U.S. For the most part, the parents of the HLS of this study came to this country when they were in their twenties. HLS were born and raised in the U.S. and they were exposed to Spanish from an early age (as the language of parents and grandparents who frequently live in the same household). They use Spanish with relatives and close friends and in various situations that require family interactions (e.g. birthday parties, Christmas). However, English is the language HLS use on a daily basis for most interactions (e.g. with friends, at school, and at work). On the other hand, SLS have English as their home language and all SLS come from monolingual English-speaking households. 30 2- Formal education in Spanish: All speakers, both in the HLS and SLS group, were taking advanced Spanish courses in college at the time of data collection for this study. Speakers were majoring or minoring in Spanish; thus taking the same courses in the Department of Spanish (see appendix D). Additionally, all participants had studied at least four semesters of Spanish at high school. All the participants in this study went to schools in which the language of education was English and none of the HLS attended bilingual schools. The Spanish classes they took while in high school were part of the foreign language requirement, and no HLS participated in Spanish for native speakers. Hence, the type and amount of Spanish classes that all HLS and SLS took prior to college were similar. Consequently, the difference between HLS and SLS lies in the level of informal contact with the language. Whereas HLS have been exposed to informal varieties at home, SLS have acquired their knowledge of Spanish mainly through formal channels without much contact with the casual varieties. 5.2. Situations of data collection This dissertation examines three different communicative situations, academic and non-academic, placed at different points in the speakers’ register continuum (Biber 1989, 1995; Valdés and Geoffrion-Vinci 1998). These situations range on a scale from more involved and casual to more informational and formal: (1) class oral presentation; (2) interview; (3) conversation. 31 The parameters that characterize these three situations are defined by various factors that affect register variation. These factors are a combination of the different aspects that some of the key studies on register variation have examined (Labov 1972; Bell 1984; Finegan and Biber 1994), as I explained in section 3. Specifically, the three registers analyzed in this dissertation differ in the expected degree of attention paid to speech by the speakers, the composition of the audience, the possibility of careful production, the degree of formality, the purpose, and the topic, as illustrated in table 3. REGISTERS IN THIS DISSERTATION FACTORS THAT AFFECT REGISTER VARIATION Presentation Interview Conversation Attention paid to speech High degree of attention paid to speech Moderate degree of attention paid to speech Low degree of attention paid to speech Audience Professor and fellow students Large audience Non-interactive The researcher Interactive A close friend Interactive Situational parameters Planned, careful production Informational and non-involved Academic topics Formal On-line production Informational and somewhat involved Nonpersonal topics Semi-formal On-line production Affective purpose, highly involved Personal topics Informal Table 3. Characteristics of the registers analyzed in this dissertation. The characteristics summarized in table 3 are described in more detail in the following paragraphs for each of the situations analyzed in this dissertation: 1- Class oral presentations: Class presentations are a type of academic register in which the speaker is required to address an audience with the primary purpose to inform. Presentations 32 are normally planned and offer the possibility of careful production. Furthermore, in this academic setting, the student is not only informing the audience (the fellow students) but is also expected to display his/her competence, ability, and mastery of the subject (Valdés and Geoffrion-Vinci 1998: 447). I consider this register the most formal of the situations examined in this dissertation. 2- Interviews: Interviews are produced on-line and face to face with interaction possible between interlocutors. This register is primarily informational in purpose and only secondarily inter-personal (Finegan and Biber 2001: 249). The interview is considered the ‘semi-formal’ situation of this study: It is a type of register that is more formal than a conversation between friends but, at the same time, less formal than the presentation situation. In the interview, there is only one interlocutor and not a large passive audience; additionally, the interviewer (the author of this dissertation) tried to make the speaker feel comfortable and relaxed. The interviews were designed to address a variety of topics that did not include personal topics or the narration of personal experiences. Instead, more abstract, “serious” subjects and controversial topics were discussed (such as politics, the war in Iraq, gay marriage, illegal immigration in the U.S., etc.). These were thought to elicit a type of register more formal than a conversation between friends but less formal than the presentations. 33 3- Conversations: Conversations are also produced on-line and face to face with more interaction between interlocutors. This register typically has affective purposes with greater involvement of the participants and it is concerned with inter-personal relations as well as the exchange of personal information (Finegan and Biber 2001: 248). The conversations are the most informal situation of the present study; they involved two close friends chatting casually. Although this situation was free, speakers were given some general guidelines about the topics that they could talk about during the recording. These included high school anecdotes, opinions on school or common current courses, weekend and leisure activities, describing one’s family and relationship with parents and other relatives, plans for the future, etc. 5.3. Data collection procedures and materials After completing the language background questionnaire, each speaker was recorded at three different times, according to the situations described above. A DM-1 Olympus digital recorder with a lapel microphone was used for the audio recordings. The first situation of data collection was the interview, which had an average length of fifty minutes per speaker. Each HLS and SLS was also recorded during a class presentation. This academic presentation was not arranged specifically for the purpose of this study; rather, I recorded the oral presentations that the participants of this study were scheduled to give for the Spanish courses they were enrolled in at the time of the data collection. Since the speakers were enrolled in different advanced Spanish courses, the presentations vary in topic and length (for a 34 detailed discussion of the courses in the Spanish department see Appendix D). The average length of the presentations is fifteen minutes. Finally, conversations were recorded by the speakers themselves without the researcher being present during the recordings so as to minimize any effect of the researcher’s presence on the speakers’ production. The students recorded themselves conversing with friends, mainly with classmates who were also participants in this study. The average length of the conversations is thirty-five minutes. Thus, a total of approximately 1600 minutes or more than twenty-six hours of audio-recordings are analyzed in this dissertation. These data were transcribed by two native speakers of Spanish: the researcher and an independent transcriber who served as inter-rater for the purpose of ensuring reliability. There was a high percentage of agreement of 98.9% 4 . 5.4. Statistical analysis Every instance of the linguistic features analyzed in this dissertation was manually counted for each speaker in each of the recordings. However, since the amount of recorded material varied for each situation, in analyzing and reporting the results of this dissertation, I normalized the use of each linguistic feature to 15 minutes of audio-recorded material per speaker, which corresponds to approximately 1,000 words. Consequently, each of the lexical features analyzed can be compared across speakers and across different situations. 4 This percentage is established according to the total number of words transcribed. 35 Finally, the data obtained from the transcriptional examination of each of the linguistic features were analyzed using Chi-square and t-tests for statistical significance. The statistical analyses were performed with the Statistical Package for the Social Sciences (SPSS, version 11.5 for Windows). The graphs and figures are designed using Microsoft Excel (version of Microsoft Excel 2002 for Windows) as well as the SPSS program. 36 Chapter 2. Discourse particles and contractions 1. Introduction There have been numerous studies on discourse markers in Spanish (e.g. Martín Zorraquino and Montolío 1998; Portolés 1998; Martín Zorraquino and Portolés 1999), including studies that deal with variation in the use of discourse markers (e.g. Cortés 1998). Similarly, there have been several studies indicating the phonological reductions -or contractions- of certain words in informal and uneducated varieties of Spanish (e.g. Lance 1975; Sánchez 1983; Valdés 1988). However, there has been relatively little attention paid to register variation in the use of discourse markers and contractions in the Spanish spoken by heritage speakers in the U.S. The purpose of this chapter is to examine these two linguistic features across different registers in order to examine whether there are register and style differences reflected in the Spanish of HLS. Additionally, this chapter also examines data from SLS, looking at the use of discourse markers and whether SLS produce contractions, with the purpose of comparing with the production of HLS in similar registers. In this chapter, I analyze the use of discourse particles; namely, connectors, pragmatic markers and punctors. The term ‘discourse particle’ will be used henceforth as a cover term for the items analyzed here. The use of discourse particles has been shown to vary across registers and styles (Biber 1989, 1995; Cortés 1991; Martín Zorraquino and Montolío 1998). For example, Cortés (1998) points out that 37 the choice of discourse particles may be conditioned by stylistic factors, that is, they depend on the formality of the register or on the type of text typology, such as narrative texts or descriptive texts (Cortés 1998: 152). In addition, Sankoff et al. (1997) point out that discourse particles are of particular interest because they constitute an aspect of the language not explicitly taught in school and, for this reason, they may be an accurate indicator of the speaker’s linguistic ability in different registers. Biber (1989) includes discourse particles in the first dimension of his five- dimensional model of variation. In his model, linguistic coccurrence is analyzed in terms of underlying dimensions of variation. Dimensions are continuous scales of variation that consist of groups of features that co-occur frequently in a text. He calls the first dimension “involved versus informational production” (p.10), that is, dimension one is a continuum from highly informational, non-interactive texts to highly affective and interactional texts. Dimension one consists of two groupings of features that represent sets of features that occur in a complementary way. For example, some features that coccur with informational density include high type/token ratio and attributive adjectives. On the other hand, features such as contractions and discourse particles are among the features that co-occur with ‘involved’ rather than ‘informational’ production. Based on the findings of these and similar studies, I believe that the examination of discourse particles produced by HLS and SLS in different situations of use with various degrees of formality is useful in the analysis of register/style 38 variation in Spanish as a non-dominant language (both in heritage and second language contexts). Depending on the register, speakers’ production of discourse particles may vary both qualitatively (the type of discourse particle) and quantitatively (the frequency of use). For example, HLS may produce more discourse particles in involved interactions, such as conversations, and fewer in informational registers such as academic presentations. Additionally, SLS may differ from HLS both qualitatively and quantitatively in the production of discourse markers. This chapter also deals with the contraction of some words including prepositions, conjunctions, and verbs (e.g. toy instead of estoy ‘I am’). Finegan and Biber (1994: 322) state that contractions are an instance of economy (phonological in this case) in language and find that this feature usually correlates with situations that are less planned, more affective and less informational, and with shared background between interlocutors. Some researchers have also suggested the loss or lessening of part of phonetic body in high frequency words such as prepositions and conjunctions in informal registers (Vincent 1993; Vincent and Sankoff 1992). In this dissertation, the analysis of contractions across different situations is intended to examine whether these items are more frequently produced when the style is more informal. This chapter is organized as follows: The second section deals with the use of discourse particles. Here, I explain the meanings and uses of the discourse particles studied. The third section is concerned with contractions, specifically phonological reductions, and introduces the types that are analyzed. The fourth part of the chapter 39 presents the hypotheses related to the use of discourse particles and contractions in the speech of HLS and SLS. Then, results and a discussion of these follow in section five. Finally, conclusions are presented in section six. 2. Discourse Particles Discourse particles or markers have been defined differently by different researchers. Some researchers consider that discourse particles signal or mark relationships across utterances (Torres 2002: 65) and may have grammatical functions as well (Aaron 2004: 162). Other researchers include punctors and other semantically ‘empty’ items within the classification of discourse particles. According to Martín Zorraquino and Portolés (1999: 4057), discourse particles are defined as linguistic items with no syntactic function at the sentence level, but serve, according to their morphosyntactic, semantic and pragmatic properties, as a guide for the interpretation of utterances. The speaker adds these particles to reduce the cognitive effort required from the hearer to interpret the utterance, by signaling which inference reflects more accurately the speaker’s meaning. As early as in 1943, Gili Gaya pointed out that the choice of discourse particles may be contingent upon the type of register. He mentions that particles such as sin embargo ‘however’, no obstante ‘nevertheless’ or por consiguiente ‘consequently’ are more appropriate in written discourse, whereas other particles are more common in oral discourse, for example, pues ‘well’, así que ‘so’. More recent 40 studies have also shown that discourse particles vary across registers and styles (Cortés 1996, 1998; Martín Zorraquino 1998). In this chapter, I include the analysis of a number of discourse particles, which are shown in table 4. These include connectors (como ‘as/like’; so, así que ‘so’, and entonces ‘so/then’) and fillers or punctors (“I mean”, “you know”, “I don’t know” in English; este ‘uhm’, sabes ‘you know’, no sé ‘I don’t know’ in Spanish). The following paragraphs describe in more detail each of these discourse particles. Table 4. Discourse particles studied 2.1. Como (‘as/like’) The focus of this chapter is on the word como, which may have a wide variety of functions as an adverb, preposition or conjunction. Furthermore, as I will show later, the use of the word como by HLS seems to be calquing some functions from the equivalent ‘like’ in American English, which como does not traditionally have in monolingual varieties. 2.1.1. Classification of como in the Spanish literature The word como is one of the most productive in Spanish (Moreno Ayora 1991; Santos Río 2003). Furthermore, some authors point to an ongoing extension of Spanish English como ‘as’, ‘like’ así que ‘so’ entonces ‘so’, ‘then’ so no sé ‘I don’t know’ o sea ‘I mean’ este ‘uh’, ‘um’ I don’t know I mean you know 41 the functions of como (Moreno Ayora 1991: 8). Como has been classified as an adverb, conjunction or preposition; como is also used in different expressions in combination with other words such as que ‘that’ or si ‘if’ (Bosque and Demonte 1999; RAE 1992; Santos Río 2003): 1. As an adverb 5 , como may be used in the following ways: A- Como expresses the ‘way’, ‘manner’, ‘form’ or ‘procedure’ of something: (1) Hazlo como te digo (RAE 1992: 519): ‘Do it as (i.e. in the way that) I tell you.’ B- Como as a comparative conveys the idea of equivalence, similarity or likeliness (Pavón Lucero 1999): (2) Es rubio como el oro. (RAE 1992: 519) ‘He’s blond as gold.’ C- Como as exemplifier introduces an instance, it has the meaning of ‘such as’ or ‘like’. (3) Genios como Galileo o Cervantes han reconocido esta gran verdad. (Santos Río 2003: 257) ‘Geniuses such as Galileo or Cervantes have recognized this truth.’ D- Como may have the meaning of ‘rather’ or ‘approximately’ and it is often followed by a numeric expression: (4) Llevo aquí como una semana. (RAE 1992: 519) 5 The uses of cómo,’ how’, as an interrogative and exclamative adverb are not included in this analysis. The interrogative/exclamative cómo is stressed and is written with an accent mark, as opposed to como without an accent mark which performs a variety of functions as an adverb, preposition or conjunction (RAE 1992; Santos Río 2003). 42 ‘I’ve been here for about a week.’ (5) Tiene como 25 años [D,HLS,f19,C] 6 ‘She is around 25-years-old.’ Furthermore, como is used as a marker of ‘non-equivalence’, ‘looseness’ or ‘approximation’ in other contexts even without a following numeric expression (Santos Río 2003: 258), which is also a common use of English like (Andersen 2000, 2001; Schourup 1985): (6) like chilly enough to where they can possibly have a coat on (Schourup 1985: 38) According to some researchers (e.g. Santos Río 2003), this como is characteristic of informal, colloquial registers: (7) Es como demasiado formal. (Santos Río 2003: 258) ‘It’s kind of too formal’ (8) Y eso crea como sentimientos que sólo pueden tener confianza en su mismo grupo. [L,HLS,m19,I] ‘And that creates like feelings that they can only trust their own group.’ E- Como may be equivalent to según, ‘as’: (9) Esto fue lo que sucedió, como fácilmente puede probarse. (RAE 1992: 519) ‘This is what happened, as it can easily be proven.’ 6 Hereafter, the code included in square brackets indicates that the example is taken from my recorded data. The code stands for the following in this order: the speaker’s initial (D, W, C, L, J, G or S, A, Co, Ge, K, Cr, R, M), heritage language speaker (HLS) or second language speaker (SLS), sex (f= female, m= male), age, and situation of use (C= conversation, I= interview, P= presentation). 43 (10) La caridad como dice Fray Luis de Granda... (RAE 1992: 519) ‘Charity, as Fray Luis de Granda says...’ 2. As a conjunction: Como may be used as a causative connector with the meaning of ‘since’ or ‘because’ (Galán Rodríguez 1999: 3619). It often follows the causal conjunction porque (‘because’), thus acting as a redundant causative marker: (11) Como llueve tanto no podemos salir al cine. (Galán Rodríquez 1999: 3619) ‘Since it’s raining so much, we can’t go to the cinema.’ 3. As a preposition: As a preposition, como may precede an NP and it has the meaning of ‘acting as’, ‘in the position of”, ‘functioning as’ (RAE 1992: 519; Santos Río 2003: 263): (12) Asiste a la boda como testigo (RAE 1992: 519) ‘He’s attending the wedding as a witness.’ (13) Y Dios como una autoridad los destruyó. [L,HLS,m19,I] ‘And God as an authority destroyed them.’ 4. Expressions with como in combination with other words: Como que may express the cause or connection between two events: (14) Me parece muy caro. ¡Como que es de oro! (Galán Rodríguez 1999: 3619) ‘I think it’s very expensive. [because, as] it’s gold!’ Additionally, como can also have an approximative value followed by que, especially in varieties of Latin American Spanish. The phrase como que conveys an idea of hypothetical comparison or vagueness, similar in many ways to the meaning 44 of “as if”, or equivalent to expressions such as ‘to seem like’, ‘to feel like/as if’ or ‘give the impression that’ (Montolío 1999: 3680; Moreno Ayora 1991: 127; Santos Río 2003: 266). (15) Muchos lo ven como que hicieron eso [la guerra] para que puedan invertir más en sus negocios. [L,HLS,m19,I] ‘Many see it as if the did that [the war] so that they can invest more in their business.’ Como si (‘as if’) introduces a hypothetical comparison (Montolío 1999: 3679): (16) Se quieren como si fueran hermanos. (Montolío 1999: 3679) ‘They love each other as if they were brothers.’ (17) Es como si ellos no tuvieran derechos tampoco… Si fuera todo así pues nadie aquí tuviera derechos. [L,HLS,m19,I] ‘It’s as if they didn’t have rights. If everything was like that then nobody here would have any rights.’ 2.1.2. Particular uses of como by HLS Apart from the functions of como listed above, which are recognized in the Spanish literature, my data show two additional uses which are not to my knowledge discussed in Spanish grammars or studies on como in Spanish. These are como used as a quotative and as an empty punctor. 45 1. Quotative como: Como introduces reported speech or a direct quotation. There have been many studies of English like that focused on the quotative use of this marker to introduce reported speech (e.g. Blyth et al. 1990; Ferrara and Bell 1995; Mathis and Yule 1994; Romaine and Lange 1991; Yule and Mathis 1992): (18) She said, “what are you doing here?” And I’m like, “nothing much,” y’know. I explained the whole… weird story. And she’s like, “Um… Well, that’s cool.” (Romaine and Lange 1991: 227) The use of como as a quotative in Spanish has not been reported in the Spanish literature but it occurs in HLS discourse: (19) A: ¿Has visto tú rivalidad entre los negros y los latinos? ‘Have you seen conflicts between Blacks and Hispanics?’ L: Sí, como “oh, no hables con ellos” [L,HLS,m19,I] ‘Yes, like “oh, don’t talk to them”.’ (20) Me dijo que preguntó algo y ellos como “¿de qué estás hablando?” [C,HLS,f20,I] ‘She told me that she asked something and they like “what are you talking about?”.’ It is possible that the quotative use of como in HLS has been transferred from American English like. 46 2. Filler or punctor como As an empty punctor, como has no semantic content and is similar to the expletive use of like in American English. Punctor like is often considered ‘colloquial’ and even ‘vulgar’, just a meaningless interjection or hesitation device (OED 1989 VIII: 946; Wilson 1993). (21) He's, like, really friendly -- someone you can talk to. It was, like, getting pretty late but I didn't want to go home yet. [from Cambridge Dictionary of American English, Landau 2001] The two sentences in 21 illustrate, according to Landau (2001), the use of like as a punctor or empty filler. This use is considered non-standard and it is very frequent in casual conversation (Landau 2001) and adolescent speech (Wilson 1993) in American English. The examples in 22 and 23 illustrate the use of como ‘like’ as a punctor in the Spanish of HLS: (22) Tenemos un restaurante que nos gusta mucho, hacen como enchiladas verdes y cuesta el plato como 8 dólares, o sea, que no es muy caro el plato y te dan como arroz y frijoles [W,HLS,f20,C] ‘We have a restaurant that we like very much, they do like green enchiladas and the dish costs like 8 dollars, so it’s not very expensive and they give you like rice and beans.’ (23) Ahora tienen trabajos pa bilingües[...]. Porque están haciendo como entrevistas pero... como de salud o algo. Y necesitan gente que habla español. 47 ‘Now they have jobs for bilinguals [...]. Because they are doing like interviews but... like about health or something. And they need Spanish speakers.’ [L,HLS,m19,C] As illustrated in 22 and 23, my data show evidence of the use of como as a punctor in Spanish. In 22, the first como is not used with an approximative value because the speaker is saying that the restaurant that she and her fiancé like serves green enchiladas (and not ‘something like green enchiladas’). Also, the speaker says that the dish comes with como arroz y frijoles but it is not ‘something like rice and beans’, indeed that restaurant serves actual green enchiladas with actual rice and beans. In the same way, the first como that the speaker uses in example 23 does not mean ‘something like interviews’ because earlier in the conversation the speaker says that he is applying for the bilingual position that he refers to, and he has scheduled his interview already, therefore, he is sure that they are conducting job interviews and not ‘something like an interview’. Nevertheless, it is sometimes difficult to tell apart como uses as an approximation marker or as a punctor. For instance, the speaker in 22 also says that the dish of green enchiladas costs como 8 dólares; here, the use of como ‘like’ may have an approximative value since the price may not be exactly 8 dollars but something close to 8 dollars. Also, in 23 one could argue that the second como (in como de salud o algo, ‘like about health or something’) is used in an approximative sense further emphasized by the following expression o algo (‘or something’), as if the speaker were not quite sure of the nature of the interviews and thinks it is something related to health. In these two cases, 48 como may be classified as an approximative marker rather than as a punctor. In order to ensure reliability of the analysis, all the examples of como in my data that I had originally classified as approximative or punctors were further checked by a second independent Spanish speaker 7 . This independent rater was asked to classify the examples of como as approximative if he thought como could convey an idea of approximation or as a punctor if no such interpretation seemed possible. There was a high percentage of agreement between both raters (almost 98%). The few instances in which there was disagreement about the interpretation of como were not included in the quantitative analysis. Other studies of bilingual varieties in contact with English have reported similar findings. For example, Sankoff et al. (1997) in their study of Anglophone speakers of Montreal French find that comme ‘as’, ‘like’ in is used as an empty punctor, a use not found in monolingual French varieties. They mention the possibility of transfer from American English ‘like’, which is commonly used as a punctor, especially among young speakers. Since neither quotative nor punctor como have been reported in monolingual varieties of Spanish, it may be the case that these uses are transferred from similar uses of like in colloquial American English. 2.2. So, así que and entonces Several studies have shown that bilingual varieties in contact with English incorporate English-origin so 8 in their oral discourse (Aaron 2004; Lipski 1994; 7 This inter-rater check was performed by the author of the paper and by a native speaker of Spanish, who is also a researcher and professor of Spanish language and Mexican literature. 8 Throughout this chapter, italicized so refers to this marker within Spanish discourse of HLS whereas “so” in quotation marks refers to the marker in English discourse. 49 Sankoff et al. 1997; Silva-Corvalán 1994). The connector so from English is used in Spanish discourse normally phonetically adapted to Spanish and realized as [so], without gliding. English “so” has two possible equivalents in Spanish: así que and entonces. The literature identifies the following functions for the connectors entonces 9 , así que and so (Aaron 2004; Alvarez 1999; Bosque and Demonte 1999; Torres 2002). The examples that illustrate these functions are taken from the data analyzed in this study: 1. Consecutive connectors As consecutive connectors, así que and entonces express the ‘consequence of an action, circumstance and quality’ (Alvarez 1999: 3741; Martín Zorraquino and Portolés 1999: 4099-4108). They introduce or follow a deduction or conclusion, as the English connector “so” does (Aaron 2004; Torres 2002). The following examples illustrate this use of así que in 24, entonces in 25, and so in 26: (24) Había mucho racismo en San José cuando estaba creciendo así que sus padres sólo permitían que hablaran inglés ella y su hermana. [W,HLS,f20,P] ‘There was a lot of racism in San Jose when she was growing up so her parents let her and her sister to speak only in English.’ (25) A: Pero, ¿ya te estás especializando en comunicación? 9 Note that “entonces” may have other functions, as a temporal adverb meaning ‘then’ as in the following example: Llegué a la escuela muy temprano, a las siete de la mañana, porque no había mucho tráfico entonces. [C,HLS,f20,C] ‘I got to school very early, at 7AM, because there wasn’t a lot a lot of traffic then.’ This use is not included in my study. 50 ‘But, are you already majoring in communication?’ C: Sí porque entré a USC con créditos de AP y de otras clases que tomé en un colegio, se llaman junior colleges aquí, entonces ya estoy tomando mis clases para mi especialidad. [C,HLS,f20,I] ‘Yes because I came to USC with AP credits and with credits from other courses that I took in a college, they’re called junior colleges here, so I’m already taking the courses for my major.’ (26) Según los Estados Unidos es la la... el país de la oportunidad so por qué no brindarles la oportunidad a otra gente de otros países [D,HLS,f19,I] ‘According to the United States, it’s the ... the country of opportunity, so why not give the opportunity to other people from other countries.’ 2. Marking discourse progression So, así que and entonces can be used to mark discourse progression (Martín Zorraquino and Portolés 1999: 4107) as illustrated in 27 and 28. (27) Fue el día 28, el martes, el tráfico estaba bien pesado y luego más al rato me dijo el de la grúa que había dos accidentes separados [...]. So yo iba entre los carros, siempre manejo entre los carros en las horas de tráfico y [...] [J,HLS,m 20,C] ‘It was last Tuesday, the 28th, traffic was very heavy and then later the towing guy told me that there were two separate accidents. So I was going between the cars, I always drive between cars during rush hour.’ 51 (28) No quería ir, entonces yo le rogué, le rogué, le rogué, porque tenía una junta con best buddies, la primera junta del semestre, y entonces yo le rogué, le rogué y le rogué hasta que la convencí. [D,HLS,f19,C] ‘She didn’t want to go, so I begged and begged and begged her because she had a meeting with best buddies, the first meeting of the semester and, so I begged, and begged and begged until I convinced her.’ 3. Unfinished utterance Like English “so”, Spanish so can also end an utterance, leaving an often implicit utterance unexpressed as in 29: (29) Me gustaría que ganara Kerry porque él quiere pasar una ley para los estudiantes que no tienen papeles y mi hermana no es... es indocumentada, so—. [D,HLS,f19,I] ‘I’d like that Kerry wins because he wants to pass a law for undocumented students and my sister isn’t… she’s undocumented, so—.’ Typically, the utterance ends in a mid-tone, pending intonation, which reinforces the sense of an implicit ending left unsaid. According to Aaron (2004: 167), this use of so is not shared by entonces. However, there are several examples in my data that show that both entonces (as in 30) and así que (as in 31) can end an utterance in the same way as “so” in English: (30) Y pelean [las pandillas salvadoreñas] también contra los mejicanos y eso es algo yo creo que mucho más irónico porque son hispanos, entonces— [C,HLS,f20,I] 52 ‘And they [Salvadorian gangs] fight also against Mexicans and that is something I think much more ironic because they are all Hispanic, so—.’ (31) Yo digo, porque si eres mamá tú te vas a dar cuenta cuando tu hija... Y es una tontería si dices que no, y si dices que no es que no conoces a tu hija y mi mamá y yo éramos muy cercanas así que— [W,HLS,f20,I] ‘And I say, if you’re a mother you’re going to know when your daughter... And it’s ridiculous to say no, and if you say no it is because you don’t know your daughter and my mum and I were very close so—.’ 2.3. Punctors Punctors, also known as fillers or hesitation marks, are a class of particles that only occur in spoken language (Vincent and Sankoff 1992) and are characteristic of informal, casual production (Brown and Yule 1983). The punctors included in this analysis are classified into: a) English-origin punctors: “you know”, “I don’t know” and “I mean”, which are produced in Spanish discourse; and b) Spanish punctors: sabes, ‘you know’, no sé ‘I don’t know’, o sea, ‘that is’, este ‘uhm’. Following Vincent and Sankoff (1992) and Sankoff et al. (1997), the words that I consider punctors are desemanticized, that is, they have lost their original meaning. As discourse particles, “you know”, “I mean”, “I don’t know” are different from the original verbal expressions of a verb of thinking or feeling, which may be followed by the complementizer que, ‘that’ (i.e. ‘you know that’, ‘I mean that’, ‘I don’t know that’). When used as discourse particles, the complementizer que is not expressed and, in the case of you know, the researcher needs to make sure that no 53 acknowledgement on the part of the interlocutor is required or given in response (Sankoff et al. 1997: 198). Therefore, examples such as 32 are not included in the analysis since no sé (‘I don’t know’) is not a discourse marker but a verbal phrase meaning literally “I don’t know” (i.e. ‘I don’t have the knowledge’). However, no sé in 33 and “I don’t know” and “you know” in 34 are included because they are used as punctors, that is, they are desemanticized: (32) Bueno, 401 K es una opción, pero no sé mucho de los programas de retiro. [C,HLS,f20,I] ‘Well, 401 K is an option, but I don’t know much about retirement programs.’ (33) Mi opinión personal es no.... a decir verdad es no, no sé, demócrata, republicano o independientes o lo que sea va a ser lo mismo [...] el propósito de todo esto es de que namás para poder decir de que hay democracia en este país, hay libertad, pero no sé, para no sé, luego no sé, hay muchas cosas... [G,HLS,m20,I] ‘My personal opinion is that no, to tell you the truth, it’s, I don’t know, democrat, republican or independent or whatever it’s going to be the same [...] the purpose of all this is only to be able to say that there is democracy in this country, that there’s freedom, but I don’t know, in order to I don’t know, then I don’t know, there are a lot of things ...’ 54 (34) Así que primero la edad, segundo, I don’t know, era el primer hombre que llevó a mi novio a Hooters you know tienen una relación así, así que cómo voy a dejar a mi novio para andar con él you know. [W,HLS,f20,C] ‘So first the age, second I don’t know, he was the first man that took my boyfriend to Hooters you know they have such a relationship, so how could I leave my boyfriend to go with him you know.’ Examples 33 and 34 illustrate the uses of no sé ‘I don’t know’ and I don’t know and you know as punctors. For example, with the use of I don’t know in 34, speaker W is not indicating that she doesn’t know that [he] was the first man that took her boyfriend to Hooters; rather, she is saying that he indeed was that man. Thus, I don’t know is used as a filler here. Furthermore, another punctor could have been used instead, such as o sea ‘I mean’ or este ‘uhm’ (e.g. Así que primero la edad, segundo, este/o sea, era el primer hombre que llevó a mi novio a Hooters o sea/este tienen una relación así [...]). 3. Contractions With the term ‘phonological contraction’, I refer to a variant that is abbreviated or reduced in some way with respect to another variant of the same lexical item. Phonological contractions have been studied as a linguistic feature that correlates with specific types of registers. For example, Finegan and Biber (1994: 322) state that contractions are an instance of economy in language and they find that this feature usually correlates with situations that are less planned, more affective 55 and less informational and with shared background between interlocutors. Some researchers have also suggested the loss or lessening of part of phonetic body in high frequency words such as prepositions and conjunctions in informal registers (Vincent 1993; Vincent and Sankoff 1992). In this chapter, the analysis of contractions across different situations is intended to examine whether these items are more frequently produced when the style is more informal. I include the study of the pronunciation of the conjunction pues ‘so/well’, the preposition para ‘so that/for’, and the verb estar ‘to be’. 3.1. Pues (‘so’/‘well’) and para (‘for’/‘so that’) Pues (‘so/well’) and para (‘so that/for’) are often pronounced as [pos] and [pa] in informal oral registers (Hidalgo 1987; Sánchez 1983; Valdés 1988). The passage in 35 illustrates the casual pronunciation of pues and para by two HLS in the conversation situation: (35) L: Pos quiero ir a… quiero seguir yendo a la Universidad. [L,HLS,m19,C] ‘I want to go to... I want to keep on going to school.’ D: ¿A la escuela de graduado?, ¿graduate school? [D,HLS,f19,C] ‘Graduate school?’ L: Sí. ‘Yes.’ D: OK, pos ¿por qué tienes un Spanish major? ‘Ok, why do you have a Spanish major?’ L: Porque me gusta...o...¡no sé!... 56 ‘Because I like it, or... I don’t know!’ D: Entonces, ¿pa qué estás aquí? ¿pa qué stás aquí? [risas] ‘Then, why are you here? why are you here? [laughs] 3.2. Estar (‘to be’) The verb estar ‘to be’ is also contracted in casual speech in the various forms of its conjugation, in particular, toy, tás, tá instead of estoy ‘I am’, estás ‘you are’, está ‘he/she/it is’. The following examples illustrate the contraction of estar: (36) B: So, ¿quieres helado? [B,HLS,f19,C] ‘So, do you want an ice-cream?’ G: Tá bien, sí, sí... A mí me gusta cookies and cream [G,HLS,m20,C] ‘All right, yes, yes... I like cookies and cream.’ (37) B: [...] porque, ay, no sé, toy bien floja... [B,HLS,f19,C] ‘ because, ah, I don’t know, I’m very lazy.’ 4. Hypotheses The hypotheses of this study are related to the uses of the particles outlined above by HLS and SLS across linguistic situations. In general, I hypothesize that speakers will vary their use of discourse particles and contractions across situations of use. More specifically, with respect to each of the linguistic features investigated in this chapter, I hypothesize that: 1- HLS will vary their use of the particle como ‘like’ across situations. In particular, I expect that more informal situations will show an increase of the non- 57 canonical uses of como in Spanish, namely como as a quotative and como as a punctor. Furthermore, the literature claims that como used with the value of approximation is more frequent in casual registers (Moreno Ayora 1991; Santos Río 2003); therefore, I expect the approximative use of como (followed by a numeric expression or not) to be higher the less formal the situation is. Additionally, I hypothesize that SLS may also use como as a punctor and quotative, transferring these uses from English into their Spanish discourse. 2- HLS will vary their use of English-origin so in Spanish producing it less frequently the higher the level of formality. I expect that así que and entonces will be used more frequently in the presentations and interviews, and occurrences of so will be higher in the conversations. On the other hand, I do not expect SLS to produce so in their Spanish discourse. So, adapted to Spanish and realized as [so], may be part of the variety spoken by HLS, but that is not the case with the SLS group, who will use así que and/or entonces instead as they are normally taught in Spanish class. 3- HLS and SLS are expected to produce an overall higher number of punctors in conversations. However, I expect that SLS will produce an overall lesser number of punctors across registers. Furthermore, I do not expect either group of speakers to produce punctors in presentations since this situation is not interactive but rather planned, formal, and non-interactive, where other discourse markers may be employed but not informal punctors. I also hypothesize that HLS will use Spanish-origin punctors more often than SLS, whereas the latter may use more English punctors due to their lesser fluency in informal oral interactions in Spanish. 58 4- HLS will produce phonological contractions more frequently in relaxed casual situations (Hidalgo 1987; Valdés 1988). Therefore, I expect to find pues and para realized as pos and pa, as well as the verb estar contracted in various forms of its conjugation (e.g. toy, tás, tá) in conversation more frequently than in the other situations. However, I do not expect phonological contractions to be found in the speech of SLS since their exposure to and acquisition of Spanish phonology derives, for the most part, from the language classroom and not from informal casual interactions. 5. Results and discussion In this section, I first report the results obtained for each of the linguistic features analyzed in this chapter. In analyzing and reporting the results of this chapter, I follow the criteria explained in the methodology section of chapter 1, namely, all results are normalized to 15 minutes per speaker, so that linguistic features can be compared across different situations. For the discourse particle so, there are alternative pronunciations that are discussed in this section (as /so/ adapted to Spanish or as /so ʊ/ with the English pronunciation). These variants are identified by the researcher with the aid of an inter-rater for the sake of reliability. Both raters have had training in phonetic transcription. No acoustic programs was deemed necessary for the analysis of this feature since the instances of /so/ or /so ʊ/ were clear in the recordings and there was 59 no disagreement between the researcher and the inter-rater in any of the tokens discussed. A discussion of these results follows. 5.1. Como The results show that HLS use non-standard functions of como, namely, como as a quotative and as a punctor. Although approximative como is a use recognized in the Spanish literature, I have also included como with an approximative value in my analysis because it has been claimed that this use is characteristic of informal registers (Santos Río 2003); therefore, it is important to examine the difference, if any, in the use of approximative como by HLS across registers. Table 5 gives the rate of use of como produced by HLS in the three situations analyzed. The rows indicate the different speakers and the columns include the functions of como, which are classified as: (1) approximative (2) quotative, and (3) punctor. Figure 2 displays graphically the percentages of use of these three functions of como according to the register. CONVERSATION INTERVIEW PRESENTATION Approx. Quot. Punct. Approx. Quot. Punct. Approx. Quot. Punct. B 6.1 1 3.4 1.8 0.8 1.2 1.3 0 1 C 4.6 0.6 2.5 0.5 0 0.4 0.3 0 0 D 4 0.4 3.1 1.5 0.9 1.7 2.4 0 0.2 E 2.4 0 1.4 0.4 0 0 0 0 0 W 3.3 1.2 2 1.3 0.3 0.5 0 0 0.6 J 5.4 1.2 2.4 3.2 0 1.9 3.1 0 0.8 L 5.2 0.3 3.7 2.1 0.2 0.7 2.6 0 1 G 4.4 1.7 3.2 2.6 1.2 2.1 1.6 0 0 Total 35.4 6.4 21.7 13.4 3.4 8.5 11.3 0 3.6 Table 5. Rate of use of como produced by HLS across registers 60 59 22 19 65.3 34.7 0 25.2 10.6 64.2 0 10 20 30 40 50 60 70 80 90 100 Conversation Interv iew Presentation Percentage Approximative Quotative Punctor Figure 2. Percentage of use of como across registers by HLS As figure 2 shows, the approximative, quotative, and punctor uses of como are higher in the conversation than in the interview or the presentation. The presentation has the lowest percentages of use of punctor como, 10.6%, as opposed to 25.2% in the interview and 64.2% in the conversation. Furthermore, there are no occurrences of quotative como in the presentation. When direct speech is present, speakers may use an overt introducer such as como or a null or zero introducer (Ferrara and Bell 1995). Therefore, the use of como as a quotative depends on whether speakers include direct speech in their discourse. In my data, the lack of quotative como is not due to a lack of possible contexts in which como as quotative could have occurred since there are several instances of direct quotations in the presentations. However, speakers choose other ways to introduce direct quotations in the presentations such as the verb decir ‘say’. In the same way, all speakers produced direct quotations during the interviews, although not all of them used como as an introducer of direct speech. Table 6 shows the 61 frequency and percentages of direct quotations in each situation and the use of como as an introducer as opposed to decir ‘say’ or a null introducer. Introducers of direct quotations como decir null Total Conversation 6.4 27.6% 10 43.1% 6.8 29.3% 23.2 100% Interview 3.4 18.7% 12.1 67% 2.6 14.3% 18.1 100% Presentation 0 0% 2 24 % 6.3 76% 8.3 100% Total 9.8 19.8% 24.1 48.6% 15.7 31.6% 49.6 100% Table 6. Frequency of the three introducers of direct quotations used by HLS in each register The percentages in table 6 show that como is less frequent as an introducer of direct speech than decir ‘say’ or a null introducer. Como as a quotative is more frequent in the conversations (27.6%) than in the interviews (18.7%). Furthermore, it does not occur in the presentations, thus confirming the hypothesis that como as quotative would be more frequent in informal contexts. SLS also produce approximative, quotative, and punctor como. The patterns of use are in some respects similar to those of HLS but there are some important differences. Table 7 shows the total rate of use and percentages of como in the different situations. Conversation Interview Presentation Total Como-approx. 25.7 72.6% 6.7 19% 3 8.4% 35.4 100% Como-quotative 3.3 71.7% 1.3 28.3% 0 0% 4.6 100% Como-punctor 45.6 64% 16.7 23.4% 9 12.6% 71.3 100% Table 7. Rate of use of como produced by SLS across registers Similar to HLS, the approximative, quotative, and punctor uses of como are also higher in the conversation than in the interview or the presentation in the SLS data. In addition, SLS do not produce quotative como in the presentations either. On 62 the other hand, SLS produce fewer approximative and quotative como across situations than HLS (a total rate of 35.4 and 4.6 respectively as opposed to 60.1 and 9.8 for HLS). Additionally, they produce significantly more punctors than HLS in all situations; the total rate of punctor use for SLS is 71.3, whereas HLS have only 33.8. However, the most interesting difference between HLS and SLS is that the latter code-switch to English like quite often across situations, especially during conversations, as illustrated in the following examples: (38) Está bien porque yo realicé que like es muy estúpido pensar eso. [Ge,SLS,f20,C] ‘It’s all right because I realized that like it’s very stupid to think that.’ (39) La cosa que no me gusta es que en esta universidad los hombres no son de like ninguna calidad. [Co,SLS,f20,C] ‘The thing that I don’t like is that in this university men aren’t like of any quality.’ In 38 and 39, the speakers are producing a single switch to English like as a punctor within their Spanish discourse. The data show that SLS produce English like within Spanish discourse with the value of approximation, as a quotative, and as a punctor. Table 8 gives the rates of use of these functions of like across registers. 63 Table 8. Rate of use of like produced by SLS across registers The results of the production of like by SLS are interesting when compared to HLS, who rarely switch to English like and rather produce como for approximative, quotative, and punctor uses. There are only three instances of switches of like as a punctor and one as a quotative in HLS data. The instances of punctor like, however, correspond to code-switching occurrences in which a whole phrase, not only the word like, is being switched to English. Furthermore, these are all produced by the same HL speaker. The following examples show these occurrences of punctor like: (40) Tienen comida bien rica, o sea comida de... no sé... tienen como desayuno, tienen... comida... no like random...sí pero lo que sea. [B,HLS,f19,C] ‘They have very nice food, well, food as... I don’t know... they have like breakfast, they have food... not like random... yes but whatever it is.’ (41) B: Sí, pero la policía cuando te para no va a saber. [B,HLS,f19,C] ‘Yes, but when the police stops you they are not going to know.’ friend: A no ser que compruebe, ¿no? ‘Unless they check, no?’ B: Bueno, depende, like I guess... ‘Well, it depends, like I guess...’ Conversation Interview Presentation Total Like-approx. 6 100% 0 0% 0 0% 6 100% Like-quotative 6 75% 2 25% 0 0% 8 100% Like-punctor 8.7 50% 6.7 38.5% 2 11.5% 17.4 100% 64 (42) El tráfico...no... it’s just a dirt, like, one road... el tráfico va a estar horrible [B,HLS,f19,C] ‘Traffic...no... it’s just a dirt, like, one road... traffic is going to be horrible.’ In these examples, speaker B switches to English more than just the word like, as in “like random” or “like I guess” in 40 and 41. Example 42 shows even a longer sequence produced in English, with like as a punctor. This use of like by HLS differs from SLS in that SLS frequently produce single-switches, as illustrated in examples 38 and 39. Another difference is that SLS use like as a quotative in their Spanish discourse, whereas there is only one occurrence of quotative like produced by HLS: (43) I was like... Le dije “tú vas a perder muchos amigos [...]” [W,HLS,f20,C] ‘I was like... I told him “you’re going to lose many friends [...]” ’ Yet, in example 43, quotative like is followed by a pause and then paraphrased by the Spanish quotative verb decir ‘say’ introducing direct speech. Therefore, switching to English like is more frequent among SLS than HLS. Figure 3 shows the distribution of each of these uses of como and like for SLS: Figure 3. Percentage of SLS use of como and like across registers 0 20 40 60 80 100 Conversation Interview Presentation Percentage como-punctor like-punctor como-quotative like-quotative como-approximative like-approximative 65 In figure 3, we can see the general pattern of higher percentages of approximative and non-canonical uses of como in conversations, thus coinciding with the HLS pattern and confirming the initial hypothesis. However, figure 3 also shows that SLS produce more como as a punctor than HLS as well as como switched into English like across all registers. 5.2. So, así que and entonces HLS use the connector so from English in their Spanish discourse as well as the equivalent connectors in Spanish entonces ‘so/then’ and así que ‘so’. It is hypothesized that so will be more frequent in the conversations than in the interviews and the presentations. The reason for this hypothesis is that HLS will probably pay more attention to their speech in formal registers and, therefore, will try to avoid features such as the use of so transferred from English (Smead 1998; Valdés 1988); especially, in a formal classroom presentation in Spanish where students are expected to show their competence and mastery of both the content (i.e. the subject) and the form of the presentations (Spanish academic language) (Valdés and Geoffrion-Vinci 1998). The results reported here include seven of the eight HLS participants since one of the speakers, L, did not produce so, así que or entonces in any situation. Instead, speaker L uses other discourse particles more frequently such as pues ‘well/so’ or the conjunction y ‘and’ where other speakers would use so, así que or entonces. For example, L uses pos (i.e. the contracted form of pues) in (44) whereas speaker D uses so in a similar context in (45): 66 (44) Según este país es pa la libertad, pos de todos modos deben tener estos derechos [los inmigrantes ilegales] [L,HLS,m19,I] ‘This country is for freedom, so anyway they [illegal immigrants] should have these rights.’ (45) Según los Estados Unidos es la… el país de la oportunidad, so por qué no brindarles la oportunidad a otra gente de otros países [W,HLS,f20,I] ‘The United States is the country for freedom, so why not give the chance to other people from other countries.’ Figure 4 displays the percentages of use of so, así que and entonces for HLS across registers: 49.5 51.9 44.2 25.8 15.1 32 33 23.8 24.7 0 10 20 30 40 50 60 70 80 90 100 Conversation Interview Presentation Percentage so asi que entonces Figure 4. Percentage of use of so, así que, and entonces by HLS across registers There is an overall higher percentage of use of so across situations (of the total production of so, así que and entonces considering the entirety of the data, 50.3% correspond to so, 22.1% to así que and 27.6% to entonces). However, so is not more frequent in the conversations than in the interviews as it had been predicted 67 and, furthermore, there is a higher percentage of so in the interviews and presentations than both así que and entonces, which also contradicts the expectations. If we analyze the results of so, así que and entonces depending on the register, the differences in the production are not statistically significant (p> .4). However, if we look at the use of these three connectors by the individual HLS, it seems that the choice of so, así que and entonces depends on the speaker’s preference, rather than on register considerations. Figure 5 illustrates the results of the use of so, así que and entonces by individual. 0 10 20 30 40 50 60 70 80 90 100 so asi que entonces G J W C D B E Figures 5. Use of so, así que, and entonces by individual HLS in the three registers Figure 5 shows that four of the speakers (i.e. speakers G, J, B, and D) prefer so over the Spanish connectors así que and entonces across registers. On the other hand, speaker W produces most of the occurrences of así que in every situation, and C produces most of the occurrences of entonces. The results indicate that the use of 68 0 10 20 30 40 50 60 70 80 90 100 conversation interview presentation percentage so entonces so, así que and entonces does not vary depending on register but rather on speaker’s idiosyncratic use. SLS produce an overall lesser amount of these connectors than HLS across registers. In fact, no SLS produces así que at all. Unlike HLS, individuals in the SLS group behave similarly, for this reason, speakers are pooled in the graph and table shown in figure 6. Rate of use Conversation Interview Presentation So 15.4 12.2 0 Así que 0 0 0 Entonces 55.4 44.3 15 Total 70.8 56.5 15 Figure 6. Percentage and rate of use of so, así que and entonces by SLS across registers Apart from the fact that SLS do not produce the connector así que, another difference with the HLS group is that entonces is the connector that all SLS produce the most across registers (unlike HLS who overall favor so). Even though SLS produce “so” in conversations and interviews, the analysis of these instances confirms the expectation that SLS would not adapt so to the Spanish phonology; rather, “so” is pronounced with gliding, beginning in a mid back vowel and ending in a high back lax vowel (i.e. /so ʊ/). Furthermore, SLS do not produce “so” in the 69 presentations. In addition, all the instances of “so” in the SLS data are produced at the end of an utterance, ending in a mid-tone, pending intonation which reinforces the sense of an implicit ending left unsaid (Aaron 2004: 167). The following examples illustrate the use of “so” by SLS: (46) Es que yo puedo verme haciendo muchas cosas que me hacen feliz, entonces no tengo mucho miedo de entontrar trabajo, so— [K,SLS,f20,C] ‘The thing is that I can see myself doing many things that make me happy, so I’m not afraid of finding a job, so—’ (47) La realidad es que no soy muy activo. Tengo mis opiniones y todo esto y, completamente, no estoy de acuerdo con esta guerra y muchas otras cosas en este país, pero la realidad es que no hago mucho para pararlo, so— [Cr,SLS,m22,I] ‘The reality is that I’m not very active. I have my opinions and all that and, absolutely, I don’t agree with this war and many other things in this country, but the truth is that I don’t do much to stop it, so—’ The rest of the examples of “so” produced by SLS are similar to 46 and 47. Therefore, this is another difference between HLS and SLS since the latter use “so” only at the end of utterances in the data while the former produce it: a) more frequently in all registers, including presentations; b) adapted to Spanish phonology, as /so/; c) in all possible positions, that is, at beginning of an utterance, in the middle, and at the end. 70 5.3. Punctors I report here the results of the Spanish punctors sabes ‘you know’, o sea ‘I mean’, no sé ‘I don’t know’, and este ‘uhm’ as well as the English punctors (produced in English within otherwise Spanish discourse) “you know”, “I don’t know”, and “I mean”. The use of como as a punctor is not included here since I reported those results together with other functions of como in section 5.1. Table 9 illustrates the percentage of use by HLS and SLS of the punctors analyzed in this chapter. Group HLS (total) SLS (total) Situation Conv. Interv. Pres. Total Conv. Interv. Pres. Total sabes 21.2 1.3 0 22.5 3.1 0 0 3.1 no sé 73.4 121.4 0 194.8 61.5 45.8 0 107.3 o sea 11.2 9.8 0 21 0 0 0 0 este 19.1 1.3 0 20.4 0 0 0 0 you know 18.3 1.6 0 19.9 9.4 2.4 0 11.8 I don’t know 9.7 23.2 0 32.9 7.2 5.7 0 12.9 I mean 7.2 9.2 0 16.4 6.5 10.7 1 18.2 Total 160.1 167.8 0 327.9 87.7 64.6 1 153.3 Table 9. Rate of use of punctors by HLS and SLS across registers Table 9 shows that both groups produce more punctors in the conversation than in the interview and no punctors are used in presentations (except for one instance of “I mean” by one SLS). Chi-square tests indicate that that the differences observed are statistically significant for both groups (p< .001). For both groups the most frequently used punctor both in interviews and conversations is the Spanish punctor no sé ‘I don’t know’ (59.4% of total punctor use 71 for HLS and 70% for SLS). However, there are differences between HLS and SLS: Overall HLS use more punctors than SLS, as was predicted. SLS do not use the Spanish o sea ‘I mean’ and este ‘uh’ at all in any of the registers and they have very few occurrences of sabes ‘you know’. Thus, SLS produce more English punctors, except for the use of no sé. On the other hand, HLS produced all the punctors analyzed in both the conversation and the interview situation. All punctors are more frequently used in HLS conversations except for the Spanish punctor no sé and the English equivalent “I don’t know”, which have higher rates of use in interviews. These results may be pointing at a difference between no sé and “I don’t know” and other punctors. No sé and “I don’t know” may have more than only the function of being a conversational marker or punctor. These discourse particles may function as a disclaimer, that is, as a discourse device to weaken an assertion. For example, in 33, repeated here as 48 for ease of reference, the speaker may be using no sé as a way of trying to mitigate his political opinion: (48) Mi opinión personal es no.... a decir verdad es no, no sé, demócrata, republicano o independientes o lo que sea va a ser lo mismo [...] el propósito de todo esto es de que namás para poder decir de que hay democracia en este país, hay libertad, pero no sé, para no sé, luego no sé, hay muchas cosas ... [G,HLS,m20,I] Speaker G is expressing his opinion regarding the U.S. presidential election that was taking place at the time of the interviews. His opinion is rather cynical and he expresses his lack of hope for a change regardless of the political party in power. 72 However, he seems to try to tone down his cynicism by the use of the discourse particle no sé, as if he could not make up his mind, rather than stating his position in a more direct manner. 5.4. Contractions The results of phonological contractions of pues ‘so/well’, para ‘for/so that’, and estar ‘to be’ are shown in the following set of figures. SLS did not produce any contractions; therefore, only HLS are included in the results reported in this section. Figure 7 shows the production of pues vs. pos. 0 10 20 30 40 50 60 70 80 90 100 pues pos percentage conversation interview presentation Figure 7. Percentage of pues vs. pos across registers by HLS There is an overall higher production of pos in conversation with 68% versus 32% use of pues. This percentage of pos decreases in the interview (55%) although it is still higher than the percentage of pues (45%). Not surprisingly, the biggest contrast is found in the presentation situation with 74% overall use of pues as opposed to only 26% of pos. Chi-square tests confirm the statistical significance of these contrasts (p< .003). Figure 8 shows the results of para vs. pa across situations. 73 0 10 20 30 40 50 60 70 80 90 100 para pa percentage conversation interview presentation Figure 8. Percentage of use of para vs. pa across registers by HLS Results are also statistically significant for the contrasts found in the production of para and pa across registers (p< .006). Figure 8 shows that HLS did not produce pa in the presentations and the percentage of pa in conversations is higher than in interviews. However, the percentages of pa are very low compared to pos (see figure 7). This difference between the contractions pos and pa is interesting and it may indicate that pa is more socially stigmatized than pos. Even if both are considered informal pronunciations, pa is regarded as more characteristic of rural or uneducated people (Sánchez 1983), whereas the use of pos is more widespread, perhaps motivated by pronunciation in faster speech production. Furthermore, the pronunciation of pues as pos is less salient as it only involves vocalic sounds (namely, the monophthongization of [ue] to [o]), whereas in pa an entire syllable is lost, which is a more salient feature. Finally, figure 9 shows the results of the contractions of estar in the different forms of its conjugation. Actually, the most frequent contractions of estar correspond 74 to the third person singular (i.e. está ‘he/she/it is’), followed by the first person singular (estoy ‘I am’). The rest of the forms are rarely contracted although there are few examples of the third person plural (están ‘they are’) and the first person plural (estamos ‘we’); therefore, I have grouped the latter forms together for the purpose of reporting the results, as figure 9 shows. 0 10 20 30 40 50 60 70 80 conversation interview presentation percentage ta toy tan/tamos Conversation Interview Presentation Total tá (está) 4.5 58.4% 2.8 36.4% 0.4 5.2% 7.7 100% toy (estoy) 2.4 60% 1.6 40% 0 0% 4 100% tán/tamos (están/estamos) 2.7 73% 1 27% 0 0% 3.7 100% Figure 9. Rate of use and percentage of estar-contractions across registers by HLS Similar to the contractions of pues and para, the verb estar is more frequently contracted in conversations, followed by interviews and only few contractions in presentations. In fact, there is only one occurrence of tá by one speaker in his presentation. These differences are statistically significant across situations (p< .02). 75 5.5. Discussion The results confirm that HLS and SLS show variation across registers in the use of some of the discourse particles and contractions analyzed in this chapter. This evidence suggests that the participants in the present study are not monosylistic but, rather, they make use of linguistic features, discourse particles in this case, differently. HLS and SLS construct what they consider appropriate discourse according to the formality of the situation: academic or non-academic. For example, there are fewer uses of punctor and quotative como in the presentations and punctors, contracted estar, and pa (for para ‘for/so that’) are not produced in that situation. Let us look at each of the discourse particles in detail. 5.5.1. Use of como by HLS and SLS This study shows that the use of the word como by HLS seems to be calquing some functions from the equivalent ‘like’ in American English, which como does not traditionally have in monolingual varieties. It was predicted that HLS would vary their use of como across situations. In particular, it was expected that non-canonical uses of como would be more likely to occur in casual registers such as the conversation situation in this study. Specifically, these are the quotative and punctor uses of como, which have not been described in the Spanish literature and seem to be characteristic of the Spanish of HLS. HLS have a higher rate of use of quotative como in the conversations than in the interviews and there were no occurrences of quotative como in the presentations. However, there are examples of quotative como in conversations and interviews. In 76 fact, considering the entirety of conversations and interviews 19.8% of the direct quotations in the data are introduced by como. This is a very interesting development of the discourse particle como, which has not been studied in monolingual or bilingual varieties of Spanish. However, como introducing direct speech is not likely to occur in a class presentation not because of the Spanish proficiency of a HLS but rather because of the nature of academic presentations. Direct speech is characteristic of narrations of personal experience stories (Goffman 1981; Ferrara and Bell 1995) and direct quotations are one of the strategies that speakers often use for dramatic effects, to make the story come alive (Silva-Corvalán 1994: 66). In contrast, presentations are primarily informational in purpose, directed to a non-interactive audience; therefore, direct quotations are less frequent, and indirect speech is preferred to report someone else’s words. Nevertheless, the reason for the absence of como as a quotative in the presentations is not due to a lack of contexts where quotative como could have been produced. Table 3 shows that there are instances of direct quotations in the class presentations and, therefore, it was possible for HLS to choose como as an overt introducer. However, none of the direct quotations in the presentations are introduced by como. Instead, HLS use a null introducer or the verb decir ‘say’ to introduce direct speech: (49) Está trinchando esta cebolla y dice “quién eres, qué estás haciendo con tu vida”. [L,HLS,m19,P] 77 ‘[She] is chopping this onion and says “who are you, what are you doing with your life”.’ Speakers do not produce this use of como in the register that is considered more formal and in which they are more likely to pay more attention to their speech. There have been very few studies on the academic production of HLS in Spanish. The study by Valdés and Geoffrion-Vinci (1998) analyzes class oral presentations by HLS in Spanish, but does not analyze discourse particles in detail nor do they mention the uses of como as quotative. Therefore, I cannot compare my results with previous studies, which may also be a further indication that this is a recent innovation in Spanish as heritage language and that may take time for this use to be recognized in the literature. Sankoff et al. (1997) studied bilingual speakers of Montreal French and also found the usage of French comme ‘like’ as introducer of reported speech, which is not found among monolingual varieties of French. Sankoff et al. (1997) report an average of 14% of quotative comme, similar to the percentage of quotative como in the data analyzed in this study. These authors state that this use of comme is a recent feature in Montreal French and suggest that it may be transferred from the quotative use of like in American English (p. 208). This may also be a likely explanation for the patterns observed in my study since no monolingual variety of Spanish has been shown to use como as an introducer of direct speech. The use of como as a punctor is also a characteristic of the Spanish discourse of HLS although it has not been found in other varieties of Spanish. The results of 78 this study confirm the hypothesis that punctor use of como is more frequent in the conversations (64.2%) than in interviews (25.2%), and it is even less frequent in the presentations (10.6%). Sankoff et al. (1997) report that the most frequent use of comme ‘like’ among English dominant bilinguals in Montreal French is as a desemanticized punctor. Their results indicate a total average of 35% of comme- punctor use in their data when compared to other punctors. The example in 50, taken from their paper (p.205), illustrates this use: (50) Comment est-ce que je peux comme prendre un petit promenade après? ‘How can I like take a little walk afterwards?’ This use of comme as a punctor is a recent development in Canadian French that seems to be increasing among younger speakers (Sankoff et al. 1997; Vincent and Sankoff 1992). Moreover, Sankoff et al. (1997) compare data collected in 1984 with data recorded in 1993-1994 and report a 22.3% increase of use of punctor comme from 1984 to 1993. Comparing punctor como to other punctors used by HLS, my data show a lower percentage (9.3%) than the 35% of punctor comme reported in Sankoff et al. (1997). It may be possible, however, that the use of Spanish como as a punctor is also a recent development in the Spanish of bilingual communities in the U.S., which may be increasing among HLS in the same direction of comme in Montreal French. Furthermore, it is also possible that the use of como as a punctor in the Spanish of HLS may be due to transfer from the use of like as punctor in English, as Sankoff et al. (1997: 209) propose for the development of punctor comme in Montreal French. 79 Sankoff et al. (1997), however, do not differentiate among registers or levels of formality and, therefore, we do not know if bilingual Montreal speakers vary their use of punctor comme depending on the type of register. The data analyzed in this dissertation provides evidence that HLS use como as a punctor differently across registers. As we can see in the results illustrated in Table 5 and Figure 2, the use of como as a punctor decreases as the formality of the situation increases. This is an interesting finding that suggests that HLS are able to use discourse particles differently depending on the requirements of the register. Finally, my data confirm to some extent the hypothesis that approximative como occurs more frequently in more informal registers. The results show that como with an approximative meaning is more frequent in the conversations (59%), than in the interviews (22%) and presentations (19%). The results of SLS are similar to those of HLS in that non-canonical uses of como are more frequent in conversations than in interviews or presentations with the latter having the lowest percentages of approximative and punctor como. Furthermore, quotative como does not occur in presentations in the SLS data either. Therefore, we can see variation across registers for the SLS group too, although, the production of como by SLS is much lower compared to that of HLS. I believe that the use of como as quotative and punctor by SLS is a result of the individual’s transfer from their own native English into their Spanish. In the SLS data, not only do they produce less como than HLS, but they also switch to English like quite frequently, which is not a characteristic of the formal Spanish that they have been 80 exposed to at school. On the other hand, the results of HLS may reflect a transfer that is a much more widespread contact phenomenon at the level of Spanish speaking communities in the U.S. This is reflected in the systematic use of como as a punctor and quotative by all heritage language speakers in this study, rather than switching to English like. Therefore, HLS may be exposed, as native speakers of a contact variety, to these uses of como as punctor and quotative, which may be becoming part of the Spanish that is spoken in their communities. 5.5.2. Use of so, así que and entonces One of the characteristics of the Spanish spoken by HLS in the U.S. is the incorporation of single-word loanwords from English produced with Spanish morphology and pronunciation (Silva-Corvalán 1994; Smead 1998). So, from English “so”, is commonly inserted in spoken Spanish discourse and adapted to Spanish phonology (it is pronounced [so], without gliding) (Mendieta 1999: 144- 145; Silva-Corvalán 1994: 171). Further evidence that this [so] is part of the HLS Spanish variety in my data is that SLS do not have this feature. Instead, SLS produce “so” with an English pronunciation, and much less frequently than in the HLS data. There is great variation in the use of so, así que ‘so’ and entonces ‘so/then’ by the HLS participating in this study and differences observed across registers are not statistically significant. Of the seven speakers that produce so, así que or entonces, four speakers use so more frequently; Furthermore, así que was produced mainly by W and entonces by C (90.6% of the total occurrences of así que correspond to W and 41.2% of the total production of entonces to C). These results 81 do not confirm the hypothesis that the greater attention paid to speech in formal registers, especially in a formal classroom presentation in Spanish (Valdés and Geoffrion-Vinci 1998), will result in HLS’ avoidance of features such as the use of so transferred from English (Smead 1998; Valdés 1988). Several studies have investigated whether so is replacing the Spanish forms así que or entonces. In the present study, the results of the use of so, así que or entonces by HLS show that there is no variation across registers and it appears that the choice of connector depends mainly on the speaker’s preference. A possible explanation for the variation in my speakers’ use of so, así que and entonces may be related to the amount of daily use of Spanish. Several studies have observed that speakers that more frequently insert English function words like so into Spanish discourse speak primarily English (Phillips 1967: 643; Lipski 1994, 2004). For example, Lipski (2004: 15) states that Spanish bilingual speakers who employ so insertion normally spend more time each day speaking English than Spanish, due to the social conditions under which Spanish and English are used in most of the United States. Although the speakers in our study completed a language background questionnaire to make sure that all participants had similar patterns of use of Spanish, it is nevertheless possible that some of the HLS speak Spanish somewhat more frequently on a daily basis than others. It could be the case that speakers W and C, who rarely produce so in the data, use Spanish more frequently than G, J or D, who have high percentages of so across situations. As I explained in the methodology section, I took into account these external factors and aimed at 82 controlling them for the purpose of obtaining comparable data by selecting a homogeneous population. Therefore, all HLS have indeed similar patterns of Spanish use according to their own evaluations and they have, furthermore, similar amounts of formal education in Spanish. Nevertheless, these methodological precautions cannot guarantee that HLS are going to use exactly the same amount of Spanish on a daily basis and thus, the amount of Spanish use by individual HLS is not an implausible explanation. In the case of SLS, however, similar results were obtained across speakers. The preferred connector is by far entonces in all situations, whereas así que is not produced at all. The reason why SLS do not produce así que may be due to the fact that it is somewhat more informal than entonces, which is more neutral and can be used across registers from formal and more literary to informal and conversational. SLS may not have been exposed to así que as much as to entonces in their Spanish classes. Finally, the results confirm the hypothesis that SLS would not produce so adapted to Spanish and realized as [so]. It appears that [so] may be part of the variety spoken by HLS, but not by the SLS group, who use entonces instead as they are normally taught in Spanish class. 5.5.3. Use of punctors In this chapter, I hypothesized that punctors would be less frequent in presentations, due to the characteristics of this register. The results confirm this hypothesis since neither HLS nor SLS use punctors in the presentations. There is no significant difference either between the total production of punctors in the 83 conversation and in the interview. Of the seven punctors analyzed in this study, no sé ‘I don’t know’ is by far the preferred punctor by both groups of speakers in both the conversation and interview situations. These results do not totally coincide with Valdés and Geoffrion-Vinci’s (1998) study of the oral production of HLS in academic settings (university class presentations). They report that HLS use fillers, or “disfluency markers” (p.479), in class presentations. In particular, the speakers of Valdés and Geoffrion-Vinci’s study produce este quite frequently, whereas este in the data analyzed here mainly occurs in the conversations. Valdés and Geoffrion-Vinci (1998) analyze data from class presentations only. The difference between their study and this dissertation is likely related to differences in the data analyzed. When Valdés and Geoffrion-Vinci (1998: 479) report on the use of punctors, they include autobiographical presentations of the speakers, in which they talk about their background and interests. This type of text is more involved and personal rather than informational; therefore, it is more likely to include punctors than the presentations in my data in which no personal information is included. The data analyzed in this dissertation show that HLS and SLS use English punctors “you know”, “I don’t know”, and “I mean” in Spanish discourse. The insertion of English punctors in bilingual varieties in contact with English is well- established and many studies have observed this feature (Lipski 1994, 1996; Sankoff et al. 1997; Silva-Corvalán 1994). For example, Sankoff et al.’s (1997) study of 84 Montreal French also observes the production of English-origin “you know” in French discourse, as illustrated in 51 (Sankoff et al. 1997: 199): (51) C’est comme, you know on était des jeunes, on avait dix and onze ans ‘It’s like, you know we were young, we were about ten eleven years old.’ However, not many studies of bilingual varieties have analyzed punctor variation across situations of use or degrees of formality (e.g. Sankoff et al. 1997 study interviews; Valdés and Geoffrion-Vinci 1998 focus on presentations) and I cannot compare the results of this study to the use of punctors in different registers by other communities or groups of speakers. In any case, the results of this study show that speakers do not use punctors in the presentations. This suggests that there is stylistic variation across registers and that HLS (and, in the case of punctors, also SLS) adjust their production to suit a formal academic register. 5.5.4. Contractions Spanish spoken in various heritage communities in the U.S. is characterized by the use of linguistic features that are normally found in informal and casual registers (Hidalgo 1987; Sánchez 1983; Valdés 1988), such as the pronunciation of pues as [pos] and para as [pa], and the elimination of the first syllable of estar in the various forms of its conjugation (toy, tá, tás, etc.) (Cedergren 1979). However, as HLS gain access to different styles and registers in Spanish, they are expected to monitor their speech in more formal registers, where non-canonical forms such as pos and pa would be inappropriate (Valdés 1995). It was hypothesized that the HLS in this study would use the reduced variant more frequently in the conversations 85 since it is a more informal situation, and, on the other hand, the full forms pues, para, and estar would be more frequent in the interviews and, particularly, in the presentations. As the results show, this hypothesis is confirmed in the present study for HLS. This is especially evident in the use of pues versus pos across situations: In the conversations there is 32% of pues vs. 68% of pos, almost the mirror image of the presentations which have 74% of pues vs. 26% of pos. Additionally, HLS do not produce pa or contracted estar (except for only one instance of tá by one speaker) in the presentations. The pronunciation of para as pa as well as contracted estar are much less frequent even in informal situations than pos. This is the most interesting finding with respect to the production of the contractions analyzed in this study because it suggests that the status of these forms may not be the same. In the literature, these forms are considered equally non-canonical, typical of rural varieties or uneducated speakers (Hidalgo 1987; Sánchez 1983). However, pos is very frequent and even if its use decreases as the formality of the register increases, it is still common in the interviews and it also occurs, even if less frequently, in the presentations. On the contrary, pa and contracted estar are not as frequent and very few instances are produced in presentations. I suggest that the reason for this contrast may be related to the perceived prominence of these reduced forms. The pronunciation of pues as [pos] is less salient as it only involves one vocalic sound, whereas in pa and in the contracted forms of estar an entire syllable is lost, which makes it more noticeable. This may contribute to the perception of pa, and the forms of estar (tá, toy, etc) as more socially stigmatized forms than pos. 86 6. Conclusions The goal of this chapter has been to explore the use of discourse particles and contractions in different situations in order to find out whether register and style differences are reflected in the production of speakers’ non-dominant language. Overall, the results of the use of the discourse particles and contractions analyzed across different registers in this study suggest that HLS and SLS vary their production of certain particles according to the situation. First, HLS produce uses of como ‘like’ as quotative and punctor more frequently in informal registers. These non-canonical uses are an innovation of the variety spoken by HLS, which has not been illustrated in the Spanish literature. The possibility of transfer from the quotative and punctor functions of English like is suggested. Second, HLS use full forms such as pues ‘so/well’, para ‘for/so that’, and estar ‘to be’ more frequently in formal situations than the reduced pronunciation of these items as pos, pa, and tar. Third, HLS do not produce punctors in the presentations, which agrees with the view that the presentation is recognized as a more formal register with features different from casual spoken registers. The evidence presented in this chapter suggests that HLS are able to modify their Spanish depending on the exigencies of the register even when their linguistic resources may be more limited in their non-dominant language. The less frequency of production of non-canonical uses in academic registers is an indication of their ability to differentiate among registers with varying degrees of formality. Thus, it 87 does not seem to be the case that the Spanish of HLS analyzed here is a monostylistic variety. The results also indicate that HLS differ among themselves in the use of some discourse particles. This is the case with the connectors so, así que ‘so’, and entonces ‘so/then’, which did not vary across registers but rather across speakers. SLS also show variation across registers as far as the use of discourse particles, although the rate of use of most of the items analyzed is lower than that of HLS, especially in the case of punctors. Furthermore, phonological contractions are not found in SLS data at all. Discourse particles and informal pronunciations are not acquired in the formal context of the language classroom but rather through contact with the local vernacular (Sankoff et al. 1997: 193); thus, the results obtained in this study confirm the expectation that the input and type of exposure to Spanish makes a difference between the Spanish oral production of HLS and SLS. To conclude, the results presented in this chapter show that speakers vary the use of some linguistic features in their non-dominant language depending on register and style considerations and this is reflected in the variation of the use of some discourse particles and contractions. 88 Chapter 3. Lexical Choices 1. Introduction Traditionally, the study of variation has been concerned with the existence of several linguistic variants, that is, alternative utterances to express a similar fact or idea but differing in some additional information; for example, some information qualifying the speaker as to his membership in some social group ─as in the case with social class or regional varieties. The lexicon of a language varies according to the situation of use and the level of formality of the communicative situation. For example, the sentence ‘he came too soon’ and ‘he arrived prematurely’ convey the same information but differ in the choice of words (Hockett 1958; Spillner 1974). From the perspective of register variation, lexical choice depends upon a number of factors such as the purpose of the communicative situation (e.g. transfer of information in a class presentation), level of formality (e.g. the use of colloquial vocabulary in informal conversations), or the topic (e.g. the use of technical vocabulary in the presentation of a scientific problem). The analysis of lexical variation is important for the present study as I try to investigate variation in Spanish as a heritage language and as a second language depending on the register and level of formality. In this chapter, I include the study of formal versus informal vocabulary, the use of technical items, and some cases of lexical transfer including borrowings, code-switches, and creation of new words. Furthermore, I include the analysis of attributive adjectives as a feature of lexical 89 elaboration as well as the analysis of type/token ratio as a measurement of lexical diversity. These lexical features are analyzed in the three registers considered in this study (i.e. conversations, interviews, presentations) and compared across the two groups of speakers (i.e. HLS and SLS). This chapter is organized as follows: In section 2, I discuss the lexical features that are analyzed in this chapter. First, I deal with the use of informal, formal, and technical vocabulary across registers. Next, I discuss some of the relevant issues of transfer in the lexicon of HLS and SLS and propose a model for the study of lexical transfer in my data. Then, I define attributive adjectives and type/token ratio. Section 3 presents the hypotheses related to the use of these features in the production of HLS and SLS across registers. The results and discussion are presented in section 4. Section 5 offers conclusions related to the use of all the aforementioned lexical features across registers and across speakers. 2. Lexical features 2.1. Informal, formal, and technical vocabulary Academic registers are characterized by the use of a formal and technical vocabulary (Achugar 2003; Biber 1988; Solomon and Rhodes 1995) both in English and in Spanish. Conversely, casual conversations are characterized by informal lexicon, such as the use of idioms and slang items (Celce-Murcia and Larsen- Freeman 1983). Studies on the Spanish produced by HLS in academic settings have pointed out the use of colloquial words and expressions considered ‘informal’ and 90 not appropriate for formal or academic situations (Achugar 2003; Hidalgo 1987; Sánchez 1993; Valdés 1995). According to Valdés and Geoffrion-Vinci (1998), the use of ‘colloquialisms’ in academic registers indicates that speakers have not developed formal registers or that they are still in the process of acquiring the ‘formal’ variety appropriate for academic settings. With these considerations in mind, this study includes the analysis of informal, formal, and technical vocabulary and examines the results from the data of HLS and SLS across registers. The term technical vocabulary applies to words or expressions characteristic of a specific area of knowledge or specialization, e.g. linguistics, science, politics. 10 Technical vocabulary is typically used in formal situations such as academic settings (e.g. talk at a professional conference, article in a scholarly journal). Nevertheless, the terms ‘technical’ and ‘formal’ are not synonymous: technical lexicon applies to the specialized vocabulary used within a specific discipline, whereas the use of formal vocabulary is related to the formality of a given communicative situation. For example, we may find technical terms in a very informal conversation between two friends who are doctors and are talking about the cases they are handling. Among other features, formal language is language use characterized by (Ferguson 1959): (a) speech before a passive audience; (b) the assumption of a role by the speaker; (c) the use of a ‘high’ dialect or language in preference to a ‘low’ one. For example, in this study the presentation situation has these characteristics as it is a situation in 10 Definition of ‘technical vocabulary’ adapted from Merriam-Webster Online Dictionary, http://www.m- w.com/ and Diccionario de la Real Academia Española (versión electrónica), http://buscon.rae.es/diccionario/drae.htm. 91 which there is a passive audience, the speaker assumes the role of lecturer or information provider, and a standard or ‘high’ dialect is expected in such situation. Formal English words include polysyllabic words (generally of three or more syllables), many with origins in Latin, Greek, and French (Celce-Murcia and Larsen- Freeman 1983). Academic registers are characterized by the use of a formal and technical vocabulary (Achugar 2003; Biber 1988; Solomon and Rhodes 1995) both in English and Spanish. On the other hand, casual interactions are characterized by the use of informal vocabulary, idioms, and slang items (Celce-Murcia and Larsen-Freeman 1983). Informal vocabulary is generally defined by opposition to formal, that is, vocabulary characteristic of or appropriate to ordinary, casual, or familiar use. Idioms are expressions that have a meaning that cannot be derived from the conjoined meanings of its elements (e.g. ‘kick the bucket’). The term slang refers to an informal nonstandard vocabulary composed typically of coinages, arbitrarily changed words, and extravagant, humorous, or exaggerated figures of speech 11 . Slang is often used by certain social groups and, in that respect, it also functions to establish and reinforce in-group identity and cohesiveness and to exclude non-group members from the conversation (Cohen 1985; Coleman 2004). There are difficulties in trying to identify all the different lexical items used across situations and determine what is technical, formal, or informal. Although it is relatively easy to distinguish some cases, for example specialized lexemes (e.g. using 11 Definition of ‘slang’ adapted from Merriam-Webster Online Dictionary, http://www.m-w.com/. 92 the word cefalea instead of dolor de cabeza ‘headache’), formal (e.g. nupcias instead of boda ‘wedding’), and colloquial terms which can be recognized right away (e.g. tipo or chavo ‘guy’ instead of hombre ‘man’) it is for the most part very difficult to classify words as clearly ‘formal’ or ‘informal’. With these considerations in mind, this study focuses on smaller sets of lexical items that may be more readily identified and compared across situations and speakers. The following paragraphs describe in more detail the aspects that will be analyzed with respect to informal, technical, and formal vocabulary. 2.1.1. Informal vocabulary: Colloquialisms, slang and idioms. My study focuses on a set of common colloquial words, including slang terms and idiomatic expressions, which can be clearly identified in order to examine the production of colloquial lexicon across registers. In order to analyze the use of colloquialisms, it is necessary to determine what items are likely to occur in the speech of HLS and SLS. For this purpose, a number of dictionaries of colloquial words and expressions in Spanish have been selected, especially those that focus on Mexican-American Spanish (Hamel 1996; Hernández-Chávez et al. 1975; Lara Ramos 1996), since in this study HLS are of Mexican-American origin. The use of dictionaries containing details about stylistic information is very useful to determine the degree of formality of the words produced by the speakers in this study. Specifically, Batchelor’s (1994) dictionary of Spanish synonyms is particularly helpful as it consists of an extensive compilation of words (including nouns, adjectives, verbs, and adverbs) and expressions, which are classified by grouping 93 synonyms together. In Batchelor’s dictionary, register is conceived as the most important organizing criterion of the book. Each entry indicates the register level, for which he has four distinctions: level 3, which he calls R3, designates high level of formality, level R2 is neutral, R1 colloquial and R1* vulgar, which according to the author is to be used with caution. This information is used in my analysis as an indication of the degree of formality of words found across situations of use. Furthermore, it is important to verify the choice of these items with native speakers of Spanish for the sake of reliability. This inter-rater check was independently performed by a native speaker of Spanish with linguistics training, who is familiar with heritage varieties of Spanish in the U.S. The following are some of the informal items considered in the present analysis: (1) Informal Item Possible equivalent Meaning in less informal/neutral style nomás solo, solamente ‘only’, ‘just’ ándale exacto ‘right’/ ‘come on’ (also used to get someone to do something fast) órale/ ándale pues de acuerdo ‘okay’ ahorita/orita ahora ‘now’ chavo hombre, joven ‘guy’ puro solamente ‘just’ ‘only’ mero justo, exactamente (used to emphasize or reinforce a statement) 94 Apart from the items presented in 1, this study also examines a variety of words and expressions that are classified as slang (see 2 for examples) 12 . Slang items are characteristic of informal registers and communication exchanges among familiar interlocutors. Often, these are words used mainly by the youth and often include words and expressions that are potentially offensive, as in 2: (2) onda: buena/ mala onda ‘a positive/negative thing or action’ agarrar la onda ‘to understand’ güey ‘stupid’, also may be used among friends as ‘buddy’ pendejo ‘stupid’ cagarla ‘to do something wrong’ joder ‘to bother’ valer (madre) ‘ignore, not care about something’ Finally, within informal vocabulary, this study also examines idioms, which are common sayings characteristic of informal registers. As mentioned above, an idiom is a metaphorical expression whose meaning cannot be derived from the individual meanings of its components. Rather, idioms have a figurative meaning assigned by conventional use. It must be pointed out that some types of expressions that I consider idioms here may be classified as proverbs in other studies. There is no clear dividing line between idiomatic phrases and proverbial expressions and, in fact, idioms and proverbs are frequently analyzed together (e.g. Galván 1995; Serrano and Serrano 1999); even some authors, for example Serrano and Serrano (1999), consider 12 The translations given in 1 and 2 are adapted from the definitions of the online glossary Jergas de Habla Hispana: <http://www.jergasdehablahispana.org>. 95 idioms a type of metaphoric proverb. In general, the difference between the two is that, unlike idioms, proverbs normally seek to give advice or warn against wrong behavior based upon popular wisdom or experience, for example ‘an apple a day keeps the doctor away’, or ‘when in Rome…’. In my data, nevertheless, there are no instances of prototypical proverbs that encapsulate popular wisdom such as the examples just mentioned; therefore, all the expressions that I include in my analysis are classified as idiomatic expressions. The following are examples of idioms used by HLS: (3) G: So, fue a Disneylandia con su novio? [G,HLS,m20,C] ‘So, did she go to Disneyland with her boyfriend? J: Tal vez…Hay más peces en el agua [J,HLS,m20,C] ‘Maybe…There is more fish in the water’. (4) Pienso que es muy raro todo, como pasó lo del 9/11 y luego la guerra así sin razones…eeh, reales. Ahí hay gato encerrado. [J,HLS,m20,I] ‘I think it’s all very strange, like 9/11 happened and then the war like that without… uhm real reasons. I smell a rat there’. In 3, speaker J has just found out that the girl he likes already has a boyfriend, so he has to abandon his hopes but on brighter note, he says, “there are more fish in the water” meaning that there are more girls around. In 4, speaker J expresses his lack of trust for the current administration and says that he can “smell a rat” in connection with 9/11 and the war in Iraq. Both of these are common idiomatic expressions, 96 especially frequent in colloquial Spanish, and which speakers know to mean something different than the literal meaning. Words and expressions such as those illustrated in 1 (colloquialisms), 2 (slang) and 3-4 (idioms) are characteristic of informal or casual registers. However, HLS sometimes produce colloquial items such as these in their oral presentations and written compositions, for which they are corrected by their professors (Valdés 1995). SLS speakers are expected to produce lower rates of informal terms than HLS in all the registers analyzed, even in casual conversations, as they have had mostly formal contact with the language. However, they may have attained a proficiency level to appropriately produce colloquial vocabulary, slang and idioms in casual interactions. It must be taken into account that the types of colloquialisms found amongst SLS will depend on the variety of Spanish spoken in the communities of native speakers that they have interacted with or been exposed to. Therefore, not all SLS may produce colloquial words and expressions typical of Mexican-American HLS varieties and, for this reason, this study should also consider colloquialisms produced by SLS that these speakers may have acquired through contact with informal registers while traveling to a Spanish-speaking country or by interacting with Spanish-speaking people from different parts of the world. The passage in 5 illustrates the use of colloquial vocabulary by SLS during the conversation: (5) K: ¡Ay, joder! ¿Cómo se dice? Me frustra mucho cuando quiero hablar y no me acuerdo de la palabra. [K,SLS,f20,C] 97 ‘Oh, shit! How do you say it? I get so frustrated when I’m trying to speak and I can’t remember the word’. M: Sí, yo también. Ahorita estaba pensando eso... como... es como ¡ay que digo! A veces no tiene ni pies ni cabeza lo que hablo. ‘Yes, me too. I was now thinking about that like... it’s like... oh, what am I saying! Sometimes one can’t make head or tails of what I’m saying’. [M,SLS,m21,C] In 5, speaker K who spent a semester in Spain, uses a slang word joder (which can have a variety of meanings but in the context of 5 could be translated as ‘shit’) to express her frustration when she cannot remember a word. This word is very frequent in informal interactions in Spain, especially among young people. And speaker M uses the colloquial ahorita ‘now’, which is frequent among HLS informal registers, as well as the idiom no tiene ni pies ni cabeza ‘can’t make head or tails’. In order to examine the use of informal vocabulary, the present study compares the frequency of use of these items across situations to determine: (a) whether speakers use colloquialisms, slang, and idioms in presentations, interviews, and conversations; and (b) whether there is a difference in the rate of use of informal items among situations, as expected (i.e. conversations having the highest percentage of use and presentations the lowest). 2.1.2. Formal vocabulary Despite the difficulties involved in the identification of formal vocabulary, it is important for the study of register variation to identify those items that can be 98 classified as formal or characteristic of formal situations in order to be able to compare across situations. For the purpose of identifying formal vocabulary in the data, three methods are employed: (1) use of vocabulary lists and dictionaries with information on register and stylistic ratings for words, (2) inter-rater reliability, and (3) word length. As in the case of colloquial words, Batchelor’s (1994) dictionary of Spanish synonyms, in which the entries are classified according to the formality level, is employed to help determine the degree of formality of the words produced by the speakers in this study. The second criterion was consultation with native speakers of Spanish. For the examination of the vocabulary in my study, Spanish speakers were regularly and systematically consulted at various stages of the data analysis. For the study of formal lexicon, I consulted native speakers first while compiling the list of formal words that I could find in my data and, then, once a body of formal items was compiled as inter-raters for reliability check. Finally, word length is also considered. Formal vocabulary is characterized by longer words in languages such as English, in which many of these words are of Latin or Greek origin and normally longer than the Anglo-Saxon equivalent (e.g. terminate vs. end). However, in romance languages such as Spanish, it may not be the case that longer words from Latin or Greek indicate formal vocabulary as the majority of the lexicon derives from Latin (or Greek via Latin). There are nonetheless several studies that deal with word length not specifically related to the 99 origin of the word and could be useful to my study of formal vocabulary across registers in Spanish. For example, Biber (1988) found that word length correlates with communicative situations that require high informational focus (Biber 1988: 105) since longer words convey more specific and specialized meanings than shorter ones (Biber 1988; Chafe and Danielewicz 1987). Also, there has been research investigating the relation between word length and word frequency and it appears to be the case that more frequent items are shorter than less frequently used ones (Zipf 1949; O’Regan and Jacobs 1992). In more recent studies, it has also been shown that word length affects both processing (Hudson and Bergman 1985; Juhasz and Rayner 2003; Balota et al. 2004) and production time (Levelt et al. 1999; Santiago et al. 2000) and, therefore, longer words are less likely to be used for everyday interactions, but rather to be confined to written registers and more formal interactions. With these findings in mind, I am taking into account word length as a factor that may help determine the formality of words in Spanish. For the most part, the words classified as formal in my study consist of three or more syllables (see examples in 11). Based on these three methods, I examined the vocabulary in the data to determine what items would be classified as formal. In examples 6-10, I illustrate the use of words and phrases from my own data. These are rather formal and can be compared with 6’-10’, in which speakers express a similar message in a more informal way: 100 (6) utilizar métodos anticonceptivos [W,HLS,f20,I] ‘to use methods of contraception’ (6’) Sí, por ejemplo, como los condones […].[B,HLS,f19,I] ‘Yes, for instance like condoms […]’ (7) Tiene negocios para recaudar fondos para la escuela [C,HLS,f20,P] ‘He has business to raise funds for the school’ (7’) […] y yo dije, pos dónde vamos a sacar dinero para la escuela [L,HLS,m19,C] ‘and I said, so where are we going to get money for school?’ (8) […] tiene relaciones íntimas.[C,HLS,f20,P] ‘[…] (he) has intimate relationships’ (8’) se acostó con dos o tres muchachos [B,HLS,f19,I] ‘(she) went to bed with two or three guys’ (9) Lo encarcelaron por 105 días [E,HLS,f20,P] ‘He was incarcerated for 105 days’ (9’) Y no es justo que los agarren y los metan en la cárcel [D,HLS,f19,I] ‘And it is not fair that they catch them and they put them in jail’ (10) Mi hermana es indocumentada [D,HLS,f19,I] ‘My sister is undocumented’ (10’) Hay mucha gente sin papeles [W,HLS,f20,I] ‘There are a lot of people without papers’ Some of the formal words and expressions are: 101 (11) Formal Possible informal equivalent(s) conversar ‘converse’ platicar ‘talk’ identificar ‘identify’ saber qué/quién ‘know what/who’ considerar ‘consider’ pensar, creer ‘think’ mencionar ‘mention’ decir ‘say’ agrandar ‘to enlarge’ hacer más grande ‘make bigger’ aceptar ‘accept’ decir que sí ‘say yes’ recaudar ‘collect’ sacar, recoger ‘take out’,‘pick up’ fondos ‘funds’ dinero ‘money’ acomodado ‘wealthy’ rico ‘rich’ significante ‘significant’ importante ‘important’ ceremonia nupcial ‘wedding ceremony’ boda ‘wedding’ All words classified as formal are counted in each situation and then frequencies are compared across situations to check whether interviews and presentations differ significantly and also whether formal items occur in the conversations. Additionally, formal vocabulary is compared between HLS and SLS to find out whether speakers differ in usage. 2.1.3. Technical vocabulary In order to identify instances of technical vocabulary in the data, the present study analyzes the lexicon produced in relation to the different subjects of the presentations. Namely, the data from class presentations of HLS and SLS can be classified under three general themes: (1) language and literature, (2) business, (3) social issues and culture of the Spanish-speaking world. In addition, dictionaries and 102 glossaries of specific subjects are also used to identify technical lexicon according to the subject (e.g. literature, business, linguistics, etc.). The following words exemplify the use of technical words and expressions produced by HLS and SLS in presentations: (12) trama ‘plot’ (language and literature) préstamos lingüísticos (language and literature) aliteración ‘alliteration’ (language and literature) franquicias ‘franchise’ (business) acciones ‘stocks’ (business) inversionistas ‘investors’ (business) The examples in 12 refer to the fields of language, literature, and business. As far as the presentations related to social and cultural issues of the Hispanic countries, it is somewhat more difficult to differentiate technical items since most of the words and expressions typically employed are not unique to that field. For example, some of the words and expressions employed by HLS and SLS in this type of discourse include concepts such as discriminación sexual/racial ‘sexual/racial discrimination’, justicia ‘justice’, sindicatos ‘work unions’, salario mínimo ‘minimum wage’. We may argue that these words are not technical in the same sense that aliteración ‘alliteration’ or troqueo ‘trochee’ are technical within the field of literature. All the presentations in this study are informational in purpose but the difference is that most of the presentations dealing with linguistics and business as well as some on literature are 103 more technical than the ones related to social issues. Consider the following examples from the presentations by HLS and SLS: (13) Linguistics: Algunas características de los dialectos meridionales son el seseo, es decir, la pronunciación del sonido [ θ] como [s]; la aspiración del sonido fricativo [s] como [h] ante consonantes, por ejemplo obi[h]po en vez de obi[s]po. ‘Some characteristics of the meridional dialects are the pronunciation of the [ θ] sound as [s]; the aspiration of the fricative sound [s] as [h] before consonants, for example, obi[h]po instead of obi[s]po’. (14) Social Issues: El reto verdadero era formar una unión para la protección y servicios de los trabajadores campesinos, la cual pudiera brindarles servicios de apoyo y seguir con la batalla de la injusticia de los salarios, con la coordinación de campañas contra la discriminación racial y económica que existe con los campesinos. ‘The true challenge was to form a union for the protection and support of the farmers, which could offer support services and continue the battle against the injustice of the wages, with the coordination of campaigns against racial and economic discrimination that exists with farmers’. In example 13 a sentence from a linguistic presentation, we find technical words specific to linguistics such as seseo, the pronunciation of the letters <z> and <c> before e and i as /s/, or aspiración the weakening of the articulation of a consonant to 104 an aspirated sound /h/. However, in example 14, we find words which are not as technical but which are, however, common in a discussion of social issues such as salario ‘salary’, or discriminación racial ‘racial discrimination’. Biber (1988) in his typology of English texts calls texts such as the example in 13 ‘scientific presentation’ and texts similar to 14 ‘learned exposition’. Biber finds that these texts differ in the level of technical and abstract style and his analysis shows that most of the texts that coincide with ‘learned exposition’ belong to social sciences and humanities. Since all speakers were recorded in a class oral presentation, we can compare the production of technical terms across speakers to determine whether all speakers have similar patterns as far as use of technical vocabulary according to their presentation subject, and whether there are differences in the frequency of use of technical terms between HLS and SLS. As far as comparison among situations of use, it is possible to compare segments of the interviews and conversations with the presentations. In the interview situation, one of the questions asked speakers to talk about the Spanish courses they were currently taking, specifically, speakers were asked about the topic of their presentations. In the same way, during the conversations, most speakers talked about their classes and assignments, including their presentations in Spanish. This offers the possibility of comparing the use of technical words when the topic is similar but the audience, setting, purpose and level of formality are different. 105 2.2. Lexical transfer This study of the vocabulary produced by HLS and SLS across different registers needs to take into account the possibility of lexical transfer from English into Spanish since transfer is a common phenomenon in situations of language contact (as in the case of heritage speakers) and in situations of language acquisition (as in the case of second language learners). This chapter analyzes several types of transfer phenomena, which I classify as code-switches, borrowings, and lexical creations. When stretches of English elements appearing in Spanish discourse are several words long, it is easy to identify them as cases of switching (Otheguy et al.1989). However, it is hard to differentiate between single-word code-switches and single-word borrowings. The difference, according to Poplack (1987) and Poplack et al. (1989), lies in the integration of these items: borrowings pattern according to the grammar of the recipient language while code-switches pattern according to the donor language. According to Poplack and Meechan (1998), most of the lone- English items found in contact varieties are borrowed, even if used only once by a speaker (which the authors call “nonce-borrowings”). The authors refer to various studies of lone English-origin words in a number of languages (Adamar and Tagliamonte 1998; Budzhak-Jones 1998; Meechan et al. 1996), which claim that most of these words pattern as the recipient language and therefore are borrowings. For the study of lexical transfer in Spanish produced by HLS and SLS, I propose a model that differentiates between single switches and borrowings (Poplack 106 1987; Poplack et al. 1989). I consider all phonologically integrated single-word items as cases of borrowings, following the model of lexical transfer presented by Otheguy et al. (1989), who take into account phonological assimilation as the criteria to distinguish switches from borrowings. Therefore, this chapter analyzes cases of lexical transfer classified as single-word switches (items that preserve English phonology) and single-word borrowings (items adapted to Spanish phonology). In this latter group, Otheguy et al. (1989) differentiate between loans (the transfer of forms with their meanings) and calques (the transfer of meanings only). Additionally, I follow Silva-Corvalán’s (1994) study, which extends Otheguy’s typology to the study of multiple-word cases of loans and calques. Moreover, Silva- Corvalán distinguishes calques that do not alter the semantic and/or grammatical features, such as máquina de contestar ‘answering machine’ (instead of contestador automático), and lexico-syntactic calques, such as tener un buen tiempo ‘to have a good time’ (instead of pasarlo bien), which do. Furthermore, my typology includes a category of items that I call lexical creations. This category is similar to loans in that the speaker transfers the form and the meaning. However, the difference is in the frequency and the degree of social integration and spread among speakers. Words such as troca ‘truck’ or lonche ‘lunch’ are well established among HLS communities (even among predominantly monolingual Spanish-speakers) and are considered loans. However, certain words such as endurar ‘endure’ or enviromento ‘environment’ occur rarely and are not widespread but idiosyncratic to individual speakers and, furthermore, not included in 107 any dictionary of Spanish. These items are here classified as lexical creations. Figure 10 illustrates the classification of items that I analyze in this section: LEXICAL TRANSFER Code-switches Lexical-creations Borrowings - single or multiple-word - single-word retain English phonology - transfer of form and meaning - not shared by all speakers - not frequent Loans Calques -single or multiple-word -single or multiple-word -transfer of form and meaning -transfer of meaning only Figure 10. Lexical Transfer Categories The following examples illustrate each of the lexical phenomena analyzed in this section: code-switches (15), loans (16), calques (17), and lexical creations (18) produced by HLS and SLS. These are underlined in the examples: (15) ¿Puedo ir al wedding? [Co,SLS,f20,C] ‘Can I go to the wedding?’ (16) En mi mente yo pensé que me iba a pegar la troca, so ah… frené y cuando frené se meneó la motocicleta y se resbaló so… [J,HLS,m20,C] ‘In my mind I thought that the truck was going to hit me, so I braked and when I did the bike swerved and slipped so...’ (17) No realizan de que toma mucho trabajo [el sueño americano].[G,HLS,m20,I] They don’t realize that it takes a lot of work [the American dream]’. 108 (18) No puedo envisionarme, en 10 años, viviendo en Europa [Cr,SLS,m22,I] ‘I can’t envision myself, in ten years, living in Europe.’ 2.3. Attributive adjectives Attributive adjectives are used to modify nouns within noun phrases (e.g. my silly smile) as opposed to predicative adjectives (e.g. my smile is silly). According to some researchers (Finegan and Biber 1994, 2001; Hudson 1994), more elaborate and literate registers are characterized by the use of a high frequency of attributive adjectives. However, the term ‘attributive’ in Spanish is used differently by different scholars. For example, Bello (1847) referred to all adjectives that appear immediately before the noun as predicados ‘predicates’ or epítetos ‘epithets’, and he calls atributos ‘attributes’ those adjectives following a copulative verb. For some authors (e.g. Butt and Benjamin 1994; Smith 1993), attributive adjectives are those that replace a noun + de ‘of’ (presión sanguínea ‘blood pressure’ = la presión de la sangre ‘pressure of the blood’), which are, on the other hand, called relacionales ‘relational’ by other authors (Judge and Healey 1983; Bosque 1993). I will follow the Real Academia (1973) and Demonte’s (1999) definition of attributive adjective in Spanish as those adjectives which are adjacent to a noun. Traditional Spanish grammars distinguish a subclass of adjectives and pronouns: these are possessives, quantifiers and demonstratives. Demonte (1999) differentiates between these, which she calls ‘determinantes’ (example 19) and ‘calificativos’ (example 20) since the former are a closed grammatical class of words without lexical meaning whereas the latter is an open lexical class: 109 (19) Mis/Estas/Dos casas ‘My/These/Two houses’ (20) Mi casa grande/bonita/ruidosa ‘My big/nice/noisy house’ I follow Demonte’s distinction between determinantes (possessives, quantifiers, and demonstratives) and adjetivos calificativos, and, in this study, I only consider the latter as in 20. In Spanish, some adjectives in attributive position can appear before or after the noun, as in 21: (21) Un lejano ruido/ Un ruido lejano ‘A distant noise’ Many factors combine to determine whether adjectives are placed in a pre- nominal or post-nominal position. Demonte (1999: 147), following Bello (1847), points out that the position of the adjective with respect to the noun is related to the way in which the adjective modifies the noun, mainly by restricting the noun (in post-posed adjectives) or by simply highlighting one characteristic of the noun that it modifies (non-restrictive as in pre-posed adjectives). Demonte adopts this classification of adjectives as restrictive and non-restrictive. Demonte illustrates this difference with some pairs of examples (1999: 146), two of which are shown in 22: (22) a. La niña tímida—La tímida niña ‘The shy child’ b. Las paredes verticales—Las verticales paredes 110 ‘The vertical walls’ The first pairs of 22a and 22b refer to a specific subgroup of objects, restricted by the adjective, that are selected from a universe in which other girls and walls are supposed to exist. However, the second pair of sentences, there is no restriction whatsoever and what is highlighted is a characteristic or trait of the object which may be intrinsic to it (as in 22b) or not (as in 22a). Even though many non-restrictive adjectives can be placed either before or after the noun, the choice of position is not random, it is a complicated issue and there is a vast body of research investigating which factors play a role in determining this choice (such as meaning, style, and convention). For the purpose of my study, however, I do not take into consideration whether an adjective precedes or follows the noun. In my analysis, I consider only whether an adjective appears in attributive, as opposed to predicative position. Adjectives are considered a feature of elaboration because they serve to expand and elaborate the information presented in a text (Chafe 1985; Silva- Corvalán 1994). Attributive adjectives occur more frequently in elaborate registers than predicative adjectives because they are highly integrative in their function, whereas the latter might be considered more fragmented (Biber 1988: 237). Thus, attributive adjectives are expected to occur more frequently in registers that are more elaborate, such as the academic presentation. In order to determine whether this is the case in the speech of HLS and SLS, this study compares the frequency of attributive adjectives across situations. 111 2.4. Type/token ratio Type/token ratio is a statistical measure which computes the number of different lexical items in a text. Type/token ratio expresses the ratio between all the running words (tokens) and all the different words (types) in a text. The number of running words shows nothing but the length of the text, while the number of types shows number of different word forms in the text. Type/token ratio represents lexical diversity and elaboration and is commonly used as a measure of the vocabulary richness of a text (Biber 1988; Finegan and Biber 2001). Several studies have shown that literary texts have richer vocabulary than non-literary texts. For example, McEnery and Wilson (1996) examined three extensive corpora, one of them containing literary texts (fictional writings of a variety of authors) and the other two composed of non-literary miscellaneous material. The authors found that the corpus containing literary texts had the richest vocabulary. Other studies have also shown high levels of lexical diversity in formal academic texts (Biber 1986; Chafe and Danielewicz 1987). For instance, Biber (1988, 1995) shows that a high type-token ratio reflects high density of information and represents a more careful word choice and more exact presentation of the information. Lexical diversity, however, is higher in written texts than oral due to the restrictions of online production (Biber 1988: 238). Nevertheless, we can expect differences in the level of lexical diversity depending on the characteristics of the oral register. For instance, it is reasonable to expect higher type/token ratio in oral academic texts than in private conversations as the former is a type of register that 112 requires a more precise presentation of the information, whereas the latter is not typically informational in purpose. For Spanish, Valdés and Geoffrion-Vinci (1998) found that bilinguals (heritage speakers) displayed greater lexical diversity in academic argumentations than in the informal narration of personal experiences. The statistical measure of type/token ratio is examined in my study in order to investigate whether HLS and SLS have different degrees of lexical diversity depending on the register. However, type/token ratios are very sensitive to the size of corpora. A 1,000 word article might have a type/token ratio of 40%; a shorter one might reach 70%, 4 million words will probably give a type/token ratio of about 2%, and so on. Since this study compares texts of varying length, caution is needed when computing this measure in order to avoid skewed results in comparisons of short and long texts. Therefore, in order to be able to compare data sets of different sizes, I will calculate the number of types in the first four hundred words of each text, regardless of the total length, following Biber (1988). 3. Hypotheses The main hypothesis tested is that both HLS and SLS will show lexical variation depending on the register. More specifically, I expect the following with respect to the lexical features analyzed: 1- As far as the use of informal, formal, and technical items, we may expect a higher frequency of formal vocabulary and technical words associated with a particular subject in class presentations than in interviews, whereas conversations 113 may show few or no use of technical items. On the other hand, we may expect the use of informal lexicon (colloquial words, slang, and idioms) in the conversation situation, especially in the data from HLS. The SLS group is expected to produce lower rates of informal terms than HLS in all the registers analyzed, even in casual conversations, as they have had mostly formal contact with the language. Nevertheless, it is expected that SLS produce some informal vocabulary, slang and idioms in casual interactions. 2- In general, I expect conversations to have a high frequency of borrowings and lexical creations as well as more code-switching. On the other hand, in academic and more formal registers, it is expected that HLS and SLS may try to avoid the use of English words (Valdés and Geoffrion-Vinci 1998). Furthermore, HLS are expected to produce more loans and calques since their variety of Spanish may already use these items due to the societal contact with English. For example, loans such troca ‘truck’, and calques such as carpeta ‘carpet’ are well-established in Spanish-speaking communities in the U.S. Conversely, SLS are expected to produce more switches to English, especially in informal settings to keep the flow of the conversation when they do not know or remember the word in Spanish. In academic settings, SLS may produce more lexical creations instead of English switches. 3- Due to the fact that academic presentations are planned and informational in nature, a higher frequency of attributive adjective is expected in this situation than in interviews or conversations. Additionally, interviews may display higher rates of attributive adjectives than conversations. 114 4- As far as lexical diversity, richer variety of vocabulary is expected in academic texts. Therefore, presentations will show higher type/token ratios than conversations and interviews for both HLS and SLS. 4. Results and discussion In this section, I first report the results obtained for each of the linguistic features analyzed in this chapter. In analyzing and reporting the results of this chapter, I follow the same criteria as in the previous chapter and as explained in the methodology section of chapter 1, namely, all results are normalized to 15 minutes per speaker, so that lexical features can be compared across different situations. A discussion of these results follows. 4.1. Informal, formal, and technical vocabulary The words and expressions used by HLS and SLS were analyzed and classified as informal, formal, or technical according to the criteria explained in section 2.1. of this chapter. In reporting the results for technical terms, the following method was adopted: a) Presentations are analyzed in their full length; b) for the interviews, I only included those segments in which speakers were asked to talk about the Spanish courses and their presentations; c) for the conversation situation, I considered the passages where speakers talked about their Spanish presentations, including their classes and assignments in Spanish. Then, the results are normalized as with the rest of the features. This offers the possibility of comparing the use of 115 technical words when the topic is similar but the audience, setting, purpose, and level of formality are different. The following figures show the rates of use of informal vocabulary (colloquialisms, slang terms, and idiomatic expressions), formal, and technical items across situations. Figure 11 displays the rates of use of HLS and figure 12 shows the results of SLS. Figure 11. HLS use of informal, formal, and technical lexicon 66.8 31.4 9 18.7 73.1 71.2 8.3 11.4 102.2 0 20 40 60 80 100 120 conversation interview presentation rate of use informal formal technical Figure 12. SLS use of informal, formal, and technical lexicon 111.7 40.2 19.7 11.3 52.4 63.9 4.3 5.14 90.3 0 20 40 60 80 100 120 conversation interview presentation rate of use inform al form al technical 116 Figures 11 and 12 show that the results of the use of informal, formal, and technical vocabulary confirm the hypothesis that conversations will have higher rates of informal vocabulary as opposed to interviews and presentations in which speakers produce more formal lexicon. This is the case for both groups of speakers, HLS and SLS. For example, as we can see in figure 11, HLS have a rate of informal vocabulary use of 111.7 in conversations as opposed to only 19.7 in presentations. On the other hand, they have a lower rate of formal vocabulary use in conversations (11.3) than in presentations (63.9). In the case of SLS the pattern is similar, as we can see in figure 12. Namely, the rate of use of formal vocabulary is much lower in conversations (18.7) than in presentations (71.2), while the contrary is true of the use of informal vocabulary with a higher rate in conversations (66.8) than in presentations (9). Furthermore, technical terminology is more frequent in the presentation register for both groups, as expected, although there is still a higher rate of use of technical terms in interviews than in conversations. The results shown in figures 11 and 12 are statistically significant according to the Chi-square tests performed ( χ 2 = 214.095 (4) p= .000 for HLS and χ 2 = 184.022 (4) p= .000 for SLS). 4.2. Lexical transfer Figures 13 and 14 show the results obtained for HLS and SLS respectively as far as lexical transfer features, namely loans, calques, creations, and switches. 117 0 20 40 60 80 100 interview presentation rate of use loan 13.8 14.9 8.4 calque 19.2 12 6.6 creation 5.7 8.1 19.7 switch 85.7 18.9 6.6 conversation Figure 13. HLS rate of lexical transfer across registers 0 20 40 60 80 100 120 interview presentation rate of use loan 5.2 2.6 4 calque 15.3 6.6 6 creation 14.8 18.1 18 switch 101.3 25.5 8 conversation Figure 14. SLS rate of lexical transfer across registers Figure 13 shows that HLS produce higher rates of loans, calques, and switches in conversations and interviews than in presentations. However, the 118 difference between conversations and interviews is not statistically significant as far as the use of loans and calques (p> .3), whereas the difference in the use of switches to English is statistically significant across situations (p<.01). On the other hand, lexical creations occur more frequently in presentations than in interviews and conversations, which is also statistically significant across situations (p<.01). As shown in figure 14, SLS also produce significantly more switches in conversations than in any other situation. However, the differences observed for the rest of the features are not statistically significant (p> .2). To summarize, figures 15 and 16 display the percentages of use for the lexical features reported thus far. We can compare these percentages across registers as well as comparing the production of HLS in figure 15 to that of SLS in figure 16. 0 10 20 30 40 50 60 70 80 90 100 informal formal technical loan calque creation switch percentage conversation interview presentation Figure 15. HLS percentages of lexical features by register. 119 0 10 20 30 40 50 60 70 80 90 100 informal formal technical loan calque creation switch percentage conversation inteview presentation Figure 16. SLS percentages of lexical features by register Figures 15 and 16 show that there are differences across registers for most of the lexical features produced. These differences are overall statistically significant ( χ 2 = 317.298 (12), p= .000 for HLS and χ 2 = 314.740 (12), p= .000 for SLS). However, as I mentioned earlier, statistical analysis performed on individual features show that the differences between conversations and interviews in loans and calques produced by HLS are not statistically significant. For SLS differences are statistically significant except for loans and creations. Both groups behave in a similar way as far as the distribution of the production of these features in the different registers. However, one of the differences that can be observed comparing figures 15 and 16 is that SLS produce lexical creations fairly evenly across the three situations, while HLS do so predominantly in the presentations. Additionally, SLS produce the least percentage of loans in the interview situation, which is not the case with HLS. 120 Finally, figure 17 shows the rates of use of the lexical features analyzed by speaker group regardless of the register: 0 20 40 60 80 100 120 140 160 180 200 informal formal technical loan calque creation switch rate of use HLS SLS Figure 17. Rate of use of lexical features by group of speaker If we compare the production of SLS to HLS, we can see that SLS utter an overall lesser amount of informal words (114), loans (16), and calques (34) than HLS (180, 45, and 44, respectively), whereas SLS produce more formal (128) and technical (139) lexicon as well as more creations (56) and switches (173) than HLS (104, 109, 39, and 132, respectively). The biggest differences are in the use of informal words and loans for which SLS have much lower rates of use than HLS. 4.3. Attributive adjectives The results confirm the hypothesis that attributive adjectives occur more frequently in presentations than in interviews and conversations. Figure 18 shows the percentage of attributive adjectives across situations for HLS and SLS. 121 0 10 20 30 40 50 60 70 80 90 100 presentation interview conversation percentage HLS SLS Figure 18. Percentage of attributive adjectives in different registers Figure 18 shows the percentage of all the attributive adjectives that HLS and SLS produced and how they are distributed over the three registers. We can see that both groups produce most of the attributive adjectives in the presentation situation, which confirms the initial hypothesis of variation across registers also for this feature. Nevertheless, the total number of attributive adjectives is much lower than that of predicative adjectives for all situations including the presentation situation. 4.4. Type/token ratio Type/token ratio is computed counting the number of different lexical items that occur in the first four hundred words of each speaker for each situation and then divided by four. This method of computing type/token ratio was adopted, following Biber (1988), in order to be able to compare texts of varying length. As I mentioned earlier, the relation between the number of types in a text and the total number of words is not linear; therefore, we cannot compute type/token ratio over the entire length of the text because we would get skewed results with shorter texts always 122 having higher ratios than longer ones. The following table shows the results for the feature of type/token ratio. Type/token ratio Mean for all speakers Minimum value Maximum value Range Standard deviation HLS Mean HLS Conversation 39.3 35 44.5 9.5 3.6 Interview 44.0 39.8 48.1 8.3 2.6 Presentation 50.1 44.2 53.7 9.5 2.8 SLS Mean SLS Conversation 36.2 30.1 45.2 15.1 4.8 Interview 39.3 34.4 46.5 12.1 3.8 Presentation 47.5 42.2 56.2 14 4.1 Table 10. Mean values for the type/token ratios of HLS and SLS across registers Table 10 shows that there is a higher type/token ratio in the presentation situation. Furthermore, interviews have in turn higher ratios than conversations. The results also show that HLS and SLS are similar in that there is a gradual increase in the lexical diversity from conversations to presentations. However, SLS have overall lower ratios than HLS. 4.5. Discussion The results of all the features analyzed in this chapter confirm the hypothesis of variation across registers in the Spanish of HLS and SLS. HLS vary their use of informal, formal, and technical vocabulary across the three registers analyzed in the expected direction, that is, the more formal the register the less informal the lexicon. It is important to notice, however, that HLS do produce informal lexicon in the presentation situation (19.7 rate of use or 11.5%), which is not expected due to the fact that this is an academic, formal situation. These results are similar to previous studies on HLS academic production. For example, Valdés 123 and Geoffrion-Vinci (1998) focus their study on class presentations and they also find that students use a number of colloquial words and expressions considered informal and not appropriate for formal or academic situations (cf. Achugar 2003; Valdés 1995). According to Valdés and Geoffrion-Vinci (1998), the use of colloquialisms in academic registers indicates that speakers have not developed formal registers or that they are still in the process of acquiring the formal variety appropriate for academic settings. In my study, a detailed analysis of the production of informal lexicon in presentations of HLS reveals that only certain colloquialisms are produced. In reporting the results, I put together colloquial terms, slang, and idioms under the category of informal lexicon, as I explained in section 2.1. of this chapter. Nevertheless, only certain colloquialisms, in particular nomás, and ahorita occur in presentations, while slang terms and idioms are not produced in this situation. This may indicate that the HLS in my study, as in the case of the bilinguals in Valdés and Geoffrion-Vinci’s study, might also be in the process of forming an appropriate academic register in their non-dominant language. Table 11 shows the distribution of colloquialisms, slang, and idioms across situations for the HLS group: Table 11. Rate of use of colloquial, slang, and idioms across situations by HLS The following examples from the presentations illustrate the use of colloquialisms: HLS Informal Lexicon Colloquial Slang Idioms Total Presentations 21 0 0 19.7 Interviews 25.5 14.4 0.2 40.2 Conversations 75.8 34.2 1.8 111.7 Total HLS 42.5 19.5 0.7 62.7 124 (23) Ahorita es…hay una oportunidad muy importante para invertir en México. [G,HLS,m20,P] (business presentation) ‘Now uh… there is a very important opportunity to invest in Mexico’. (24) No ha logrado lo que quería, nomás es ama de casa. [L,HLS,m19,P] (literature presentation) ‘She did not get what she wanted, she is only a housewife’. The words nomás and ahorita illustrated in examples 23 and 24 account for most of the colloquial words produced by HLS in presentations which are reported in the results section. Additionally, speakers show variation between words that are just colloquial and items such as those classified as slang in this study since these are mostly produced in conversations, especially words categorized as vulgar by the dictionaries consulted (e.g. Batchelor’s dictionary marks slang words that are especially vulgar or potentially offensive with a star, for instance joder* ‘to bother’). Speakers show awareness of register differences trying to produce a more formal vocabulary in academic situations as evidenced in self corrections illustrated in the following examples: (25) Pero nomás lleg--solamente llegó a octavo grado en su educación. [D,HLS,f19,P] (social and cultural issues) ‘But he only go--only got to the eighth grade in his education’. (26) Una noche ella descubrió a su amiga con el chav--con el hombre que ella quería. [L,HLS,m19,P] (literature) 125 ‘One night she discovered her friend with the guy--with the man that she loved’. In example 25, the speaker is talking about how César Chávez had to work from an early age and was not able to finish his education, getting only to eighth grade. She starts the sentence saying nomás ‘only’, but right before finishing the verb she corrects herself and produces solamente ‘only’, which is not as informal as nomás. Similarly in 26, the speaker starts to utter the informal word chavo ‘guy’ but immediately corrects himself and restarts the phrase substituting for chavo ‘guy’ the more neutral hombre ‘man’. Furthermore, there are fewer colloquialisms and slang in interviews too and, as with presentations, speakers seem to be monitoring their lexical production and trying to avoid slang. There are several instances in the interview data in which speakers seem to realize that a word is not appropriate as they are uttering it and consequently pause and do not finish it as in 27. On other occasions, as in the previous two examples, speakers correct themselves by substituting the slang item for a more neutral one, as in 28: (27) y nos tratan así, como si fuéramos pendej…[pause] [G,HLS,m20,I] ‘and they treat us like that, as if we were stup…[pause]’ (28) pues ahí se la pasan de huevo… de borrachos, nomás ahí sin hacer nada. [L,HLS,m19,I] ‘well, they are all the time as…. as drunks, there, not doing anything.’ 126 The word pendejo in 27 is a vulgar term of abuse. To call a person a pendejo is essentially to call him stupid, but it also carries implications of stubbornness and ineptitude. The dictionaries and sources consulted rank it between moderately to highly offensive (Batchelor 1994; The American Heritage Spanish Dictionary 2001). The speaker seems to realize that the word is not appropriate as he stops himself before finishing the word and does not finish the sentence. In 28 the speaker almost says the word huevones referring to some of his classmates, a vulgar word that means lazy, or good for nothing, but he immediately substitutes for it borrachos ‘drunks’, which does not have the vulgar connotations of huevones. In conversations, however, there are no instances of self corrections or lexical substitutions to avoid a vulgar word. In the informal lexicon of HLS, it is also interesting to notice that there are very few idioms: taking into account the total number before normalization, only one idiom occurs in an interview and a total of five idioms in conversations (see examples in 3 and 4). Furthermore, four of these five idioms, are produced by one of the speakers. Many studies in second language acquisition have pointed out that idioms are difficult to learn (e.g. Cooper 1998; Irujo 1986) since they do not make sense literally and require an understanding of the figurative meaning of the phrase. Some studies have also pointed out that bilinguals too have difficulty in mastering idioms in their non-dominant language (Wilson 1984). In the case of the HLS in this study, although Spanish is their home language and they are fluent speakers, it is still their non-dominant language. Therefore, HLS may not know enough idioms in 127 Spanish or may not have the confidence to produce idioms spontaneously in discourse even if they can understand idioms when they hear them in context. The difference in the production of idioms among the speakers may be related to the amount of language use in the familiar context. Even if all the speakers have similar characteristics as far as language background and language use (see Appendix A), it may still be the case that some speakers may use more Spanish on a daily basis. Speaker J, who produces most of the idioms, lives with his parents and commutes to campus on a daily basis. His parents are Mexican immigrants who came to the U.S. when they were adults (the father was in his 30s and the mother in her 20s) and, according to speaker J, they speak little or no English. Therefore, J speaks Spanish on a daily basis at home with his parents as well as other relatives. Conversely, the rest of the HLS speakers are living in or around campus and only go back home for the weekends or holidays. It is not surprising that SLS produce an overall lesser amount of informal vocabulary and almost no idioms (a total of two in the conversations, see an example in 5). SLS have had primarily formal input, that is, they have learned Spanish in class as a foreign language. However, typically, colloquialisms, slang, and idioms are learned through interaction with native speakers in everyday situations and very little of it is actually learned through the formal context of the language classroom. Nevertheless, the SLS in this study either have friends who are native speakers and/or have traveled to Spanish-speaking countries, therefore, they do have exposure to some informal input outside of the classroom, which may account for the 128 production of some colloquialisms and slang as the results show. On the other hand, SLS produce higher rates of formal and technical vocabulary during interviews and conversations. The results discussed so far suggest that although SLS are fluent in Spanish and can compete with HLS in academic situations, still the difference in exposure to informal situations is an important factor accounting for the differences observed between both groups in the production of formal, informal, and technical vocabulary. SLS realize that they lack the ability to produce native-like Spanish in casual interactions and they share their frustration for their lack of fluency with other SLS during the conversation. There are several examples of this, as the following passage illustrates: (29) Co: Sí pero Mark es…. [Co,SLS,f20,C] ‘Yes, but Mark is…’ Ge: …Es un asshole [Ge,SLS,f20,C] ‘…He’s an asshole’ Co: Sí. ¿Cómo se dice? No sé las palabras malas en español. ‘Yes, how do you say that? I don’t know bad words in Spanish’. Ge: Yo tampoco… No es bueno. ‘Me neither… It’s not good’. Co: No, yo tengo que aprender las palabras malas… quiero, quiero aprender las palabras de la gente… ‘No, I have to learn the bad words… I want, I want to learn the words of the people…’ 129 Ge: Coloquial ‘Colloquial’ Co: Sí, coloquial. ‘Yes, colloquial’. Ge: Sí, yo también. Hay muchas cosas que aprender, me cuesta mucho porque lo que quiero hacer es hablar mucho, you know, con la gente… pero en situaciones informales es muy muy difícil. ‘Yes, me too. There are many things to learn, it’s hard for me because what I want to do is to talk a lot, you know, with the people…but in informal situations is very very difficult’. Co: I know, lo mismo aquí…I mean, para mí. Ojalá aprendemos esto en clase, ¿verdad? [risas]… Es…sería más divertido. ‘I know, same here, I mean, for me. I wish we would learn this in class, right? [laughs]… It’s… It would be more fun’. The second set of results, which concerns the use of loans, calques, creations, and switches, shows that the feature that varies the most among registers is switches to English. This is the case for both groups of speakers, HLS and SLS, with conversations having the highest rate of switches to English. These results suggest that speakers try not to switch to English in those situations where Spanish is expected, even if switching would be the easier way out when they cannot remember or do not know how to express something in Spanish. In this study, those situations are the interview since the interviewer is a native speaker (and an instructor) of 130 Spanish who does not belong to the speakers’ community and, mainly, the academic presentation since not only the content but also the Spanish of the presentation is being evaluated (Achugar 2003). The fact that HLS produce more lexical creations in the presentations is likely related to their attempt to not resort to English when it would not be appropriate. Evidence for this can be found in the hesitations and lexical creations based on the English equivalent that speakers produce in presentations, as illustrated in the following examples: (30) En 1985 Chávez y la UFW organizaron una huelga entre los…los… --[in a lower voice] cómo se dice--… piqueadores de uva.[D,HLS,f19,P] ‘In 1985, Chavez and the UFW organized a strike among the grape pickers’. (31) los precios bajaron un 15% y también el rato de ocupancia [de los hoteles] bajaron un 25% [Co,SLS,f20,P] ‘prices decreased 15% and also the rate of occupancy [of hotels] decreased 25%’. In 30, a HLS is having difficulty in finding the word for fruit pickers and after some hesitation, she produces piqueadores in Spanish (instead of the standard recogedores or recolectores), which is clearly based on the form of the English word. Still, she decides to give this word a try instead of saying the word in English. The example in 31 is produced by a SLS who is giving a class presentation about business in Spanish. She produces rato de ocupancia, which does not exist in Spanish, but is phonologically adapted to Spanish from the English ‘rate of occupancy’ (rather, in Spanish, one could say tasa de ocupación or tasa de clientela when referring to a 131 hotel’s ‘rate of occupancy’). However in 31, unlike 30, there is no hesitation or indication that the speaker is thinking, which may indicate that this SLS produces rato de ocupancia in the belief that this is the correct Spanish equivalent for ‘rate of occupancy’. In any case, the latter, which implies switching to the phonology of English, is not produced. There are many examples similar to 30 and 31 in the presentations of both HLS and SLS. In contrast, speakers freely make switches to English in the conversation, as the results show. Furthermore, HLS produce significantly fewer lexical creations in conversations (a rate of use of 5.7 as opposed to 19.7 in presentations), which also suggests that lexical creations are more likely to be produced in order to avoid the use of English when the situation creates the need or pressure to speak only in Spanish. In contrast, the results of SLS do not display a significant difference in the production of lexical creations across registers (namely, the rates of use are: 14.8 conversations, 18.1 interviews, 18 presentations). I suggest that this indicates an important difference between HLS and SLS. HLS belong to a community of speakers in which loans from English (e.g. troca ‘truck’, lonche ‘lunch’) and some degree of code-switching already exist in everyday interactions as part of their variety of Spanish in constant contact with English (Hernández-Chávez et al. 1975). On the other hand, SLS are learning a standard variety based on monolingual varieties of Spanish not in contact with English. SLS are more self-conscious of the use of English and regard it as an indication of lack of fluency in Spanish, which is a concern that several SLS in this study expressed during their interviews and 132 conversations. It seems to me that the higher overall rate of lexical creations as well as the lack of differentiation of this feature among registers may reflect SLS’s willingness to be considered ‘good’ speakers of Spanish in all situations, speakers who do not want to resort to English. Even in informal situations as the passage in (29), speakers want to be able to produce colloquial words in Spanish. Although SLS switch to English in all situations –following the same pattern as HLS (i.e. more switches in informal registers)–, the rate of switches may have been higher had they not produced lexical creations throughout the three situations analyzed. The results of the linguistic features of attributive adjectives and type/token ratio also support the hypotheses tested in this study: the frequency of attributive adjectives and the ratio of lexical diversity are higher the more formal the situation. Both HLS and SLS follow this pattern. These results are consistent with previous work; for example, Biber (1988) found that in his dimension 1, ‘involved versus informational production’, a high frequency of attributive adjectives as well as higher type/toke ratios co-occur with texts that are highly informational, such as official documents and academic prose. On the other hand, in Biber’s study, telephone and face-to-face conversations have low frequency of attributive adjectives and lower type token/ratio. Although, unlike Biber’s work, my study does not include written registers, I believe that the results can still be compared as the type of discourse found in academic class presentations is closer to written academic prose than to spoken casual conversations among friends. 133 5. Conclusions In this chapter, I have analyzed a number of lexical features with the purpose of investigating the possibility of variation in the lexical choices of HLS and SLS across linguistics registers. First, we have looked at the lexicon in terms of formal, informal, and technical. The results demonstrated that there are indeed significant differences, both for HLS and SLS in the production of technical terms and formal vs. informal lexicon. Although, it may seem obvious that in informal situations the vocabulary used by speakers will also be informal, or formal for formal situations, the fact is that there are few studies that investigate this relation in the non-dominant language of heritage speakers taking into account different registers. The present study provides evidence of significant variation across situations that vary in the degree of formality. Second, this chapter has also looked at lexical transfer in various forms: loans, calques, switches to English, and lexical creations. The results obtained also show variation across registers. However, the differences are small for loans and calques, whereas there is a great difference in the number of switches to English, which both HLS and SLS avoid in academic registers. I argue that there are several conclusions that can be drawn from these data: First, switches to English are perceived as not appropriate for academic situations by both HLS and SLS. Second, in the case of HLS, the production of switches is related to that of lexical creations: the more switches, the fewer lexical creations and vice versa. Third, the differences 134 observed between HLS and SLS in switches and creations might be related to a different way of approaching register variation for these two groups of speakers: it may be the case that HLS use English switches as elements that form part of their informal Spanish variety, whereas SLS use them as a resort to make up for lack of adequate informal lexicon. Third, we have also seen variation across registers as far as use of attributive adjectives and type/token ratio. For these two features too, the results are consistent with previous studies that show that attributive adjective and higher lexical diversity correlate with informative and more elaborate registers (e.g. Biber 1988; Finegan and Biber 1994), in the case of the present study, that situation is the academic presentations. To conclude, this chapter provides a picture of lexical use in the non- dominant language of HLS and SLS. Through the examination of a number lexical features, the data analyzed here have shown how different lexical choices are made depending on the register. These registers vary in the purpose for which they are intended, in the participants or audience involved in each of them, the topics discussed, and the degree formality. I believe that all these factors play a role in the speakers’ choice of lexical items. For example, the informational nature of presentations is connected with the higher rates of attributive adjectives and type/token ratio; the type of audience in the interview (the fact that the interviewer is a researcher, and an instructor of Spanish) is a factor that contributes to the speakers’ avoidance of slang and vulgar items; the topic of the interviews also accounts for the 135 higher rates of formal vocabulary than in conversations; and so on and so forth for all the lexical features that we have seen throughout the chapter. Based on the results obtained for the data analyzed in this chapter, I conclude by arguing that the lexical variation that we observe across registers is evidence against the idea of monostylization in the Spanish of HLS. Additionally, this chapter indicates that both heritage and second language learners have the ability to differentiate linguistic registers and attempt to produce what is appropriate for each situation of use. 136 Chapter 4. Conclusions This dissertation has examined the oral production of HLS and SLS in different situations of use in order to investigate the hypothesis of variation across registers. The framework for the study of register variation presented here has incorporated some of the main issues in the literature of register and style variation as well as the research on Spanish as a minority language in the U.S. In particular, the literature reviewed in this dissertation approached the study of register variation by looking into the following aspects: attention paid to speech (Labov 1972), audience design (Bell 1984), situational parameters (Finegan and Biber 1994), and the co-occurrence of linguistic features with specific registers (Biber 1989). The framework I have used in this dissertation considers all these aspects to play a role in intra-speaker variation. Accordingly, in order to investigate register variation, this study was planned to obtain data from the same speakers in different situations of use, which I have referred to as registers throughout my study. These registers were designed to: (a) Be related to different degrees of self-monitoring (Labov 1972). Accordingly, greater attention paid to speech was expected in academic presentations than in casual conversations. (b) Have different audiences (Bell 1984); namely, a large passive audience consisting of students and a professor in presentations, a researcher/instructor for interviews, and a close friend for conversations. 137 (c) Differ in their situational parameters including planning, topic, and purpose among others (Finegan and Biber 1994, 2001). In the presentation situation, there was the possibility of careful production but not in the interview and conversation since these are produced on-line without previous planning. In each register there were different predominant topics (the topic of the academic discipline for presentations, discussion of current issues of general interest for interviews, and personal topics for conversations, such as talking about the day’s activities, studies, job issues, family, etc.). Each register has a different purpose (the primary purpose of informing for presentations; informational and partly interpersonal purpose in interviews; and primarily affective in conversations). Additionally, each register is associated with a different degree of formality. The three situations that I examine, i.e. class oral presentations, interviews, and conversations, range on a scale from more to less formal. (d) Obtain different characteristic linguistic features (Biber 1988). Research has shown that academic presentations are characterized by more elaborate language, such as high lexical diversity and attributive adjectives (Finegan and Biber 1994); whereas informal situations such as conversations with friends are characterized by the use of economy features, for example, punctors and contractions. Furthermore, in this study, I have taken into consideration the insights from research on contact situations and the issue of transfer between Spanish and English in heritage and second language situations. Both groups of speakers are exposed to Spanish and English on an almost daily basis: For HLS, Spanish is the language of 138 the home; for SLS it is their second language and the one associated with academic settings; for both groups Spanish is the non-dominant language in relation to English. Research has observed that phenomena characteristic of situations of societal bilingualism (e.g. transfer) are shared with other situations including second language acquisition (Silva-Corvalán 1991, 1994). Therefore, I have examined a number of features that have been shown to be the locus of transfer from English in situations of bilingualism and language acquisition. The main question investigated in this study was whether Spanish as the non- dominant language of heritage and second language speakers would show variation across registers. In investigating this main question, this dissertation has examined the occurrence and frequency of linguistic features across different registers and the question of whether HLS and SLS have similar patterns of use across situations. The hypothesis was that HLS would show variation and that this variation would be reflected in the different use of some linguistic features according to the situation of use. Additionally, it was hypothesized that SLS would also vary across registers but that their use of certain features would differ from that of HLS due to the different circumstances in which Spanish was acquired. In general, the results presented in this dissertation confirm the hypothesis that variation across registers exists in the Spanish production of HLS and SLS and is evidenced in the different use of linguistic features. 139 1. Summary of the findings The following paragraphs summarize the results obtained in this dissertation and relate these results to the questions that were posed in chapter 1 regarding: (1) the distribution and frequency of the linguistic features examined across registers and (2) the comparison between groups of speakers. 1.1. Linguistic features in the speech of HLS across registers Regarding the first question, this study analyzed a number of linguistic features that previous research has shown to vary according to register or style differences. Furthermore, the linguistic features that this study analyzed are especially relevant in the case of Spanish in contact with English both as a heritage language spoken in the U.S. and as a second language taught to native speakers of English. In particular, chapter 2 studied discourse particles and contractions, while chapter 3 examined a number of lexical choices including informal, formal, and technical vocabulary; transfer of lexical items from English; attributive adjectives; and type/token ratio. The data for the study consisted of transcriptions of audio- recordings of each speaker in three situations: presentations, interviews, and conversations. The results of the use of the discourse particles analyzed across different registers in this study show that HLS vary their production according to the situation. First, HLS produce uses of como ‘like’ as quotative and punctor more frequently in informal registers. These uses are an innovation of the variety spoken by HLS, which has not been illustrated in the Spanish literature. I argue that these uses of como are 140 transferred from the quotative and punctor functions of English ‘like’. Second, HLS do not produce punctors in the presentations, which agrees with the view that the presentation is recognized as a more formal register with features different from casual spoken registers. There is also variation that can be seen in the production of contractions: HLS use full forms such as pues ‘so/well’, para ‘for/so that’, and the various forms of estar ‘to be’ more frequently in formal situations than the pronunciation of these items as pos, pa, tá. The results shown in chapter 3 also provided evidence of differentiation in the use of lexical items across registers. First, the results of the use of formal, informal, and technical vocabulary demonstrated that there are indeed significant differences in the production of technical terms and formal vs. informal lexicon. Second, the analysis of various forms of lexical transfer (loans, calques, switches to English, and lexical creations) also showed variation across registers; in particular, there is a great difference in the number of switches to English, which speakers rarely produce in academic registers. Third, we also saw variation across registers as far as use of attributive adjectives and type/token ratio, which co-occur with academic presentations; namely, the situation that is most elaborate, informational, and formal. This dissertation offers evidence supporting the main hypothesis of variation across registers that was proposed in chapter 1. The outcome confirms that HLS are able to modify their Spanish depending on the exigencies of the register even when 141 their linguistic resources may be more limited in their non-dominant language. The scarcity of non-canonical uses in academic registers is an indication of their linguistic ability to differentiate among registers with varying degrees of formality. In this dissertation, it is not the case that the Spanish production of HLS is a monostylistic variety. 1.2. Comparison between HLS and SLS This dissertation shows that, in general, HLS and SLS follow similar patterns of variation reflected in the linguistic features reported across the registers examined. Both groups produce fewer non-canonical uses in the presentation situation, such as como, punctors, as well as loans, calques, and switches to English. Conversely, these features are more frequent in the conversation situation as well as the use of informal vocabulary. However, a closer look at the results reveals that there are several differences between the HLS and SLS group in terms of frequency of production of the linguistic features analyzed. First, regarding the use of discourse particles, the production of como by SLS is much lower than that of HLS. SLS do not produce so (adapted to Spanish as [so]) in their Spanish discourse. Furthermore, SLS also use fewer punctors than HLS. Second, there are differences observed between HLS and SLS in lexical transfer since SLS produce an overall lesser amount of informal words, loans, and calques than HLS; by contrast, SLS produce more formal and technical words as well as more creations and switches. 142 I claim that the differences observed between HLS and SLS in this study are related to two main factors: (1) the nature of the Spanish language input: casual versus academic; and (2) the variety of Spanish to which speakers are exposed: in intensive contact with English in the case of HLS, but not so in the case of SLS. Regarding the formality of the input in Spanish, SLS are mainly exposed to Spanish in the formal setting of the classroom instruction which may account for less access to colloquialisms and slang items as well as punctors and informal pronunciations such as the contractions analyzed in this study. On the other hand, HLS are exposed to those characteristics of oral informal registers in Spanish, as it is the variety they use for familiar interactions in their community, with relatives, and close friends. As far as the variety of Spanish that HLS are exposed to, it is a variety that is in contact with English and considered a minority language within the U.S. As many studies on Spanish in the U.S. have shown, this situation of contact results in a number of linguistic phenomena such as transfer (Hernández-Chávez et al. 1975; Otheguy et al. 1989; Silva-Corvalán 1994; Bernal-Enríquez and Hernández-Chávez 2001), which in this study is manifested in the following ways: a) Extension of function: The incorporation of some of the uses of like from English into Spanish como, namely the use of como as a quotative and as a punctor in the speech of HLS. 143 b) Borrowing: The use of loans (e.g. troca ‘truck’, lonche ‘lunch’, so) and calques (e.g. grados ‘degrees’, carpeta ‘carpet’), frequent in the speech of HLS and characteristic of the variety of Spanish spoken in heritage communities in the U.S. On the other hand, SLS are exposed to monolingual varieties of Spanish, both in class, as they acquire a standard monolingual variety, and outside class in their trips to Spanish-speaking countries for vacation or with study abroad programs. Therefore, SLS do not produce items such as [so] or troca ‘truck’ in their Spanish discourse as such items do not belong to the variety of Spanish that SLS acquire. When SLS produce calques such as realizar for ‘realize’ instead of the Spanish darse cuenta or aplicar ‘apply’ for Spanish solicitar, they do so as result of their own transfer from English, at a rather individual level, and not as a result of acquiring that calque in the variety of Spanish that they are used to –which is the case of HLS. Although SLS and HLS have the possibility of getting together in class or around campus, there is little interaction in Spanish between them (see Appendix A for the patterns of use of Spanish of the speakers in this study). As I explained in the first chapter, HLS use Spanish primarily at home, with relatives and close friends who are also of Hispanic origin. As far as SLS, they use Spanish in class and some of them at work but none of them use Spanish with relatives or close friends. Therefore, for the most part, the differences observed between HLS and SLS can be accounted for by the two factors proposed above, that is, exposure to different amounts of informal input, as well as to a different variety of Spanish. 144 2. Pedagogical implications The results obtained in this dissertation have a number of pedagogical implications for the teaching of Spanish, specially related to heritage education. There has been a steady increase in the number of people whose home language is other than English in the U.S. The U.S. Census reports 11 percent of the population in 1980 spoke a language other than English at home, this increases to almost 14 percent in 1990, 18 percent in 2000 and, in 2005, the American Community Survey 13 reports almost fifty-two million Americans or 20 percent of the U.S. population who speak a language other than English at home. As a result, foreign language teaching professionals across the country are encountering increasing numbers of heritage learners. There is a general recognition that the needs of heritage learners are different from second language learners, and a growing body of literature identifies those needs. Some researchers have indicated that heritage learners experience a lack of familiarity with the full range of stylistic registers available to the educated monolingual speaker. Scholarly literature investigating language loss and shift among Spanish speakers in the U.S. (Silva Corvalán 1994; Valdés and Geoffrion- Vinci 1998) suggests the loss of stylistic variation among heritage speakers. In the case of HLS, students pursuing higher education in Spanish frequently start college without having acquired the full range of styles and level required in the academy (Porras 1997; Valdés 1997; Valdés and Geoffrion-Vinci 1998). In particular, this 13 Survey available from the US Census website site (http://www.census.gov/) 145 dissertation has provided evidence of use of certain colloquialisms, non-canonical uses of discourse markers, casual pronunciations, and lexical transfer in academic presentations. Even though when compared to interviews and, mainly, conversations, the frequency of use of these features is lower in presentations, this still indicates that HLS do not master the academic register for which these uses are not considered appropriate. Yet, many scholars have underscored the importance of the expansion of bilingual’s register range to include forms of the language such as those required for formal interactions, academic situations, as well as general writing skills. For instance, Valdés (1995, 2001) identifies four central goals in Spanish heritage education: 1) Spanish language maintenance, 2) acquisition of a prestige variety of Spanish, 3) expansion of bilingual range, and 4) transfer of literacy skills. The development of proficiency in academic registers is essential to gaining access to and succeeding in college and in the job market. Valdés (1995) notes that one of the problems associated with learning a second dialect 14 is the fact that learners are not always aware of the exact differences between registers. Often the oral variety that HLS use at home presents informal features while the academic variety does not; thus, one of the educational goals is to help the learner develop awareness of register variation to assist in the adequate use of features depending on the social situation. One of the challenges in heritage language pedagogy is how to respond to sociolinguistic issues related to the indexing of social values to certain varieties of 14 Valdés uses the term ‘dialect’ to what I refer to as ‘register’. 146 the language. For example, what is known in most languages as the ‘standard’ variety can be described in fact as the prestige social dialect. In the case of heritage speakers, the ‘standard’ is often associated with monolingual varieties and, conversely, their own contact variety is considered ‘non-standard’. According to some researchers (e.g. Valdés 1995; Villa 1996), the stigmatization of vernacular varieties of Spanish in U.S. communities has led to linguistic insecurity and inhibition that directly interferes with the language development process. In order to tackle this situation, Martínez (2003) argues for a pedagogical approach that specifically addresses the relationship between language, power, and social groups. Martínez follows the pedagogy of dialect awareness developed by Wolfram and his associates (Wolfram 1999; Wolfram, Adger, and Christian 1999), supporting a pedagogy that explicitly introduces students to the following linguistic facts: (a) all languages vary, (b) all dialects are equal from a linguistic point of view (rule- governed, systematic), and (c) variation occurs at different levels. Martínez, however, goes one step further for heritage education and proposes the addition of a social framework to the linguistic framework developed by Wolfram. Thus, addressing themes such as the functions that different registers serve, the distribution of different registers and the social evaluation of different varieties of language. Martínez’s approach seems to me to be a very useful pedagogy to address not only the specific needs of heritage speakers but also the sociolinguistic issues that face many heritage learners of Spanish. 147 This dissertation confirms that HLS and SLS are not identical regarding their register variation. Although both groups show some degree of variation in the use of linguistic features across registers, they do have important differences, as I summarized in the previous section. The results shown in this dissertation shed some light regarding the characteristic strengths and weaknesses of HLS and SLS and the specific needs of each group. As I pointed out, for HLS the academic register is less developed than their everyday oral register. With regard to SLS, the speakers that participated in this study had a high level of competence and fluency in Spanish in the presentation situation and, also, although to a lesser extent, in the interviews. In fact, if we were to listen to SLS’ academic production in the classroom presentation only, we could presume a very high level of Spanish fluency almost native-like (except for the pronunciation aspect, for HLS native-like, whereas SLS have a noticeable foreign accent). Still, it was clear in the conversation situation that SLS lacked the resources that made the bilingual HLS group sound native, that is, the knowledge and ability to fluently use colloquialisms, slang, conversational markers, casual pronunciations, etc. in situations in which such linguistic features are not only appropriate but also expected. This dissertation suggests that the strengths and weaknesses in the non- dominant language of HLS and SLS are at different ends. These observations can be framed in terms of the conceptual distinction made by Cummins (1984) between basic interpersonal communication skills (BICS) and cognitive academic language 148 proficiency (CALP). Communicative approaches emphasize BICS, whereas formal, more grammar-oriented approaches emphasize CALP. As the present dissertation shows, the advantage that HLS learners bring to the classroom, if placed beside SLS, is related to BICS, not CALP. Spanish as a foreign language is often taught following communicative approaches, which would not address the specific needs of HLS. On the other hand, focus on conscious learning of grammatical rules when not done at the appropriate level often results in HLS falling behind as they may not have learned or acquired these items before coming to college. SLS sometimes outperform HLS on grammar tests and get higher grades in the language class, even though the non-heritage speaker may not be fluent communicating in everyday interactions (Krashen 2000). The field of second language acquisition has a much longer tradition than the field of heritage education. There is a tremendous amount of scholarly work proposing different theories of second language acquisition and development as well as a great number of pedagogical proposals regarding teaching practices for second or foreign language. I mentioned earlier communicative approaches, which have been since the 1990s a popular methodology in language departments for the teaching of foreign languages. These approaches address the need of developing a communicative fluency in the oral informal registers of the learners; essentially, addressing the development of BICS in SLS. The results of this dissertation are consistent with the research by Cummins (1984) and Krashen (2000) in that HLS and SLS show different skills when using Spanish in the same situations. For example, 149 HLS show more developed BICS than SLS producing more colloquial and slang items, casual pronunciations, and conversational markers in informal situations than SLS. Until recently, heritage language education has been following the pedagogical models of second language acquisition. The differences observed between HLS and SLS in this study support a growing body of literature that recognizes that Spanish as a heritage language and as a second language should meet different educational needs in order to reach the goal of expanding the learner’s register range in their non- dominant language. Based on recent research on heritage education (Martínez 2003), I consider that a pedagogical approach that specifically addresses register variation in class, paying attention to a wide range of communicative situations, will be an effective methodology to expand the range of HLS abilities in the non-dominant language. 3. Direction of future research and concluding remarks The analysis of register variation presented here raises some issues that future research needs to address both in terms of the linguistic features of HLS and SLS in their register range, and in the pedagogical implications of these findings. Future studies on the register and style differentiation of heritage languages need to examine a greater variety of linguistic features as well as a wider range of registers than has been done so far. As Biber (1995) has pointed out, distinctive indicators of register are rare and the different distribution of the same linguistic feature can distinguish among registers. Furthermore, a comparison of the HLS’ 150 academic production with the production of non-HLS native speakers in a context in which Spanish is the societal language can shed light on the process of acquisition of academic registers. By comparing HLS to monolingual Spanish speakers, we can examine the ways in which a situation of societal contact (as with HLS) can affect the development of the register range of the speakers in contrast to non-contact situations. In addition, a study of the register variation in the English of HLS and SLS can also allow us to examine the relation between the development of registers in the dominant and non-dominant languages of the students. To conclude, the results presented in this study show that speakers have variation in their non-dominant language and this is reflected in the different use of a number of linguistic features depending on the situation. This dissertation provides evidence of variation in a relatively small range of registers in the speakers’ non- dominant language. In particular, I have looked at three registers that are all oral and that correspond to social situations that are not radically different. If, on the other hand, I had compared a written register, for instance, a formal legal text with the text of a casual online chat between friends, I would have certainly increased the possibilities of getting variation in many linguistic features. However, the fact that this study finds regular patterns of variation across a small range of oral registers makes it even more certain that the fundamental sociolinguistic principle that “there are no single-style speakers” (Labov 1970) prevails even when the use of the language is largely restricted to a particular domain (Dorian 1994). In this study, heritage language speakers use Spanish primarily at home and family interactions. 151 They may have not developed a full range of registers in their non-dominant language (Valdés and Geoffrion-Vinci 1998). Yet, there is no loss of stylistic range that would justify the idea of ‘monostylism’ for Spanish as a non-dominant language in HLS and SLS. Studies such as the one presented here are important as they provide valuable insights about the development of different varieties of a language as well as the characteristics of the register range in a non-dominant language (Valdés and Geoffrion-Vinci 1998). This dissertation contributes to further our understanding of bilingualism by examining Spanish as a heritage and as a second language across different registers, which has not been previously investigated. Because of the great number of Spanish-speaking population in the U.S., the investigation of the characteristics of the Spanish spoken by growing generations of HLS is especially important. The Hispanic/Latino population is the largest minority group in the country, with Spanish as the most spoken language in the U.S. after English (data from the 2000 U.S. Census). However, even though there has been considerable research on discourse and register variation in general and, especially, in English (Chafe 1982, 1985; Finegan and Biber 1994 among others), there is still very little research on register variation in heritage language speakers in the U.S. Some studies have looked at specific registers: For example, academic oral production (Achugar 2003; Valdés and Geoffrion-Vinci 1998) or written production (Montaño-Harmon 1991; Schwartz 2003). Nonetheless, a complete study controlling different variables (situation, topic, interlocutors) and aiming at a wide range of 152 registers remains to be done. By examining register and style variation, this dissertation is a step further in bridging that gap in the literature of Spanish as a heritage language. 153 Bibliography Aaron, Jessi Elana. 2004. “‘So respetamos un tradición del uno al otro’. So and entonces in New Mexican bilingual discourse.” Spanish in Context 1.161- 179. Achugar, Mariana. 2003. “Academic registers in Spanish in the U.S.: A study of oral texts produced by bilingual speakers in a university graduate program.” In Mi Lengua: Spanish as a Heritage Language in the US, edited by Ana Roca and M. Cecilia Colombi. Washington DC: Georgetown University Press, 213- 234. Adalar, Nevin and Sali Tagliamonte. 1998. “Borrowed nouns; Bilingual people: The case of the “Londrall” in Northern Cyprus.” International Journal of Bilingualism, 2:2, 139-159. Alvarez, Alfredo I. 1999. “Las construcciones consecutivas.” In Gramática Descriptiva de la Lengua Española, edited by Ignacio Bosque and Violeta Demonte. Madrid: Espasa-Calpe, 3739-3804. American Heritage Spanish Dictionary. 2001 (2 nd edition). By the editors of the American Heritage Dictionaries. Houghton Mifflin. Andersen, Gisle. 2000. “The role of the pragmatic marker ‘like’ in utterance interpretation.” In Pragmatic Particles and Propositional Attitude. Pragmatics and Beyond, 79, edited by Gisle Andersen and Thorsten Fretheim. Amsterdam: John Benjamins, 17-38. Andersen, Gisle. 2001. Pragmatic Particles and Sociolinguistic Variation: a Relevance-Theoretic Approach to the Language of Adolescents. Amsterdam: John Benjamins. Balota, David A., Michael J. Cortese, Susan D. Sergent-Marshall, Daniel H. Spieler, and Melvin J. Yap. 2004. “Visual word recognition of single-syllable words.” Journal of Experimental Psychology: General, 133, 2, 283-316. Batchelor, Ronald E. 1994. Using Spanish Synonyms. Cambridge: Cambridge University Press. Bell, Allan. 1984. “Language style as audience design.” Language in Society, 13:2, 145-204. 154 Bell, Allan. 1990. “Audience and referee design in New Zealand media language.” In New Zealand Ways of Speaking English, edited by Allan Bell and Janet Holmes. Clevedon, UK: Multilingual Matters. Bell, Allan. 2001. “Back in style: Reworking audience design.” In Style and Sociolinguistic Variation, edited by Penelope Eckert and John R. Rickford. Cambridge: Cambridge University Press, 139-169. Bello, Andrés. Gramática de la Lengua Castellana Destinada al Uso de los Americanos (1847). Series Colección Bibliotheca Philologica, Madrid: Arco/Libros, 1988. Bernal-Enríquez, Ysaura and Eduardo Hernández Chávez. 2001. “La eseñanza del español en Nuevo México: ¿Revitalización o erradicación de la variedad chicana?” In Mi Lengua: Spanish as a Heritage Language in the US, edited by Ana Roca and M. Cecilia Colombi. Washington DC: Georgetown University Press, 96-115. Biber, Douglas. 1986. “Spoken and written textual dimensions in English: resolving the contradictory findings.” Language 62: 384-414. Biber, Douglas. 1988. Variation across Speech and Writing. Cambridge: Cambridge University Press. Biber, Douglas. 1989. “A typology of English texts.” Linguistics 27: 3-43. Biber, Douglas. 1995. Dimensions of Registers Variation: A Cross-linguistic Comparison. Cambridge: Cambridge University Press. Biber, Douglas and Edward Finegan (eds.).1994. Sociolinguistic Perspectives on Register. Oxford: Oxford University Press. Blyth, Carl Jr., Sigrid Recktenwals, and Jenny Wang. 1990. “I'm like, ‘say what?!’ A new quotative in American oral narrative.” American Speech 65, 3: 215-227. Bosque, Ignacio. 1993. Las Categorías Gramaticales, Relaciones y Diferencias. Madrid: Síntesis. Bosque, Ignacio and Violeta Demonte (eds.). 1999. Gramática Descriptiva de la Lengua Española. Madrid: Espasa-Calpe. Brown, Gillian and George Yule. 1983. Discourse Analysis. Cambridge; New York: Cambridge University Press. 155 Budzhak-Jones, Svitlana. 1998. “Against word-internal codeswitching: Evidence from Ukrainian-English bilingualism.” International Journal of Bilingualism, 2:2, 161-182. Butt, John and Carmen Benjamin. 1994. A New Reference Grammar of Modern Spanish, 2 nd edition. Lincolnwood, IL: NTC. Cambridge Dictionary of American English. 2001. Prepared by Sydney I. Landau. Cambridge: Cambridge University Press. Campbell, Russell N. and Donna Christian. 2003. “Directions in research: Intergenerational transmission of heritage languages.” Heritage Language Journal 1(1), http://www.heritagelanguages.org/. Celce-Murcia, Marianne and Diane Larsen-Freeman.1983. The Grammar Book: An ESL/EFL Teacher’s Course. Rowley, Mass.: Newbury House. Cedergren, Henrietta. 1973. The interplay of social and linguistic factors in Panama. Ph.D. Dissertation, Cornell University. Chafe, Wallace. 1982. “Integration and involvement in speaking, writing, and oral literature.” In Spoken and Written Language: Exploring Orality and Literacy, edited by Deborah Tannen. Norwood, N.J.: Ablex, 35-54. Chafe, Wallace. 1985. “Linguistic differences produced by differences between speaking and writing.” In Literacy, Language and Learning, edited by David Olson, Nancy Torrance, and Angela Hildyard. Cambridge: Cambridge University Press, 105-123. Chafe, Wallace and Jane Danielewicz 1987. “Properties of spoken and written language.” In Comprehending Spoken and Written Language, edited by Rosalind Horowitz and S. J. Samuels. NY: Academic Press, 83-114. Cohen, Gerald Leonard. 1985. Studies in Slang. Frankfurt; New York: P. Lang. Coleman, Julie. 2004. A History of Cant and Slang Dictionaries. Oxford: Oxford University Press. Cooper, Thomas C. 1998. “Teaching idioms.” Foreign Language Annals 31, 255-66. Cortés Rodríquez, Luis. 1991. Sobre Conectores, Expletivos y Muletillas en el Español Hablado. Málaga: Ágora. 156 Cortés Rodríquez, Luis. 1996. Español hablado. Bibliografía sobre aspectos teóricos y empíricos (morfosintácticos y sintáctico-pragmáticos). Cáceres: Universidad de Extremadura. Cortés Rodríquez, Luis. 1998. “Marcadores del discurso y análisis cuantitativo.” In Los Marcadores del Discurso: Teoría y Análisis, edited by María Antonia Martín Zorraquino and Estrella Montolío Durán. Madrid: Arco/Libros. 142- 160. Cummins, Jim. 1984. Bilingualism and Special Education: Issues in Assessment and Pedagogy. Clevedon, UK: Multilingual Matters. Demonte, Violeta. 1999. “El adjetivo: clases y usos. La posición del adjetivo en el sintagma nominal.” In Gramática Descriptiva de la Lengua Española, edited by Ignacio Bosque and Violeta Demonte. Madrid: Espasa-Calpe, 129-215. Dorian, Nancy C. 1981. Language Death: The Life Cycle of a Scottish Gaelic Dialect. Philadelphia: University of Pennsylvania Press. Dorian, Nancy C. 1994. “Stylistic variation in a language restricted to private-sphere use.” In Sociolinguistic Perspectives on Register, edited by Douglas Biber and Edward Finegan. Oxford: Oxford University Press, 217-232. Dressler, Wolfgang U. 1982. “Acceleration, retardation, and reversal in language decay?” In Language Spread, edited by Robert L. Cooper. Bloomington: Indiana University Press, 321-336. Dressler, Wolfgang U. 1988. “Language death.” In Linguistics: The Cambridge Survey, vol. 4 (Language: The Socio-cultural Context). Cambridge University Press, 184-191. Dressler, Wolfgang U. 1991. “The sociolinguistic and patholinguistic attrition of Breton phonology, morphology, and morphophonology.” In First Language Attrition, edited by Herbert W. Seliger and Robert M. Vago. Cambridge: Cambridge University Press, 99–112. Eckert, Penelope. 2000. Linguistic Variation as Social Practice: The Linguistic Construction of Identity in Belten High. Malden, Massachusetts and Oxford, U.K.: Blackwell. Eckert, Penelope and John R. Rickford. (eds.). 2001. Style and Sociolinguistic Variation. Cambridge: Cambridge University Press. Ferguson, Charles A. 1959. “Diglossia.” Word 15: 325-340. 157 Ferguson, Charles A.1964. “Baby talk in six languages.” In The Ethnography of Communication, edited by John Gumperz and Dell Hymes. American Anthropologist 66, 6: 102-113. Ferguson, Charles A. 1994. “Dialect, register and genre: Working assumptions about conventionalization.” In Sociolinguistic Perspectives on Register, edited by Douglas Biber and Edward Finegan. Oxford University Press, 15-30. Ferrara, Kathleen and Barbara Bell. 1995. “Sociolinguistic variation and discourse function of constructed dialogue introducers: The case of be + like.” American Speech 70, 3: 265-290. Finegan, Edward. 2004. “American English and its distinctiveness.” In Language in the USA: Themes for the Twenty-first Century, edited by Edward Finegan and John R. Rickford. Cambridge: Cambridge University Press, 18-38. Finegan, Edward and Douglas Biber. 1994. “Register and social dialect variation: an integrated approach.” In Sociolinguistic Perspectives on Register, edited by Douglas Biber and Edward Finegan. Oxford: Oxford University Press, 315- 347. Finegan, Edward and Douglas Biber. 2001. “Register variation and social dialect variation: the register axiom.” In Style and Sociolinguistic Variation, edited by Penelope Eckert and John R. Rickford. Cambridge University Press, 235- 268. Galán Rodríguez, Carmen. 1999. “La subordinación causal y final.” In Gramática Descriptiva de la Lengua Española, edited by Ignacio Bosque and Violeta Demonte. Madrid: Espasa Calpe, 3597-3642. Galván, Roberto A. 1995. The dictionary of Chicano Spanish/El Diccionario del Español Chicano (2nd. ed.). Lincolnwood, Ill., USA: National Textbook Co. Gili Gaya, Samuel. 1943. Curso Superior de Sintaxis Española. Barcelona: Spes. Goffman, Erving. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press. Gumperz, John. 1964. “Speech variation and the study of Indian civilization.” In Language in Culture and Society, edited by Dell Hymes. New York: Harper and Row, 416-428. Halliday, M.A.K. and Ruqaiya Hasan. 1976. Cohesion in English. London: Longman. 158 Hamel, Bernard H. 1996. Diccionario Bilingüe Hamel de Americanismos. Los Angeles, CA: Bilingual Book Press. Hernández-Chávez, Eduardo, Andrew Cohen, and Anthony Beltramo. 1975. El Lenguaje de los Chicanos: Regional and Social Characteristics Used by Mexican Americans. Arlington, VA: Center for Applied Linguistics. Hidalgo, Margarita. 1987. “On the question of ‘standard’ vs. ‘dialect’: Implications for teaching Hispanic college students.” Hispanic Journal of the Behavioral Sciences, 9, 375-95. Hockett, Charles F. 1958. A Course in Modern Linguistics. The Macmillan Company: New York. Hudson, Patrick T.W., and Marijke W. Bergman, 1985. “Lexical knowledge in word recognition: Word length and word frequency in naming and lexical decision tasks.” Journal of Memory and Language, 24, 46-58. Hudson, Alan. 1994. “Diglossia as a special case of register variation.” In Sociolinguistic Perspectives on Register, edited by Douglas Biber and Edward Finegan. Oxford: Oxford University Press, 219-314. Hymes, Dell. (ed.) 1964. Language in Culture and Society. New York: Harper and Row. Hymes, Dell. 1984. “Sociolinguistics: Stability and consolidation.” International Journal of the Sociology of Language 45: 39-45. Irujo, Suzanne, 1986. “A piece of cake: Learning and teaching idioms.” ELT Journal 40(3): 236-242. Oxford: Oxford University Press. Judge, Anne and F.G. Healey. 1983. A Reference Grammar of Modern French. London: Edward Arnold. Juhasz, Barbara J., and Keith Rayner. 2003. “Investigating the effects of a set of intercorrelated variables on eye fixation durations in reading.” Journal of Experimental Psychology: Learning, Memory, and Cognition, 29(6), 1312-8. Krashen, Stephen. 2000. “Bilingual education, the acquisition of English, and the retention and loss of Spanish.” Research on Spanish in the United States: Linguistic Issues and Challenges, edited by Ana Roca. Somerville, MA: Cascadilla, 432-44. 159 Labov, William. 1966. The Social Stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics. Labov, William. 1970. The Study of Nonstandard English. Champaign, IL: National Council of Teachers of English. Labov, William. 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Labov, William. 1984. “Field methods of the project on linguistic change and variation.” In Language in Use: Readings in Sociolinguistics, edited by John Baugh and Joel Sherzer. Englewood Cliffs, NJ: Prentice Hall, 43-70. Lance, Donald M. 1975. “Dialectal and nonstandard forms in Texas Spanish.” In El Lenguaje de los Chicanos: Regional and Social Characteristics Used by Mexican Americans, edited by Eduardo Hernández-Chávez, Andrew Cohen, and Anthony F. Beltramo. Arlington, VA: Center for Applied Linguistics, 37- 51. Lara Ramos, Luis Fernando. 1996. Diccionario del Español Usual en México. México: Colegio de México, Centro de Estudios Linguísticos y Literarios. Levelt, Willem J.M., Ardi Roelofs, and Antje S. Meyer. 1999. “A theory of lexical access in speech production.” Behavioral and Brain Sciences, 22, 1–38. Lipski, John M. 1994. “So in bilingual Spanish: from code-switching to borrowing.” Unpublished paper presented at the XV Symposium on Spanish and Portuguese Bilingualism, Rutgers University, New Brunswick, NJ. Lipski, John M. 1996. El Español de América. Madrid: Cátedra. Lipski, John M. 2004. “Code-switching or borrowing? No sé so no puedo decir, you know.” Unpublished paper presented at the WSS2, SUNY Albany. Lynch, Andrew. 2003. “The relationship between second and heritage language acquisition: Notes on research and theory building.” Heritage Language Journal 1 (1), http://www.heritagelanguages.org/. Martín Zorraquino, M. Antonia and Estrella Montolío (eds.). 1998. Marcadores del Discurso. Teoría y Análisis. Madrid: Arco-Libros. 160 Martín Zorraquino, M. Antonia, Estrella Montolío, and José Portolés Lázaro. 1999. “Los marcadores del discurso.” In Gramática Descriptiva de la Lengua Española, edited by Ignacio Bosque and Violeta Demonte. Madrid: Espasa- Calpe, 4051-4213. Martínez, Glenn A. 2003. “Classroom based dialect awareness in heritage language instruction: A critical applied linguistic approach.” Heritage Language Journal 1 (1), http://www.heritagelanguages.org/. Mathis, Terri and George Yule. 1994. “Zero quotatives.” Discourse Processes 18: 63-76. McEnery, Tony and Andrew Wilson.1996. Corpus Linguistics. Edinburgh Textbooks in Empirical Linguistics. Edinburgh: Edinburgh University Press. Meechan, Marjorie, Sali Tagliamonte, and Shana Poplack. 1996. “Isolating systems in language contact: A variationist method.” Paper presented at Methods IX, Bangor, U.K. Mendieta, Eva. 1999. El Préstamo en el Español de los Estados Unidos. New York: P. Lang. Milroy, Lesley. 1987. Observing and Analyzing Natural Language. Oxford: Blackwell. Montaño-Harmon, M. Rosario. 1991. “Discourse features of written Mexican Spanish: Current research in contrastive rhetoric and its implications.” Hispania 74: 417- 425. Montolío, Estrella. 1999. “Las construcciones condicionales.” In Gramática Descriptiva de la Lengua Española, edited by Ignacio Bosque and Violeta Demonte. Madrid: Espasa-Calpe, 3643-3737. Moreno Ayora, Antonio. 1991. Sintaxis y Semántica de “Como.” Málaga: Ágora. Odlin, Terence. 1989. Language Transfer. Cambridge: Cambridge University Press. O’Regan, J. Kevin, and Arthur M. Jacobs. 1992. “Optimal viewing position effect in word recognition: A challenge to current theory.” Journal of Experimental Psychology: Human Perception and Performance, 18, 185-197. Otheguy, Ricardo, Ofelia García, and Mariela Fernández.. 1989. “Transferring, switching, and modeling in West New York Spanish: An intergenerational study.” International Journal of the Sociology of Language, 41-53. 161 Otheguy, Ricardo. 1993. “A reconsideration of the notion of loan translation in the analysis of U.S. Spanish.” In Spanish in the United States: Linguistic Contact and Diversity, edited by Ana Roca and John M. Lipski. Berlin: Mouton de Gruyter, 21-41. Oxford English Dictionary. 1989. (2 nd ed.). Prepared by John A. Simpson and E.S.C. Weiner. Oxford: Clarendon Press. Pavón Lucero, M. Victoria. 1999. “Clases de partículas: preposición, conjunción y adverbio.” In Gramática Descriptiva de la Lengua Española, edited by Ignacio Bosque and Violeta Demonte. Madrid: Espasa-Calpe, 565-655. Phillips, Robert, Jr. 1967. Los Angeles Spanish: A Descriptive Analysis. Ph.D. Dissertation, University of Wisconsin. Poplack, Shana. 1987. “Contrasting patterns of code-switching in two communities.” In Aspects of Bilingualism: Proceedings from the Fourth Nordic Symposium on Bilingualism, 1984, edited by Erling Wande. Uppsala: Ubsaliensis S. Academiae, 51-76. Poplack, Shana, David Sankoff, and Chris Miller. 1988. “The social correlates and linguistc processes of lexical borrowing and assimilation.” Linguistics. 26, pp. 47-104. Poplack, Shana and Marjory Meechan. 1998. “How languages fit together in codemixing.” International Journal of Bilingualism, 2:2, 127-138. Porras, Jorge E. 1997. “Uso local y uso estándard: un enfoque bidialectal a la enseñanza del español para nativos.” In La Enseñanza del Español a Hispanohablantes: Praxis y Teoría, edited by M. Cecilia Colombi and Francisco X. Alarcón. Boston: Houghton-Mifflin, 190-197. Real Academia Española. 1973. Esbozo de una Nueva Gramática de la Lengua Española. Madrid: Espasa-Calpe. Richards, Jack, John Platt, and Heidi Platt. 1997. Dictionary of Language Teaching and Applied Linguistics. London: Longman. Rickford, John R. and Faye McNair-Knox. 1994. “Addressee- and topic-influenced style shift. A quantitative sociolinguistic study.” In Sociolinguistic Perspectives on Register, edited by Douglas Biber and Edward Finegan. Oxford: Oxford University Press, 235-276. 162 Roca, Ana. 1997. “Retrospectives, advances, and current needs in the teaching of Spanish to United States Hispanic bilingual students.” ADFL Bulletin 29: 37- 43. Romaine, Suzanne and Deborah Lange. 1991. “The use of like as a marker of reported speech and thought: A case of grammaticalization in process.” American Speech 66, 3: 227-79. Sánchez, Rosaura.1983. Chicano Discourse: Socio-historic Perspectives. Rowley: Newbury. Sánchez, Rosaura. 1993. “Language variation in the Spanish of the Southwest.” In Language and Culture in Learning: Teaching Spanish to Native Speakers of Spanish, edited by Barbara J. Merino, Henry T. Trueba, and Fabián A. Samaniego. London: The Flamer Press, 75-81. Sánchez-Muñoz, Ana. (in press). “Style variation in Spanish as a heritage language: a study of discourse particles in academic and non-academic registers.” In Spanish in Contact: Educational, Social, and Linguistic Inquiries. Amsterdam: John Benjamins. Sankoff, Gillian, Pierrette Thibault, Naomi Nagy, Hélène Blondeau, Marie-Odile Fonollosa, and Lucie Gagnon. 1997. “Variation in the use of discourse markers in a language contact situation.” Language Variation and Change 9: 191–218. Santiago, Julio, Donald G. MacKay, Alfonso Palma, and Christine Rho. 2000. “Sequential activation processes in producing words and syllables: Evidence from picture naming.” Language and Cognitive Processes, 15, 1–44. Santos Río, Luis. 2003. Diccionario de Partículas. Salamanca: Luso-Española de Ediciones. Schourup, Lawrence C. 1985. Common Discourse Particles in English Conversations: "Like", "Well", "y'know". New York and London: Garland. Schwartz, Ana M. 2003. “¡No me suena! Heritage Spanish speakers writing strategies.” In Mi Lengua: Spanish as a Heritage Language in the US, edited by Ana Roca and M. Cecilia Colombi. Washington DC: Georgetown University Press, 235-256. Serrano, Juan and Susan Serrano. 1999. Spanish Proverbs, Idioms and Slang of Yesterday and Today. New York: Hippocrene Books, Inc. 163 Silva-Corvalán, Carmen. 1991. “Spanish language attrition in a contact situation with English.” In First Language Attrition, edited by Herbert W. Seliger and Robert M. Vago. New York: Oxford University Press, 151-171. Silva-Corvalán, Carmen.1994. Language Contact and Change. Oxford: Clarendon Press. Silva-Corvalán, Carmen. 1998. “On borrowing as a mechanism of syntactic change.” In Romance Linguistics: Theoretical Perspectives, edited by Armin Schwegler, Bernard Tranel, and Myriam Uribe-Etxebarria. Amsterdam: John Benjamins, 225-246. Smead, Robert N. 1998. “English loanwords in Chicano Spanish: Characterization and rationale.” The Bilingual Review/La Revista Bilingüe, 23(2), 113-123. Smith, Colin (ed.). 1993. Collins Spanish-English English-Spanish Dictionary, 3rd revised edition. Glasgow and New York: Harper Collins. Solomon, Jeff and Nancy C. Rhodes. 1995. Conceptualizing Academic Language. University of California, Santa Cruz: National Center for Research on Cultural Diversity and Second Language Learning. Spillner, Bernd. 1974. “The relevance of stylistic methods for sociolinguistics.” In Association Internationale de Linguistique Apliquée. Proceedings of the Third Congress, Copenhagen 1972. vol. II: Applied Sociolinguistics, edited by A. Verdoodt. Heidelberg, 172-183. Torres, Lourdes. 2002. “Bilingual discourse particles in Puerto Rican Spanish.” Language in Society, 31.65-83. Valdés, Guadalupe. 1988. “The language situation of Mexican Americans.” In Language Diversity: Problem or Resource, edited by Sandra L. McKay and Sau-Ling C. Wong. New York: Newbury House Publishers, 111-139. Valdés, Guadalupe. 1992. “The role of the foreign language teaching profession in maintaining non-English languages in the United States.” In Languages for a Multicultural World in Transition: 1993 Northeast Conference Reports, edited by Heidi Byrnes. Skokie, IL: National Textbook, 29-71. Valdés, Guadalupe. 1995. “The teaching of minority languages as academic subjects: Pedagogical and theoretical challenges.” The Modern Language Journal 76: 301-328. 164 Valdés, Guadalupe. 1997. “The teaching of Spanish to bilingual Spanish-speaking students: Outstanding issues and unanswered questions.” In La Enseñanza del Español a Hispanohablantes: Praxis y Teoría, edited by M. Cecilia Colombi and Francisco X. Alarcón. Boston: Houghton-Mifflin, 8-44. Valdés, Guadalupe. 2001. “Heritage language students: Profiles and possibilities.” In Heritage Languages in America, edited by Joy K. Peyton, Donald Ranard, and Scott McGinnis. Arlington, VA: Center for Applied Linguistics, 37-77. Valdés, Guadalupe and Michelle Geoffrion-Vinci. 1998. “Chicano Spanish: The problem of the ‘underdeveloped’ code in bilingual repertoires.” The Modern Language Journal 82.4: 473-501. Villa, Daniel. 1996. “Choosing a ‘standard’ variety of Spanish for the instruction of native Spanish speakers in the U.S.” Foreign Language Annals, 29: 191-200. Vincent, Diane. 1993. Les Ponctuants de la Langue et autres Mots du Discours. Quebec: Nuit Blanche. Vincent, Diane and David Sankoff. 1992. “Punctors: A pragmatic variable.” Language Variation and Change 4, 205-216. Walters, Keith. 1994. “Writing and education.” In Writing and its Use: An Interdisciplinary Hand-book of International Research, edited by Hartmut Günther and Otto Ludwig. Berlin: Warlter de Gruyter, 638-645. Weinreich, Uriel. 1953/1968. Languages in Contact. The Hague: Mouton. Wilson, Kenneth G. 1993. The Columbia Guide to Standard American English. New York: Columbia University Press. Wilson, Ruth M. 1984. The Bilingual Students: Understanding Language Imagery. In Yale-New Haven Teachers Institute, Curricular Units, vol. 3: 10. Retrieved August 2, 2006, from http://www.cis.yale.edu/ynhti/curriculum/units/. Wolfram, Walt. 1999. “Repercussions from the Oakland Ebonics controversy- The critical role of dialect awareness programs.” In Making the Connection: Language and Academic Achievement Among African American Students, edited by Carolyn Adger, Donna Christian, and Orlando Taylor. Arlington, VA: Center for Applied Linguistics, 61-80. Wolfram, Walt, Carolyn Adger, and Donna Christian. 1999. Dialects in Schools and Communities. Mahwah, NJ: Lawrence Erlbaum Associates. 165 Yule, George and Mathis Terrie. 1992. “The role of staging and constructed dialogue in establishing speaker’s topic.” Linguistics 30: 199-215. Zipf, George K. 1949. Human Behavior and the Principle of Least Effort. Cambridge, Mass: Addison-Wesley. 166 Appendix A ____________________________________________________________________ SPEAKER INFORMATION AND CODING ____________________________________________________________________ Coding HLS Origin Sex Age Education (Spanish) Use of Spanish G Mx m 20 HS 1; U 3 H a W f F f M f J Mx m 20 HS 6; U 3 H a W f F f M f L Mx m 19 HS 6; U 2 H a W f F f M f W Mx f 20 HS 2; U 4 H a W r F r M r C Mx f 20 HS 6; JC 1; U 1 H a W r F f M r D Mx f 19 HS 4; U 2 H a W r F f M f B Mx f 19 HS 6; U 1 H a W f F f M f E Mx f 20 HS 4; U 2 H a W f F f M r SLS Origin Sex Age Education (Spanish) Use of Spanish S US f 20 HS 6; U 3 H n W r Fr M r A US f 20 HS 4; U 3 H n W r Fr M r Co US f 20 HS 4; U 2 H n W f Fr M n Ge US f 20 HS 6; U 2 H n W f Fr M f K US f 20 HS 6; U 2 H n W r Fr M r Cr US m 22 HS 6; U 4 H n W f Fr M r R US m 19 HS 6; U 2 H n W r Ff M r M US m 21 HS 4; U 3 H n W r Fr M r Explanation of Speaker Codings Social features Heritage origin: Mx= Mexico US= United States Refers to the place of origin (birth place) of one or both of speakers’ parents. Sex: m= male f= female. Age: 19-22 years-old Spanish education: HS= high school JC= junior college U= university (i.e. USC) The number that appears after the code refers to the number of semesters of Spanish taken at each of those educational settings (i.e. HS 4= 4 semesters at high school). 167 Use of Spanish: H= home W= work and school F= close friends M= media (i.e. TV, music, books, newspapers) The letter that appears after these codes refers to the overall daily frequency of Spanish use as follows: a= always f= frequently r= rarely. Data Collection Situations C= Conversation I = Interview P= Presentation 168 Appendix B __ __ HLS BIOGRAPHICAL AND LANGUAGE BACKGROUND QUESTIONNAIRE ____________________________________________________________________ This questionnaire will help the researcher understand more about your language background and evaluation of your own language. Thank you for your participation! 1. Name: ____________________________ USC e-mail:__________________ 2. Gender: F F M F 3. Your native language(s) is/are: 1. _____________________ 2. _____________________ 4. Where were your parents or tutors born (specify city and country): Mother/Tutor: Father/Tutor: 5. If your parents/tutors were born in another country, at what age did they come to the U.S.? Mother/Tutor: _______________ years-old. Father/Tutor: _______________ years-old. 6. How many relatives –that you see more or less frequently- come from a Spanish speaking country (specify relation and location)? Relative: __________________ Place of origin: _________________ Relative: __________________ Place of origin: _________________ Relative: __________________ Place of origin: _________________ Relative: __________________ Place of origin: _________________ 7. Where have you been in contact with/studied Spanish? (Check all that apply): Home/Relatives F Kindergarten/Elementary F Middle School F High School F Private Language School/Lessons F University/College F Spanish-speaking country F Other(explain):_______________________ F 8. Number of semesters studying Spanish in (including present semester): 169 Elementary/Middle School…........ _________ semesters High School…............................... _________ semesters Private Language School/Lessons _________ semesters University/College…..................... _________ semesters Spanish-speaking country….......... _________ semesters Other………….. _________ semesters (specify): __________________________________ 9. Use of languages: Spanish English Both Always Frequently Rarely At home: _ _____ _ _____ With friends: _ _____ _ _____ At work/school: _ _____ _ _____ Radio/Music: _ _____ _ _____ TV: _ _____ _ _____ Books/Newspapers: _ _____ _ _____ 10. Your major at USC: __________________ 11. Spanish courses that you have taken and/or that you are currently taking at USC: ____________________________________________________________________ ____________________________________________________________________ ____________________________________________________________________ ____________________________________________________________________ 12. GPA in Spanish so far: ____________ overall GPA so far: _________________ 170 Appendix C ____________________________________________________________________ SLS BIOGRAPHICAL AND LANGUAGE BACKGROUND QUESTIONNAIRE ____________________________________________________________________ This questionnaire was distributed to native speakers of English who were taking Spanish courses at USC. 1. Name: _____________________________USC E-mail:____________________ 2. Place of birth: _________________________ 3. Is English your only native language? F Yes F No If not, explain: _____________________________________________________ 4. Place of birth of parents or tutors: _________________ 5. Any languages other than English spoken at home? F Yes F No If yes, explain: _____________________________________________________ 6. Age that you started learning Spanish: _____________Where? (Check all that apply): Home with relative(s): F with caretaker: F School: F Other: F Please, specify: ____________________________ 7. Spanish at school: Number of semesters or courses taken: Kindergarten/Elementary school F ________________________ Middle School F ________________________ High School F ________________________ College F ________________________ Other F (please, specify-e.g. travel study, tutoring...) ____________________________________________________________________ 8. Your major at USC: _________________ 9. GPA in Spanish so far: ___________ 10. Spanish courses that you have taken and/or that you are currently taking at USC: ____________________________________________________________________ ____________________________________________________________________ 171 Appendix D ____________________________________________________________________ SPANISH COURSES TAKEN BY HLS AND SLS IN COLLEGE ____________________________________________________________________ The following list includes the courses that the participants in this study were taking at the time of the data collection. The class presentations recorded for the present study were the presentations that the speakers prepared for these courses. HLS SLS Course Description of the course __ C SPAN 265 Writing-intensive course L Spanish for Communication designed to develop students' D (Society and the Media) communicative skills through B grammar review, readings on current issues and exposure to media. D A SPAN 266 Writing-intensive course B S Spanish for Communication designed to develop students' L Co (Arts and Sciences) communicative skills through J intensive readings of literary C and scientific material. E K SPAN 308 A survey of Spanish and Latin W R Survey of Poetry American poetry from the B S Middle Ages to the present, acquainting students with various critical and theoretical approaches to verse. J S SPAN 310 A systematic study of the W K Structure of Spanish structure of Spanish. C Cr Topics include fundamental R aspects of the sound system; M word classes; sentences and their meaning; linguistic change and variation; standard and colloquial usage. 172 HLS SLS Course Description of the course __ G A SPAN 316 The language and culture of a J Ge Spanish for the Professions particular area of Spanish for the W Co (Business) Professions study or profession, M such as medicine and healthcare, political and social sciences, business and the law.

Abstract (if available)

Abstract One of the fundamental principles of sociolinguistics is that language is not homogeneous and that no single person speaks in the same way all the time. Numerous studies have provided evidence of linguistic variation across situations of use in English (e.g. Bell 1984

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The locative alternation: unaccusative constructions and subject position

PDF

Adding and subtracting alternation: resumption and prepositional phrase chopping in Spanish relative clauses

PDF

A variationist study of relative clauses in Spanish

PDF

Asymetrical discourse in a computer-mediated environment

PDF

Sources of non-conformity in phonology: variation and exceptionality in Modern Hebrew spirantization

PDF

A reduplicative analysis of sentence modal adverbs in Spanish

PDF

Towards a correlational law of language: three factors constraining judgement variation

PDF

The use of culturally relevant authentic materials and L1 in supporting second language literacy

Asset Metadata

Creator Muñoz, Ana Sánchez (author)

Core Title Register and style variation in speakers of Spanish as a heritage and as a second language

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Linguistics (Hispanic Linguistics)

Publication Date 04/02/2009

Defense Date 01/31/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag bilingualism,language acquisition,language contact,OAI-PMH Harvest,register variation,Spanish,Spanish as heritage language,Spanish as second language,style

Language English

Advisor Silva-Corvalan, Carmen (committee chair), Finegan, Edward (committee member), Rueda, Robert S. (committee member), Saltarelli, Mario (committee member)

Creator Email sanchezm@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m342

Unique identifier UC1205725

Identifier etd-Munoz-20070402 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-327976 (legacy record id),usctheses-m342 (legacy record id)

Legacy Identifier etd-Munoz-20070402.pdf

Dmrecord 327976

Document Type Dissertation

Rights Muñoz, Ana Sánchez

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu