|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 LINGUISTIC AND SOCIOSTYLISTIC VARIATION OF THE GENERIC SUBJECT CLITICS ON TU/VOUS COMPARATIVE STUDY BETWEEN FRENCH L1 AND FRENCH L2 By JINGYA ZHONG A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2009
2 2009 Jingya Zhong
3 To my parents Chunmu Zhong and Wangyang Wang
4 ACKNOWLEDGMENTS First and above all, I am deeply indebted to my advisor Dr. Hlne Blondeau, without whom I would never have received the substantial academic training in the field of Sociolinguistics and without whom this study would never have come into existence. I admired and was in spired by her deep knowledge in Sociolinguistics and I was greatly influenced by her enthusiasm in sociolinguistic research. Her personality, teaching and advising style matched my character and learning style so well that it has been such an intellectuall y beneficial and enjoyable experience working with her during the years in the French M.A. program. Words even seem unable to sufficiently express my gratitude. I thank her, from the bottom of my heart, for her interests in my study, for the countless advi ce that she has offered me, for the continuous encouragement, support and trust that she has given me. Many thanks go to my committee member and teaching supervisor Dr. Theresa A. Antes, who revised my thesis so painstakingly and gave a great deal of perti nent and helpful comments that had greatly contributed to the improvement and refinement of my thesis. I also thank her for checking the language quality of my thesis writing. As my teaching supervisor, I can never forget the recognition and understanding that she has shown me. Thanks are due to her if I have become a good and confident teacher. I would like to thank Dr. Gayle Zachmann of Paris Research Center, her assistant Isabel Fuentes Rey, Dr. Brigitte Weltman Aron and Dr. Helene Blondeau, for their invaluable help on my data collection. Without them, I would not have been able to find enough participants for my study. Thanks are also extended to the participants in the study, for their willingness and interests in participating in my study. To Dr. Ca rol Murphy, who taught me how to write a literature paper academically and scientifically, just as how I do for linguistic papers. To Heloise Sailles, for her genuine help and
5 encouragement when I began teaching. Thanks also go to all other faculty and st aff in the department, for their help and advise during my M.A. studies. To my dear colleague and friend Audrey Viguier, who helped me transcribe part of the data. To my other friends and colleagues, who shared two wonderful years with me. My deep love and appreciation goes to my family, who defines me and is part of me. Thank you, dad, mom and bro, for your caring, unconditional love and constant support. Last but not least, I want to thank my husband, Mingliang Wang, for his unfailing love and indulgence An end is another beginning. The thesis is finished; the fascinating world of linguistic research has just opened to me; yet the thanks to the many people who shaped me will never fade.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .........................................................................................................................9 ABSTRACT ...................................................................................................................................10 C H A P T ER 1 INTRODUCTION ..................................................................................................................12 2 THEORETICAL F RAMEWORKS ........................................................................................16 Variationist Sociolinguistics ...................................................................................................16 Sociolinguistics and SLA .......................................................................................................19 3 LITERATURE REVIEW .......................................................................................................24 French L1 ................................................................................................................................24 French L2 ................................................................................................................................31 4 METHODOLOGY .................................................................................................................36 Participants .............................................................................................................................36 Data collection ........................................................................................................................39 Data transcription ....................................................................................................................43 Data encoding .........................................................................................................................44 Circumscribing The Variable Contexts ...........................................................................44 Data Exclusion .................................................................................................................45 Coding .............................................................................................................................48 5 RESULTS AND DISCUSSION ON FRENCH L1 DATA ....................................................57 Overview .................................................................................................................................57 Real time comparison across corpora .....................................................................................60 Linguistic variation .................................................................................................................64 Verbs ................................................................................................................................65 Syntax ..............................................................................................................................68 Pragmatics .......................................................................................................................69 Gender variation .....................................................................................................................74 Stylistic variation ....................................................................................................................76 Co occurrence between the generic TU and the discourse markers tu sais and tu vois .............................................................................................................................77
7 The generic TU and the pronoun of address tu ...............................................................80 Topic ................................................................................................................................81 Concluding remarks ................................................................................................................83 6 RESULTS AND DISCUSSION ON L2 LEARNE RS DATA .............................................84 General patterns of the use of ON, TU, LON and VOUS .....................................................84 General Patterns ...............................................................................................................84 Individual Differences .....................................................................................................86 The linguistic repertoire of expressing generic references in French L2 data ........................87 Fixed Expressions ............................................................................................................87 Sentence Structures .........................................................................................................88 Pragmatic Dimension Of The Generalizations ................................................................89 The use of TU/VOUS in the French L2 data ..........................................................................91 The use of LON .....................................................................................................................95 The acquisition of the sociolinguistic variation of the variable ON TU/VOUS ....................97 Concluding remarks ..............................................................................................................103 7 CONCLUSION .....................................................................................................................104 APPENDIX A INFORMED CONSENT ......................................................................................................108 B QUESTIONNAIRE ..............................................................................................................111 C INTERVIEW QUESTIONS .................................................................................................113 D GOLDVARB DATA ............................................................................................................114 LIST OF REFERENCES .............................................................................................................117 BIOGRAPHICAL SKETCH .......................................................................................................122
8 LIST OF TABLES Table page 41 Demographic charact eristics of participants ......................................................................36 42 General linguistic background of French L2 participants ..................................................38 43 Coding protocol .................................................................................................................49 51 The distribution of ON and TU in the whole data .............................................................57 52 The distribution of ON and TU by sex ..............................................................................59 53 Sex difference in the frequency of ON by young middle class participants in three corpora ...............................................................................................................................61 54 The count, frequency and factor weight of ON and TU according to diffe rent verbs .......65 55 The count, frequency and factor weight of the syntactic factors .......................................68 56 The count, frequency and factor weight of ON and TU according to different pragmatic factors ................................................................................................................70 57 Use of ON and TU by each participant in PFL1 ................................................................74 58 Sex difference in the use of ON TU ..................................................................................75 59 Association between the ON TU and the tu sais / tu vois by each participant ..............78 510 The distributio n of topics in the tokens of ON among Participants #7,#8, and #9 ............82 61 The count and frequency of LON, ON, VOUS, TU in the speech data of French L2 learners ...............................................................................................................................85 62 Comparison of the use of ON, TU, VOUS and LON between PFL1 and UFFL2 ...........85 63 The distribution of ON, TU, VOUS, and LON in the UFFL2 data ..................................86 64 The correspondence between address pronouns and generic pronouns in participants #11, 13, 16 and 19 ..............................................................................................................94
9 LIST OF FIGURES Figure page 51 The distribution of ON and TU in the whole data .............................................................58 52 The distribution of ON and TU by sex ..............................................................................59 53 Frequency of generic ON by young middle class participants in three corpora ................60 54 Sex difference in the frequency of ON by young middle class participants in three corpus .................................................................................................................................62 55 Continuum on the degree of formality and interaction of the pragmatic factors ...............73 56 Association between the ON TU and the tu sais / tu vois by each participant ..............78 57 The distribution of topics in the tokens of ON among Participants #7,#8, and #9 ............82 61 Comparison of the use of ON, TU, VOUS and LON between PFL1 and UFFL2 ...........86 61 U course of the development of the acquisition of TU/VOUS ........................................101
10 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts LINGUISTIC AND SOCIOSTYLISTIC VARIATION OF THE GENERIC SUBJECT CLITICS ON TU/VOUS COMPARAT IVE STUDY BETWEEN FRENCH L1 AND FRENCH L2 By Jingya Zhong August 2009 Chair: H lne Blondeau Major: French In this study I aim to examine and compare the use of the generic pronouns ON TU/VOUS in French as a first language (L1 French) and in French as a second language (L2 French). Previous studies on L1 French have shown that the variation of ON TU/VOUS is constrained by linguistic, pragmatic and social factors. However, some research problems remain to be addressed: while research in Montreal French has shown an increase in the use of the variant TU/VOUS in the place of the traditional variant ON (Laberge, 1977, 1980; Laberge & Sankoff, 1980; Thibault, 1991; Thibault & Develuy, 1989) the potential linguistic change in favor of the TU/VOUS form in France French has not been suggested (Ashby, 1992) ; although the TU/VOUS variant is excluded from formal writing and is consid ered to be nonstandard language, the stylistic nature of the variable ON TU/VOUS has not been empirically tested in spoken French. In French as a second language, with the exception of Blondeau (2008), the acquisition of the sociolinguistic variation of t he variable ON TU/VOUS has not been much studied. Based on the data in this study (a corpus of sociolinguistic interviews totalizing about 20 hours) collected from both French native speakers and American learners of French, the present
11 study contributes to the existing French L1 literature on the topic by showing the following results for the L1 participants: (1) on the linguistic level, the verbs dire and voir favor the use of ON and the verbs vouloir and avoir favor the variant TU; presentatives are found to favor the choice of ON whereas existential sentences favor that of TU; finally the participants used ON predominantly in evaluations and TU in situational insertion where they intended to engage themselves as well as their interlocutor int o their discussion; the distribution of ON and TU is quite balanced in objective description and explanation due to the neutrality of this pragmatic category; (2) on the extra linguistic level, a striking sex difference has been discovered: young men strongly preferred the variant TU while young women in the present study used the generic ON much more frequently than the generic TU; positive evidence has supported the hypotheses that the variable ON TU/VOUS is stylistically marked and is involved in an ongoing change, namely, the participants tended to opt for the generic ON in formal speech style and the generic TU in informal speech style, and the generic TU is taking the place of ON in the speech of young men. Compared to the French L1 data, little variat ion has been observed in the French L2 data in this study. The learners used ON by default and switched to other forms of generic pronoun TU/VOUS/LON occasionally. The extra linguistic factors that influence the use of ON TU/VOUS in French L1 did not affe ct much the use of the variable in French L2. Unlike the French natives who used the different variants to express gender identity and vary speech styles, the learners valued the language norms and showed a narrower sociolinguistic repertoire. Nevertheless, it is suggested that the learners are on the verge of developing towards the acquisition of the native patterns of the variation. Proficiency level in French, contact with French native speakers and explicit instruction each play a role in the use of the variants.
12 CHAPTER 1 INTRODUCTION The variable ON TU/VOUS refers to the generic subject clitics used in French to express generic referen ce. ON, which is translated to one or people in English, is the traditional variant carrying the generic function, as illustrated in Example 1 A. (1 A ) ...dans une petite ville, ON samuse bien un petit moment... ( PFL1 #10)1 While TU and VOUS (you in English) indicate in their standard uses a second person addressee(s), they are also used by French speakers for gen eric meanings and are documented as other variants in competition with ON, both in Montreal French (Laberge, 1977) and in France French (Coneney, 2003) Example 1B and Example 1C illustrate the use of TU and VOUS as generic clitic s : (1 B ) Oui mais TU penses toujours en fait, T(TU) aime s toujours l au dpart, TU vas toujours tre attire par, par l o TU vis pas ( PFL1 # 08) (1 C ) Si ON ne fait rien votre vie ntre pas utile ou trs, VOUS navez pas beaucoup de satisfaction avec votre vie. (UF F L2 # 19)2 Previous studies of French as a first language (French L1) conducted within the Labovian quantitative paradigm have revealed the linguistic and sociolinguistic factors that constrain the variable use of ON TU/ VOUS (Ashby, 1992; Laber ge, 1977, 1980; Laberge & Sankoff, 1980) The variant ON is favored in propositions headed by presentative constructions such as je pense, je trouve, daprs moi, etc, while the variant TU/VOUS is preferred in implicative constructions such as si and quand. At the pragmatic level, morals strongly favor the use of ON. Regarding the sociolinguistic characteristics of the variable, Laberge (1977) found that young men below age 40 use TU/VOUS frequently, while young women prefer ON. While the TU/VOUS variant i s 1 The corpus of Paris French in the present study is named PFL1 (Paris French L1); #10: Speaker #10. 2 The corpus of French L2 learners at UF is named UFFL2 (UF French L2); #08: Speaker #08.
13 excluded from formal writing, it has not been empirically tested as an informal marker. Coveney (2003) found that the use of TU/VOUS did not correlate with th e absence of NE, which is demonstrated in several studies as signaling in formality ( Armstrong, 2002; Ashby, 1981; Coveney, 1996) Therefore, whether TU/VOUS is indeed an informal marker in French remains a question t o be tested. After examining the L1 literature concerning ON TU/VOUS variation, one can imagine many similar research questions in the area of French second language acquisition. In French L2, how do learners use sociolinguistic variables? Do they show sim ilar patterns to those observed in French L1? Do L2 learners speak differently according to their social identities and their perception of speech formality? Second language acquisition (SLA) research, which is interested in adopting a sociolinguistic mode l to examine L2 learners acquisition of sociolinguistic competence, has demonstrated that learners L2 speech can also be influenced by social contexts such as interlocutors, linguistic contexts and time (Tarone, 2007) Applying a variationist approach to study L2 French learners speech variation, researchers in Canada also found that while L2 French immersion l earners follow the linguistic contexts fairly closely, they are not as sensitive as native speakers to styleshifting, and learners who have had more contact with native speakers produced more informal variants than those who have had less (Bayley & Regan, 2004; Mougeon, Nadasdi, & Rehner, 2002; Nadasdi, 2005; Rehner & Mougeon, 1999; Rehner, Mougeon, & Nadasdi, 2003; G. Sankoff et al., 1997; Uritescu, Mougeon, Rehner, & Nada sdi, 2004) The sociolinguistic variables studied within this perspective included, among others, grammatical variables such as the retention versus deletion of the negative particle NE, lexical variables such as verbs indicating residence ( habiter, vivre, rester, demeurer), and phonological variables such as retention versus deletion of schwa (for a complete list of the variables studied
14 and a detailed summary of findings, see (Mougeon et al., 2002; Mougeon, Rehner, & Nadasdi, 2004) Studies that looked at instructed learners sociolinguistic competence had similar findings (Dewaele, 1992, 2004a; Dewaele & Regan, 2002; Howard, 2006; Thomas, 2002) Researchers who examined the L2 speakers sociolinguistic behavior in naturalistic context where they are in daily contact with native speake rs proposed that these L2 speakers can internalize the effect of certain social factors such as gender and participate in the community norm and linguistic change (Blondeau, 2008; Blondeau & Nagy, 1998; Blondeau, Naomi, Sankoff, & Thibault, 2002) Situated in this background, the present study aims to examine and compare the linguistic and socioli nguistic variation of the ON TU/VOUS variable in French L1 as well as the French L2 spoken by instructed learners. On the one hand, based on the speech data collected in Paris in June 2008, the variation of this variable in French L1 will first be studied to see if the patterns correspond to the previous findings, which were based on data collected several decades ago in other regions of France (Ashby, 1992; Coveney, 2 003) On the other hand, speech data collected from advanced American learners of French at the University of Florida is also examined to reveal the patterns of their use of ON TU/VOUS. Instead of applying the factors that were found to influence the use of ON TU/VOUS in the French L1 literature and then verifying if they also have an effect in French L2, I am more interested in adopting a bottom up approach and examining the L2 patterns separately before comparing them with the L1 patterns. I use this m ethod because the linguistic knowledge of more than one language should have its own characteristics, instead of being a simple combination of several different language systems. Therefore, the L2 patterns might have something distinguishable from L1 patte rns that one cannot discover if just examining the pre established factors which are observed in the L1. While I do think that the patterns concerning the use of the ON TU/VOUS variable may overlap to a
15 certain extent in L1 and L2, I am also interested in finding the differences that are due to the different nature of the languages. Learner differences will also be examined to see whether and how we can categorize the learner community into different groups based on their use of the linguistic variable; for example, will a learners speech tell us that he/she belongs to the group of learners who are characterized by having had extensive contact with native speakers? Specifically the present study examines the following questions: 1. In French L1, what are the linguistic and sociolinguistic constraints on the use of the variable ONTU/VOUS? Is the variable stylistically marked? Linguistic factors, the sex of the speaker, and formality of speech will be examined. 2. The same questions as those in #1 will be examined in French L2 data. Moreover, do the proficiency level in French, contact with French native speakers and instruction of French influence the use of ON TU/VOUS in French L2 data? 3. Comparing French L1 and L2, how are they similar and different? What can account for the similarities and differences? The theoretical frameworks used in the present study are presented in the next chapter, and literature in French L1 and L2 concerning the topic is reviewed in chapter three. The methodology of this study is the foc us of chapter four, and results are revealed and discussed in the chapters that follow.
16 CHAPTER 2 THEORETICAL FRAMEWORKS The Variationist Sociolinguistics framework is used in the present study to examine French L1. In French L2, both the variationist approach and SLA interlanguage variation theories will be used to account for the observed patterns. Variationist Sociolinguistics Postulating that language is homogenous, traditional linguistics attributes phenomena that cannot be explained by formal lingui stics to free variation which results from performance errors, mixing of dialects, etc (Gadet, 1992; Laberge, 1977) Conversely, the Variationist Sociolinguistics approach believes in the heterogeneity of language, suggesting that language variation is inherent and systematic, as Gadet (1992) stated: Le point de dpart dans la variation, plutt que dans ce qui ne varie pas, permet de rvler des rgularits, trop systmatiques pour tre le fait du hasard : Labov en conc lut que lhtrognit est une di mension de la structure mme de la langue. Holding this belief, the variationist approach is interested in examining the notion of linguistic variables. A linguistic variable is the abstract representation of the linguistic element investigated, and its v ariants refer to the actual realizations of the variable. One of the first definitions of variants is that different variants of a variable are assumed to possess identical meaning (Labov, 1972a; Meyerhoff, 2006) Variation in the use of a linguistic variable can be constrained by linguistic factors (e.g. phonetic contexts, word identity, syntactic structures, etc;), social factors (e.g. age, g ender, social class, etc;), and stylistic factors (e.g. formality of speech). Categorical constraints may exclude the use of one of the variants, and variable constraints could favor or disfavor the use of a certain variant. An example of a phonological va riable is the retention and deletion of schwa in French. The schwa is obligatorily retained in some contexts such as before the aspirated H, as in le hero
17 Its deletion is normally prohibited when preceded by two consonants and at the same time followed by one or more consonants, as this would result in a cluster of three consonants which is impossible to pronounce, for example, vendredi Other than the categorical use of schwa, the surrounding phonetic contexts and the position of the schwa syllable in the word are found to influence schwa deletion. For example, Uritescu, Mougeon, Rehner, and Nadasdi (2004) revealed the hierarchy of schwa deletion by FrancoOntarian students as a function of phonetic context. According to their results, a sequence of monosyllables not following a consonant or another schwa is the context that favors most schwa deletion, as illustrated in Example 2 A, and a groupmedial monosyllable following a consonant strongly disfavors schwa deletion, as in Example 2B. (2 A ) J(e) me baig nais beaucoup. / Je m(e) baignais beaucoup. (Uritescu et al. 2004) (2 B ) Je pense qu(e) cest difficile. (Uritescu et al. 2004) Regarding stylistic factors, schwa is retained more frequently in reading a passage out loud than in interviews and conversations (Hansen, 1994) In the reading context, speakers from the least privileged social class retained even more schwas than cultivated Parisians, showing a linguistic insecurity, and young speakers drop schwa more frequently than adult speakers in spoken French (Hansen, 1994, 2000) Beyond the level of phonology, whether the variationist approach is valid in grammatical research is controversial among scholars. Gadet (1992), for example, questioned the validity of Labovs variable r ules at the syntactic level by pointing out that phonological variants are interchangeable and can express the same thing because they do not have meanings, however, this characteristic of phonological variable does not exist in syntax. In other words, dif ferent grammatical forms or structures may have different meanings or functions in some usages or
18 contexts, and this difference is always pertinent whenever one of the forms is used. So we cannot assume that all linguistic levels are homologous and equally apply the variable rules to other levels beyond phonology. Nevertheless, the contrary viewpoint was adopted by other scholars. For them, distinctions in referential value or grammatical function among different surface forms can be neutralized in discour se (D. Sankoff, 1988) This means that in some identifiable contexts where different forms are alternated, the distinctions of the forms are not intended and perceived by speakers, hence different grammatical forms or structures, as well as phonological forms, can be considered to be equivalent. Based on the case of French interrogative s tructures, Coveney (1997) demonstrated the possibility of semantic pragmatic equivalence between two forms or structures in syntax. He also stated, however, that not everything is variable in grammar, and grammatical variation is less extensive than phonol ogical variation. The variationist approach is reliable and can be applied validly in examining the ON TU/VOUS variation because this variable meets the five criteria of Coveneys definition of a grammatical variable (Coveney, 1997) : certain occurrences o f the variants of the variable are interchangeable and are able to express the same propositional meaning in some contexts (if we change ON to TU/VOUS in Example 1A, or TU to ON in Example 1B, the meaning does not change); certain tokens of the variants must have the same communicative function in certain contexts (ON or TU/VOUS can both be used to express generalizations); the same lexemes can appear in parallel enunciations that contain the variants (the same lexemes in (1a) can also work for TU/VOUS); normally the variants belong to the same grammatical category (ON and TU/VOUS are both subject clitics); social differentiation in the use of the variants must be attested in order to recognize a sociolinguistic variable (young men use TU/VOUS much more fr equently than speakers from other social groups). Since the variable ON TU/VOUS as a
19 generic clitic has been studied by researchers in Montreal and France who use this approach (Ashby, 1992; Laberge, 1977, 1980; Laberge & Sankoff, 1980) and is demonstrated to be a sociolinguist ic marker, we can be further assured of the validity of applying this approach in the present study. Methodologically, Labovianstyle researchers collect data from people of different demographic backgrounds and use sociolinguistic interviews to elicit dif ferent styles of speech. In Labovs early studies, he incorporated four structured tasks in his interviews, namely reading minimal pairs, reading a list of words in isolation, reading a short paragraph, and talking about general life. The corresponding for mality of the speech context ranged from formal to causal, because less attention is paid with each successive task (Meyerhoff, 2006) Labov also suggested that researchers can manipulate speech style by varying the topics of the interview, because speakers tend to pay more attention and hence exhibit more care ful speech when talking about more formal topics such as life philosophy or education, and less attention when talking about informal or intimate topics such as childhood games and danger of death, so the speech style in this case becomes more casual (Tagliamonte, 2006) Analytically, the varia tionist approach quantifies the occurrences of variants and compares the frequencies in different linguistic contexts or social categories. Statistical models are developed to analyze linguistic variation (Paolillo, 2001; D. Sankoff, Tagliamonte, & Smith, 2005; Tagliamonte, 2006) A multivariate analysis with Gol dvarb X will be performed to reveal the significant factors that influence the choice of the variants ON and TU/VOUS in the present study. This software will be presented in the methodology chapter (D. Sankoff et al., 2005) Sociolinguistics and SLA In second language learning, learners share some universal characteristics in the language learning process, and they also show individual differences that can be affected by cognitive
20 factors and affective factors. How ever, from a sociolinguistic perspective, it is also necessary to view the learner as a social being. Learners identities, such as social class, ethnicity and gender, are related to second language learning, and the dynamic social contexts of language lea rning may also lead to differences among learners (Mitchell & Myles, 2004) Recognizing that second language learners are social beings, it is reasonable to say that social factors such as interlocutor, topic and situational norm affect learners cognitive processing of language, and hence lead to language variat ion (Tarone, 2007) In studies examining second language variation, Mougeon, Rehner and Nadasdi (2004) distinguished two types of variation: Type 1 variation refers to the study of alternations between standard target forms and nontarget forms which are unacceptable in the target language; Type 2 variation refers to the acquisition of sociolinguistic variables th at display variation in the target language and thus would be produced by native speakers also. Although traditional SLA studies focus on the acquisition of standard varieties, it is relevant and necessary to examine Type 2 variation because second languag e learners acquisition should not only be measured by their use of standard forms, but also by their acquisition of the target language variability patterns (Bayley, 2000; Bayley & Regan, 2004; Preston & Bayley, 2008) Examining second la nguage variation from a sociolinguistic perspective, however, requires a theoretical model. From the variationist point of view, learners language also shows the characteristic of structured heterogeneity, which is observed in native language by sociolinguists. Different linguistic and social factors have an effect on the first language, as well as on the second language. As in L1 studies, using the variationist approach in L2 studies has the following advantages: it provides us with information about the distribution of variants in different contexts and among different groups of speakers; it reveals the variable constraints and
21 their relative weights on the choice of variants (Coveney, 1997; Preston & Bayley, 2008) Thus, the variable rules work in L1 studies as well as in L2 studi es, as Preston and Bayley (2008) concluded on the use of variationist approach in SLA research: Even more frankly, we do not know where the elements of a variable psycholinguistic model fit nor do we know where the linguistic levels fit into it, but we do know that native languages as well as interlanguages display facts that can be captured by a device that includes such variable weightings and offers the opportunity for appropriate interpretations of these various forces. Drawing from the resource of soc iolinguistics, but taking into consideration the coexistence of more than one language in the learners mind and the acquisition course of learners, a recent sociolinguistic model developed by Fasold and Preston in SLA was presented in Tarone (2007). This model states that language learners have two grammars in their mind: Grammar 1 is that of their first language, and Grammar 2 refers to that of an additional language. In both Grammar 1 and Grammar 2, learners can choose to use either of the variants of a form that displays variation, and the choice depends on three levels of factors, namely social contexts, linguistic contexts and time. First, sociocultural contexts, such as interlocutors, purpose of communication, social settings and norms of interaction, influence language variation in Grammar 1 as well as in Grammar 2; they can even condition the code switching between L1 and L2. For example, Thai speakers of English L2 used more Thai phonological variants when talking with Thai people than with English native speakers (Tarone, 2007) In addition to language production, social context can also influence learner s acquisition because different social contexts offer different input. Research in this strand in Canada, for example, found that learners with more contact with native speakers produced more vernacular variants than those whose social network restricted speakers to their first language (Mougeon et al., 2002; G. Sankoff et a l., 1997) Second, linguistic contexts can also cause L2 variation; for example, immersion classroom students follow the same hierarchy of linguistic constraints of schwa
22 deletion as that observed in native speakers (Mougeon et al., 2002; Mougeon et al., 2004; Uritescu et al., 2004) Finally, time is an important factor that influences th e extent to which social and linguistic factors can cause L2 variation. Linguistic forms that are acquired later in life are less internalized and automatic than forms that are acquired early in life, therefore the first requires more attention and control than the latter; the time variable also predicts the process by which learners L2 changes over time (Tarone, 2007) More examples about studies combining a sociolinguistic approach and SLA will be given in the literature review section of the present study. In the present study, I use the sociolinguistic model presented above to guide my study. Although I believ e that L1 and L2 variation can both be influenced by the three levels of factors, I also expect that the weighting of the levels or factors on the choice of variants within each level might not be as equal in L2 as in L1. Since the Grammar 2 which builts i n a learners mind might combine the features of that learners first language and the second language, it is not simply the Grammar 1 as in the mind of a native speaker (Selinker, 1972) Therefore, we might expect that for a given sociolinguistic variable, the patterns or the variables that motivate variation in L2 w ill be different from those found in L1. For example, social factors might have less impact on learners than on native speakers, because learners are often not as integrated into the language community as native speakers; therefore, their second language m ay not reflect their social identities as well as their first language. Likewise, linguistic contexts might have more weight on learners choice of variants than on native speakers; for example, advanced learners may be more aware of linguistic rules than native speakers (Coveney, 1998) A comparative study like the present study helps to shed light on the unstable nature of sociolinguistic variables in the L2. Speci fically, it examines the question of whether second
23 language can be as dynamic as first language in displaying social identities and reflecting changes in cognitive processing (change of degree of attention to speech, which leads to style shifting). Social Identity Theory proposes that human beings identify with multiple identities and perceive one or some of them to be salient in a certain situation (Meyerhoff, 2006) The question then becomes: how do second language speakers express their identities in a given situation? Methodologically and analytically, the present study adopts a quantitative paradigm to examine the data and reveal the stratification of the variants, and a qualitative discussion is also used to account for the variation found. In this chapter, I have attempted to describe and interpret the main theoretical frameworks adopted in the present study, namely variationist sociolinguistics and SLA, and explained how they are suitable for the present study. Since my motivation is to describe, explain and compare the patterning of the variants with res pect to different linguistic and nonlinguistic factors in French L1 and L2, a combination of these approaches is necessary.
24 CHAPTER 3 LITERATURE REVIEW Within the variationist framework, t he variable use of ON TU/VOUS has been extensively documented in Montreal French (Laberge, 1977, 1980; Laberge & Sankoff, 1980) Although based on a smaller amount of da ta, research on France French has also found a similar type of variation (Ashby, 1992) In French L2, more than one dozen sociolinguistic variables have been studied in immersion or instructed context s but not including the ON TU/VOUS variable (Dewaele, 1992, 2004a; Dewaele & Regan, 2002; Howard, 2006; Mougeon et al., 2002; Mougeon et al., 2004; Nadasdi, 2005; Rehner & Mougeon, 1999; Rehner et al., 2003; Thomas, 2002; Uritescu et al., 2004) Along with other sociolinguistic variables, ON TU/VOUS has been explored in the French L2 spoken by Anglophones in Montreal who are in daily contact with French speakers (Blondeau, 2008) This chapter summarizes the findings of previous studies related to ON TU/VOUS variation in French L1 and the acquisition of other sociolinguistic variables by L2 users. In addition, the contribution of the current study to the existing literature will b e discussed. French L1 The subject clictics in French include je, tu, il, elle, on, nous, vous, ils, and elles They can all be used as definite personal pronouns However, on, tu, vous, and ils can also be used as indefinite personal pronouns. Since the interests of the present study focus on ON, TU, and VOUS as generic subject clitics, on, tu and vous as definite clitics are excluded in data encoding; however, examples of on, tu and vous as definite clitics are presented in the section on data encoding. Three studies examined the linguistic and sociolinguistic variation of the ON TU/VOUS variable, namely Laberge (1977), Ashby (1992), and Coveney (2003). Laberges study on ON -
25 TU/VOUS variation was based on vernacular speech data collected from 120 Francophones in Montreal. The data was considered to represent the everyday speech of Francophones in Montreal. Large data size allowed the researcher to extract rich tokens of the variants ON and TU/VOUS, to perform multivariate analysis and to provide substantive results and conclusions. After the study by Laberge, Ashby examined the variation patterns of ON TU/VOUS in his Tour corpus formed in 1976. As for the stylistic effect of the variants, Coveney (2003) examined the correlation between the negative particle NE, of which the presence shows a formal style, and the use of ON TU/VOUS. A synthesis of the findings from the previous studies, regarding the linguistic constraints, social constraints, and stylistic constraints, is provided below. First, concerning the syntactic constraints, a number of lexical and syntactic indicators of generalizations could involve the variable use of ON TU/VOUS, and some of those indicators have a significant influence on the choice among the variants. Additionally, discursive and s equential factors also constrain the use of the variable. The enunciations where the variable ON TU/VOUS appear can be grouped into different types, based on other generalization indicators and common semantic functions. The three enunciation types labele d by Laberge (1977) are generalizations, implicative constructions, and propositions headed by presentative constructions. Generalizations refer to the enunciations where there appear a number of lexical items, and morphological and syntactic structures that are indicators of generalization. The indicators dissociate reference from specific times and places, for example, infinitive verbs such as travailler, lire adverbs like toujours, de nous jours, and aujourdhui as illustrated in 3 A: ( 3A ) Jen ai peu t tre rega gn un peu...A part a, trava iller puis lire ON samliore toujours un peu.
26 Ive perhaps made a little progress Besides, in working and in reading one always improves somewhat. (Laberge & Sankoff, 1980) Def inite articles le, la, les indicate categories, hence lead to generalized context where the generic ON or TU /VOUS can be found : ( 3B ) ON choisit les amis. One chooses ones friends. (Laberge & Sankoff, 1980) Indefinite noun phrases such as une personne or quelquun also serve as coreferent to the generic subject clitic: ( 3C ) Quand une personne levait une famille dans ce temps l, VOUS tiez pas capable davoir de luxe When a person was bringing up a family in those days, you couldnt have any luxuries. (Laberge & Sankoff, 1980) Definitional structures with cest introduce a definition where locators tend to make a generalization: ( 3D ) Linconvnient cest qu ON fait des rparations dans un logement, et puis qu ON part, on est oblig de la i sser a l. The trouble is that you make repairs in a rented place, but when you leave, you have to leave it all there. (Laberge & Sankoff, 1980) C ertain existential structures with the expression il y a or the verbs avoir, voir, trouver also help to identify generalizations: ( 3E ) Le jo ual, cest une d formation, comme TU as des patois en France. Joual is a deformation, like the patois you get in France. (Laber ge & Sankoff, 1980) The lexical and grammatical elements presented above are typical in what Laberge called generalizations
27 An examination of other types of enunciations shows rule based behavior governing the choice of variants as well. An implicative construction is composed of two propositions; the subordinate clause, normally headed by the conjunctions si or quand, introduces a clause, and the principle proposition presents the consequence or effect of the cause. The last type, the presentati ve constructions is an enunciation headed by forms such as il me semble, disons, je pense, je trouve, il est certain, etc. Speakers tend to generalize their opinions or express truths and commonly believed morals in this t ype of enunciation. Examples (3F) and (3G ) illustrate these two types of enunciations respectively. ( 3F ) Bien si ON laisse faire les hom m es cest tout des grosses btes. Well if one lets men do what they want, theyre all big brutes. (Laberge & Sankoff, 1980) ( 3G ) Daprs moi, cest pas ave c des guerres que TU russis faire un pays, TU tassis puis TU discutes. As far as Im concerned, you dont build a country with wars, you sit down and you discuss. (Laberge & Sankoff, 1980) The variable rule analysis revealed the lexical and syntacti c constraints on the choice among the variants. ON is used in frozen forms with the verbs appeler or dire, for example, comme on dit, quon va dire. Generic TU is used categorically when the generalization involves a past tense verb, and when the varia ble ON TU/VOUS co occurs with ici, because in these two contexts, the use of ON will be interpreted as we instead of one, as illustrated in the following examples from previous studies ( 3H ) VOUS aviez pas une cenne pour vous acheter un habillement toutes ces annesl. La femme non plus. You didnt have a cent to buy clothing for all those years. The wife neither.
28 ( 3I ) Oui cest pas la mme chose, a restait chaud plus longtemps aussi. Cest pas pareil. Ici VOUS chauffez l, puis le manger est froid tout d e suite. Yes its not the same, it used to stay warm longer too. Here you heat, and the food gets cold right away. For the variable contexts where either ON or TU/VOUS can be expected to occur, Laberge (1977) found that the implicative constructions fa vored the use of TU/VOUS, while the presentative constructions favored the use of ON. The hypothetical nature of an utterance with si or quand diminishes the ambiguity between the address pronoun tu and the generic pronoun TU, and thus favors the use of the generic TU, and the metalinguistic quality of the presentative forms such as je pense favors ON because of its inherent formality and distance. Although overall Ashby (1992) concluded that the ON TU/VOUS variation in France French followed similar patterns to those observed in Montreal French, he provided a different finding from Laberges regarding implicative constructions. While Laberge (1977) grouped si clauses and quand clauses into the same category, Ashby (1992) separated them in his data analysi s and found that while TU/VOUS was favored in si clause, it was not in quand clause, instead, it was ON that was predominantly used in quand clauses. In addition to the syntactic constraints presented above, some discursive factors are also found to have a n effect on the choice among the variants. Laberge (1977) examined the effect of two discursive categories, namely situational insertion and formulation of morals or truisms. The first category refers to utterances where the speaker generalizes an acti vity or context based on his or anyone elses experience, for example: ( 3J ) Ben on a qu prendre lautobus puis ON se rend compte comment les jeunes parlent ; disons a cest le mauvais language. Well, one only has to take the bus and one realizes how young people talk. Lets say its poor language. (Laberge, 1980)
29 The second category involves the contexts where speakers evaluate a situation, expressing reflections based on conventional wisdoms, like the following utterance: ( 3K ) Jai horreur dentendre crier, puis pourtant faut que je crie avec les petites. Ca me dplat beaucoup. Mais je me dis, des enfants cest des enfants, ON peut pas les faire penser comme des adultes, ON peut pas. I cant stand to hear yelling, but just the same, with the little ones, I have to yell. I really find it unpleasant. But I say to myself, children are children, one cant make them think like adults --one just cant. (Laberge & Sankoff, 1980) The results showed that ON is very much preferred in formulation of morals or truisms and TU/VOUS is more frequent than ON in situational insertion, except in those propositions headed by presentative constructions, where ON is predominant. In contrast to Laberges two way categorization of discursive types, Ashby further specified four types of discursive effects, namely referent introductio n, evaluation, situational insertion and exposition. Speakers introduce a new referent in referent introduction, which Laberge called existentials, with avoir as in example (3e). Speakers evaluate or judge a person or a thing in evaluation, which includes but expands what Laberge labeled morals and truisms. Situational insertion for Ashby only refers to those generalizations that are based on the speakers personal experience. The generalizations based on a third party or on common knowledge are grouped into the last category, exposition. GoldVarb probability analysis showed that evaluation and exposition favor the use of ON, and the other two categories favor the use of VOUS1. So Ashby (1992) confirmed and developed Laberges findings of the effect of the discursive factors on the variable use of ON TU/VOUS. A t the social level, age and sex are the common factors that Laberge (1977) and Ashby (1992) looked at. Laberge (1977) showed that the variable use of ON TU/VOUS is correlated 1 Very few tokens of TU are fo und in his data.
30 with speak ers age and sex. Older speakers above age 40 prefer the conservative form ON while their younger counterparts below age 40 opt for the innovative form TU/VOUS more frequently than ON. Looking at the interaction between age and sex, Laberge found that whil e young males used TU/VOUS very frequently, young females preferred to use ON; however, the difference of the patterns observed between older males and older females was very slight. ON is therefore considered a female variant, and TU/VOUS is frequently us ed by young males below 40. Considering the discrepancy of the use of the variable between the age groups, one might hypothesize that the variable was undergoing a linguistic change. Actually, it was confirmed by Thibault (1991) that her 25 participants increased threefold their use of generic TU from 19711984 (Thibault, 1991) Nevertheless, Ashby found that in his data, age did not play a role in influencing the choice among variants, although he did fi nd the same correlation between sex and the use of ON TU/VOUS. He therefore concluded that the variable is a stable one, but not one that shows ongoing change. More recent data is needed to look at the evolution of the variable and to shed light on the sit uation. Concerning the stylistic effect of the variants, ON is considered to be the formal and standard variant and TU/VOUS is viewed as the non standard form and excluded from formal writing. Coveney (2003) examined the correlation between the well known stylistic marker presence and absence of NE and the use of ON TU/VOUS and attempted to test the stylistic effect of ON and TU/VOUS. One of his findings that is of particular interest to the present study is that although the young participants who were l ess than 23 years old retained just 10% of NE, they still used generic ON in 76% of contexts. So is ON indeed a formal variant and TU/VOUS an informal variant, or is the NE variable not truly indicative of the overall formality of the speech style due to the frequent deletion of NE in modern spoken French (Ashby, 2001; G.
31 Sankoff & Vincent, 1980) ? The present study attempts to re examine the stylistic effect of ON and TU/VOUS and to clarify the situation. Finally, regarding the correspondance between the address pronoun TU/VOUS and their generic variants, Coveneys (2003) study found that in contrast to Montreal French in which the pronouns of address tu and vous do not match the generic subject clitics TU and VOUS, France French showed a correspondence, for example, speakers in the Picardy corpus who used tu to address the fieldworker als o opted for TU as a generic subject clitic in their discourse. In this subsection, I reviewed the French L1 literature about the ON TU/VOUS variation and proposed gaps that I attempt to fill in with the present study. Specifically, by basing my study on pr evious studies, I will look at the movement of the ON TU/VOUS variable to see whether there is linguistic change signaled by the variable in France French, and whether the variable is stylistically marked. French L2 Since the language learners used in the present study are classroom instructed learners, I will focus my review on previous studies that were also based on data collected from instructed learners and immersion students, although a comparison with L2 French speakers in a naturalistic context migh t be relevant. The second language variation involved in the present study refers to the Type 2 variation defined by Mougeon, Rehner, and Nadasdi (2004), which was explained in the last chapter (Mougeon et al., 2004) Various L2 researchers have focused on different sociolinguistic variables that display variation in French L1 and have found similar results. Regan (1995, 1996, 1997) studied the acquisition of native patterns of NE variation by native speakers of HibernoEnglish, and found that the students who had stayed a year in France approached the native patterns of NE deletion, w hile the classroom students in the study used a minimum of the vernacular variant of NE
32 (Regan, 1995, 1996, 1997) Similar findings are provided by other studies in both European and American contexts. Dewaele (1992 2002, 2004a) examined the same variable, NE, in Dutchspeaking instructed learners of French; his studies uncovered both endogenous and exogenous extralinguistic factors that are associated with the deletion of NE. Specifically, extraverts deleted NE mor e frequently than introverts, and speaking with a native interviewer triggered a higher rate of NE deletion than speaking with a nonnative interviewer, which demonstrates a convergence effect (Dewaele, 1992, 2004a; Dewaele & Regan, 2002) Howard (2006) compared the use of three sociolinguistic variables, namely liaison, /l/ deletion (in wordfinal position of subject pronouns il, elle, ils, elles ; in word initial position of object pronouns in Canadian French le, la, les, lui, leur ; in word median position of some words such as table and quelque ), and subject verb agreement in third person plural contexts, between the Irish studyabroad learners and the instructed learners in his study (Howard 2006) Using a similar approach, Thomas (2002) examined the variables liaison and schwa deletion in studyabroad students and instructed students (Thomas, 2002) Sax (1999, 2000) studied the acquisition of the sociolinguistic variation of the variable deletion of NE and /l/ (Sax, 1999, 2000) Their studies corroborated with Regans by showing that studying abroad and contact with native speakers greatly contributed to the approximation of native like patterns of sociolinguistic variation, and that after a year spent studying abroad, learners used more informal variants. Since the TU/VOUS variant as generic subject clitic is not explicitly presented in textbooks and taught in class, students can only learn it by exposing themselves to informal speech, either in naturalistic contexts or through media. Hence studying abroad undoubtedly offers an ideal opportunity for the acquisition of the informal TU/VOUS, and its variation patterns with ON.
33 Studies of the same thread in Canada, which examined the acquisition of sociolinguistic competence by immersion students, revealed similar findings. Thirteen sociolinguistic variables at different linguistic levels are examined. Variants are placed on a socio stylistic continuum: vernacular, mildly marked and formal variants. Vernacular variants do not conform to the norm of standard French, are typically used in informal speech and excluded in formal settings and writings, and are associated with speakers from lower social strata, hence stigmatized. Mildly marked variants do not conform to the rules of sta ndard French either, are also typically used in the informal speech style, but may also be used in formal situations. Formal variants are the forms that are considered to be the norm of standard French and are used in formal writing and speech. The results showed that immersion students generally followed the linguistic constraints observed in French L1 (Blondeau & Nagy, 1998) on the use of the variables. However, their stylistic repertoire is much more restricted than native speakers. Th ey made no or minimal use of vernacular variants, used mildly marked variants at a rate considerably lower than that of native speakers and over used the standard forms that they learned at school. Interaction with native speakers and exposure to French language outside of class was also found to be an important factor that led to a higher rate of informal variants (Mougeon et al., 2002; Mougeon et al., 2004; Nadasdi, 2005; Rehner & Mougeon, 1999; Rehner et al., 2003; Uritescu et al., 2004) Even in a naturalistic context of French in Montreal, French L2 speakers appropriate use of discourse markers and rate of subject doubling are influenced by their extent of integration into the Francophone community, with the speakers who had more contact with Francophones exhibiting more similar patterns of variation than those who had less cont act (Blondeau & Nagy, 1998; Blondeau et al., 2002; G. Sankoff et al., 1997) The use of the ON TU/VOUS variable by Anglophones in Montreal, examined in Blondeau (2008), provided an inspiring source of
34 comparison with the present study. While L2 speakers in a naturalistic context alternated the variants and conformed in general to the lingui stic constraints observed in French L1 (Blondeau, 2008) one wonders if the instructed learners in the present study, who had various degrees of contact and interaction with native speakers, will behave in a similar or different way? The comparison may s hed light on the influence of social contexts in which a second language is acquired. In addition, an examination of L1 transfer may contribute to the existing literature. Specifically, in spoken English, the second person pronoun you can serve as an indefinite grammatical person, sometimes to denote the speaker, sometimes to denote the interlocutor and sometimes to denote everybody or anybody (Staels, 2004) Furthermore, Staels showed that you is also anaphoric and in some cases refers to something else beyond the original antecedents referent. He also demonstrated the use of you in existential expletives. He finally ar gued that it is the polysemy and dynamism that make you form attractive and powerful as indefinite pronouns. The use of the indefinite you can be illustrated in the following extract from my data: 1. yeah and how to whats the effect of stay abroad? 2. ok, and I think it is the most important way YOU can or the best way to improve your language, its definitely immersion, living somewhere where they speak the language YOU trying to learn hum because theres always a lot of varieties what your teacher teaches YOU as opposed to what people are currently speaking, because language is always changing and then textbooks dont always keep up (UFFL2#19) So would French L2 learners transfer indefinite you into French tu or vous? This question is within the scop e of discussion of the present study. In summary, L2 French studies that focus on the acquisition of sociolinguistic variables consistently revealed that while for some variables, if not all, learners generally followed the ordering of linguistic constraints, they were markedly different from native speakers with respect
35 to style shifting. As for the independent variables that favor the acquisition of informal variants and native like patterns of variation, contact with native speakers and degree of exposur e to French were proven to be influential.
36 CHAPTER 4 METHODOLOGY In this section, I will discuss the demographic and general linguistic information about my participants, and then describe the data collection procedures, data encoding procedures and tools used for data analysis. Since I am studying a sociolinguistic variable sociolinguistic interviews were conducted to collect data. As I attempt to examine variation patterns of the ON TU/VOUS and have adopted a perspective that combines Variationist Sociolinguistics and SLA, both quantitative and qualitative analysis will be done. Participants Table 4 1. Demographic characteristics of participants Native speakers of French French learners F irst language French English Second language N/A French or othe r languages Place of birth Paris United States Sex 5 females et 5 males 7 females et 2 males Year of birth 1984 1989 1985 1988 Socioeconomic background Middle class Middle class Profession University student Universi ty student Table 4 1 gives the ge neral demographic information and background of the participants in this study. In total, ten French L1 participants were recruited in June 2008 in Paris2,. Eight were recruited through the Paris Research Center (PRC) of the University of Florida (UF) one was a friend of a friend, and one was recruited through an online poster. Except for one female participant (#02) and one male participant (#07) who worked at the PRC at the time the data was collected, all the other participants attended a university in P aris, for example, Paris III, Paris VII, Paris XIII, or FACO (Facult libre dconomie, de droit et de gestion). Some were 2 Thanks to the Conner Award provided by UF French program for funding this research and many thanks to Dr. Galye Zachmann and her assistant Iasbel Fuentes Rey from University of Florida Paris Research Center for their invaluable help.
37 undergraduate students, and the others were M.A. students. Their majors varied from Languages to Computer Science. They were all born and grew up in Paris, although some of them have parents who came from other countries. Specifically, Participant 01s family was from Thailand, the parents of Participant 03 and Participant 06 were from Spain, and Participant 10 has Cantonese parents. So me of the participants belonged to the same social network, because they were friends with each other. Since the present study is a comparative study, sex, age range and socioeconomic class were controlled. I succeeded in finding five males and five females for the French L1 group; all participants in French L1 group and L2 group were from 20 to 25 years old, and all were from middle class families. Table 4 2 summarizes the general linguistic information about the French L2 participants. Like their French L1 counterparts, they were all university students; three were M.A. students in the French program at the University of Florida, and six were undergraduates enrolled in a 4th year French undergraduate class or a graduate class. Only 2 males were recruited for the French L2 group, due to logistics in these programs. The French L2 participants participated in the present study in November 2008. The assessment of their general proficiency level of French is based on two types of information: first they are enr olled in a 4000level or 6000 level class which requires that they have already completed a certain number of classes conducted in French, and have achieved a minimum proficiency level (generally, intermediate high or better); second their self evaluation in the questionnaire indicates that they consider themselves as advanced learners. Since the present study examines their oral production, it is also necessary to know the oral proficiency level of the L2 participants. This evaluation was made by me during the first part of the interview with the participants. The assessment followed the American Council of Teaching Foreign Languages (ACTFL) guidelines and participants were evaluated as
38 Advancedlow, Advancedmid, and Advancedhigh3. Procedures of the inter view and more details on their evaluation will be presented in the next section. The French L2 participants all speak English as their native language, and most of them speak other second languages at different self evaluated proficiency levels. Except par ticipant #17, they have all had stays in France or other Francophone countries. Most of them took classes there, but some simply traveled or spent vacations there. More information about their French use during their stays abroad and in Gainesville were obtained through the interview and will be discussed in the Results and Discussion chapter. Table 4 2. General linguistic background of French L2 participants Participant # General proficiency level in French Oral proficiency level of French Years of French study Time spent in Francophone countries Purpose of stay in Francophone countries Other second languages 11 Advanced Advanced high 11 8 months in France Class, vacation, teaching Beginning Spanish 12 Advanced Advanced high 16 2 years in France class Intermediate German and Spanish 13 Advanced Advanced mid 4 2 months in Geneva class Advanced Russian, Intermediate Spanish and Arabic 14 Advanced Advanced mid 11 6 months in Lille class N/A 15 Advanced Advanced mid 6 3 weeks in Paris class Beginning Hebrew 16 Advanced Advanced low 8 6 months in France class Beginning Spanish 17 Advanced Advanced low 5 N/A N/A Advanced Spanish, Beginning German 18 Advanced Advanced low 7 8 7 months in Paris vacation Beginning German and Hebrew 19 Advanced Advance d low 7 1 week travel Beginning Dutch, Intermediate Mandarin 3 I am not ACTFL certified, but followed their guidelines to the best of my ability.
39 Data collection All participants in the present study were first asked to sign an informed consent form approved by the Institutional Review Board (IRB) of UF (see Appendix A). Individual soci olinguistic interviews were conducted and recorded with French L1 participants in a classroom at the PRC. Each interview lasted approximately one hour. By controlling the formality of topics, I attempted to elicit both formal and informal speech styles. To pics of the interviews ranged from formal to informal, including economics, education, language, society, studying, working, general life, childhood memories, childhood games, leisure activities, personal experiences, friends and families, etc. Although th e sociolinguistic interviews were conducted in free conversational mode, rather than simple question answering, some questions were preplanned and asked naturally at appropriate moments in order to elicit more tokens of the variable; for example, I asked participants personal opinions about the advantages of traveling, of studying abroad, of the economic situation in France, etc.; I also asked participants to explain to me the rules of their childhood games, how to prepare a French dish, etc. The pragmatic nature of evaluations and explanations often leads to generalizations, and hence offers more chances for participants to use the generic subject clitics ON or TU/VOUS. At the beginning or at the end of the conversation4, participants were asked to give th eir background information, and I noted this down. A very small Sony digital voice recorder was used to record the interviews, in order to make the participants feel comfortable and thus reduce, as much as possible, the observers paradox. The observer s paradox is the concern of many sociolinguists who hope to collect naturally occurring speech data. It refers to the paradox that sociolinguists want to know how 4 I asked the first participants for their background information, but later realized that it would be better to do this after the interview to avoid observers paradox.
40 people talk when they are not being observed, however, very often the presence of a voice re corder and interviewer may cause the participants to monitor their speech. One of the solutions to this problem that Labov (1978) proposed is to involve participants in topics where their emotions override the interview situation (Labov, 1978) Some of the topics that were discussed by the participants in the present study did engage the interest of the participants, as shown by their higher pitch and their gestures. Based on social network and exchange theories, Milroy (1980) suggested that, in order to have participants talk volubly, a fieldworker should provide some kind of goods or services, because in this case, participants are engaged in obligations t o return the goods or services. The tokens of exchange provided by the fieldworker may be material, or in the form of personal qualities such as sympathy or interest shown while participants are talking (Milroy, 1980) Therefore, to make sure that the participants would talk volubly, compensation was provided during the data collection of this study in exchange for a one hour interview; interest and sympathy were also shown during the conversation. Whether the participants in the present study were motivated by the compensation provided, by interest in my study, by willingness to help out or by a combination of multiple motivations, they were very willing to talk and to explain thi ngs to me, and some were even interested in my study or myself and asked me questions. Most of the time, the atmosphere was relaxed and natural. However, one must note that despite the fact that efforts are made by sociolinguists to reduce the effect of th e observers paradox and to make participants talk volubly, there are factors that fieldworkers cannot control; for example, participants personality may influence, to a certain extent, the atmosphere of the interview situation. Participants who are extro verts might talk more and talk more easily than those who are shier. In the present study, participant # 10 belonged to the latter category and exhibited a relatively formal speech style, as his linguistic features showed. It is
41 interesting, however, to turn the observer paradox to good use and compare the use of ON TU/VOUS between a more formal speech style, like that of participant # 10 and a very informal one displayed by another participant, as we will do in the next chapter. So in general, techniques w ere adopted to make the overall interview situation as natural as possible and topics were controlled to elicit both formal speech and informal speech. The linguistic features of the speech samples, which will be discussed in the chapter concerning results and discussion, demonstrate that this was the case. Speech samples of L2 French learners were collected individually, through informal interviews, in a conference room at UF. After the L2 French participants signed the informed consent form, they were the n asked to complete a questionnaire, the objective of which was to obtain background information on their previous French studies and their use of French (see Appendix B). Afterward, an informal oral interview was conducted. The interview was composed of t hree major parts: a warm up, an oral proficiency interview, and an informal interview. During the data collection stage of the present study, based on the responses provided by some of the early L2 participants, I got some interesting ideas that were not in the initial proposal of my study: namely, that a short English interview might shed light on what English equivalents of ON TU/VOUS variable would be used by the participants and might help me to see if there was L1 transfer in their use of ON TU/VOUS; and also that a reflective interview about ON TU/VOUS might also help me to interpret their use of this variable. So at the end of the interview with some of the participants, I gave a short five minute English interview that was based on some of the same questions asked in French. Additionally, I asked a couple of the
42 participants about their general linguistic awareness of the use of the ON TU/VOUS variable. Again the results will be revealed and discussed in the following chapters. Let us now look in mor e detail at each part of the interview (a list of sample questions can be found in Appendix C). Considering that the L2 participants are second language learners and might be nervous when engaging in conversation in French, it is necessary to put the parti cipants at ease. The warm up consisted of general greetings and social amenities that are normally used in everyday conversations. Afterward, in order to check the oral proficiency level of the participants, several tasks were used in the oral proficien cy interview (OPI). The design of the tasks followed the guidelines of the ACTFL Oral Proficiency Interview and Stimulated Oral Proficiency Interviews (Omaggio, 2000; Stansfield & Kenyon, 1996) I also used samples of the Test of Spoken English from ETS as a model (www.ets.org). Specifically, participants were asked to narrate a story in a movie or a book, to describe a place th at they were familiar with, to give directions to a place, to express and support opinions on certain topics such as comparing advantages and disadvantages of living in a big city and a small city. These tasks corresponded to the ACTFL OPI level check, and for some of the participants who handled these tasks well, I further added a role play to check if they could handle a task that was difficult to test directly in an interview situation, for example, they were asked to give complaints about a textbook t hey had ordered a month ago but that had not arrived. This part of the oral proficiency interview lasted about 15 minutes. Although I intended to check the participants proficiency level in French during this part of interview, I did not tell this to the participants, because I did not want the participants to perceive the situation as a formal interview. All of the L2 French learners demonstrated competence at the advanced level in performing the tasks in the level check phrase, they were able to narrat e and describe in past, present, and future tense, and they were
43 able to give and support their opinions. According to their accuracy and fluency in performing the tasks, I further grouped them into Advancedlow, Advancedmid, Advancedhigh and Superior (for a detailed description of each level, see the ACTFL speaking guidelines in Appendix A in (Omaggio, 2000) ). After an oral proficiency level check, interviews continued and focused on a variety of topics. To make sure that the speech samples collected from French L1 participants and those collected fr om L2 participants were comparable, I kept the topics of the interviews and types of questions similar. Nevertheless, given that L2 speakers may not be as voluble as L1 speakers, more questions were prepared and asked. A complete individual interview laste d about one hour. Data transcription All the recordings were transcribed precisely in Word documents and according to Standard French and English spelling. Each participant was assigned a number. The speech data of French L1 participants were transcribed and checked by me and one of my French colleagues. This is to guarantee the accuracy and quality of the transcription. Because the speech samples of second language participants were structurally less complex, they were all transcribed and checked by myself In the transcription, 1 represents the interviewer who is me, and 2 represents the interviewee. I followed the transcription convention in Thibault and Vincent (1990) (Thibault & Vincent, 1990) Transcription reflects exactly what the participants actually said, errors (from the point of view of prescriptive grammar) were also transcribed, and nothing was modified. An excerpt of a transcription is provided below: 1 : Oui. On a mliore beaucoup la langue quand on est dans le : < 2 : Oui, oui > l e pays 2 : Oui, oui je pense oui, vu qu ON e st oblig de parler cette langue < 1 : mm> Sauf si videmm ent ON a, ON reste entre groupe de : par exemple si jallais en Angleterre et qu e je res tais entre franais ON va pas beaucoup voluer. Mais si je, je vais vraiment dans une famille E uh, je pense que oui l, ON est oblig de de parler de parler anglais de : de de lire,
44 denten, ON entend que de langlais partout donc euh, a aide, a a aid e beaucoup quoi. (PFL1#01) Data encoding Circumscribing The V ariab le C ontexts Using variable rules to examine the variation of ON TU/VOUS, one must identify the variable contexts and exclude the contexts that do not contain the variants under investigation. The variable contexts identified in the present study involve those where ON, TU and VOUS appear as generic clitics, referring to an indefinite person who could be the speaker, the interlocutor, or anybody. In addition, ON, TU and VOUS should be theoret ically interchangeable in the contexts. The following examples, Example 4A from Participant # 04, and Examples 4B and 4C from Participant # 08, demonstrate the variable contexts of ON TU/VOUS: ( 4A ) [Talking about advantages of going abroad] Ben quand ON va ltranger ON apprend, cest sr que cest plus : cest bnfique pour la langue parce quon, on ON apprend beaucoup plus vite ON apprend pas seulement des mots : dont fin dont ON a besoin pour : pour le pour : lcole, fin, pour expliquer un texte et t out a, ON apprend vraiment les mots de la vie courante, cest l quON se rend compte quON a beau avoir fait 5 ans dtudes de langue, et au final, ON sait pas dire (elle parle lespagnol) fin des trucs : vraiment tout simple, et donc ON les apprend l bas (PFL1#04) ( 4B ) 1. Et pourquoi est ce que tu es all che rcher un stage aux Etats Unis, ltranger? Parce que je trouve que cest dj moi jaime beaucoup les Etats Unis, jaime beaucoup langlais, donc, ben voil. Et parce que je trouve que partir ltranger en gnral cest enrichissant (). Ben, Tapprends la langue, Tapprends des trucs, Tapprends connaitre de nouvelles cultures, de nouveaux et de nouvelles personnes rencontrer, cest intressant, cest enrichissant! (PFL1#08) ( 4C ) Oui mais TU penses tou jours en fait, T aime s toujours l au dpar t, TU vas toujour s tre attire par, par l o TU vis pas. (PFL1#08) In the examples, the participants were clearly making indefinite reference because they used present tense with the clitics, and they used other generalization markers such as cest, les, le, la, toujours (Laberge, 1977, 1980; Laberge & Sankoff, 1980)
45 After identifying the variable contexts of ON TU/VOUS, the sentence where the variable appeared was extracted, and the preceding and following sentence were also extracted, when necessary. Data E xclusion Since ON and TU/VOUS appearing in the data coul d be generic clitics or address pronouns, one needs to carefully examine the function of each occurrence of ON and TU/VOUS and identify the true variants, especially for ON, which may be interpreted as definite pronouns nous, ils, je or tu/vous (Ashby, 1992; Laberge, 1977; Laberge & Sankoff, 1980) The definite pronoun on and the address pronouns tu and vous in the transcription are italicized, and th e alternatives of the definite pronoun on illustrated in the examples of the present study are also italicized. In contrast, the generic subject clitics ON and TU/VOUS under investigation are capitalized. The function of on must be examined in the context. It can be used to substitute for nous as illustrated in the discourse of Participant # 02 in (4 D). In this example, on and nous both refer to the participant and her classmates who were preparing for their competitions. (4 D ) [explaining her year of preparation to enter university] 2. En fait cest une anne de prpa pour nous f aire rentrer dans les coles, les coles nationales franaises. 1. Donc, Pendant combien temps est ce que tu es rest dans lcole ? Un an ? 2. Mais on peut le faire en deux ans mais, enfin, si on na pas nos concours ou si on veut approfondir plus avant de rentrer dans une cole. Cest une cole prive. (PFL1#02) on can also vary with ils, je or tu as in examples (4e) and (4f) from the same participant: (4 E) [ Talking about t he advantages of living in Sweden] je sais pas ils sont socialement parlant la Sude cest quand mme un grand exemple de : je sais pas tout tout se passe bien quoi, cest vrai quy a moins de gens donc moins de pauvres et moins de de probmes sociaux, mai s quand mme cest pas mal. Et : et oui et ce qui est gnial cest que on peut faire ( ) ya des choses qui sont engages par exemple, lhiver on peut aller au boulot au en ski, donc ON a des gens dans le mtro qui font qui font avec leur ski et tout pour une station par exemple, cest pas mal. (PFL1#02)
46 (4 F) 1 Quand tu coutes quelquun parler franais, tu peux distinguer quel pays ou quel quel endroit est ce quil vient ? 2 Le sud oui : sinon, non. Je pense, une fois qu on [4 F 1] en a rencontr un, on [ 4F 2] peut on [4 F 3] peut reconnatre dautres personnes, je veux dire une fois qu on [4 F 4] a entendu laccent oui, mais : fin genre quand on [4 F 5] a rencontr vraiment des personnes, des amis, quoi je vois le sud j y vais tout le temps donc, je sais comment ils parlent, mais : mais Lille, par exemple, fin jy vais souvent Lille, mais je sais pas, des gens, ils sont, on [4 F 6] a pas tous laccent, comme : quand je parle, dans le sud, tout le monde na pas laccent du sud. (PFL1#02) In (4 E ), based on the context, one can determine that on refers to the people in Sweden, which is also what ils in the same discourse refers to. Similarly, in (4 F), both on [4 F 6] and ils nearby refer to people in Lille. In addition, je could also be used to replace th e definite on, as in Example (4F), at the beginning of the answer of the participant; the on from [4 F 1] to [4 F 5] that she used must refer to je Three indices here can help to identify the function of on as je : first of all, the interviewer asked the participant about her personal experience using the address pronoun tu in the question, it is logical that the participant was talking about herself; second, the past tense verbs following on ([4 F 1], [4 F 4] and [4 F 5]) increase the specificity of time and hence attenuate the effect of generalization; finally the following occurrences of je give a sign that the participant was referring to herself. The uses of on and tu/vous as definite pronouns are thus excluded from coding and analysis. In addition, tu and vous as address pronouns are excluded from the data analysis and hence are not extracted, as illustrated in the following examples, where tu and vous clearly referred to either the interviewer or the interviewee: ( 4G ) 1 Est ce que oui est ce que vous avez besoin dune copie, de a ? (PFL1#09) ( 4H ) 1 Donc tu tes bien adapt cette vie l ? (PFL1#09) ( 4I ) 2 Daccord pour la thse. T(tu) es en quelle anne de thse l ? (PFL1#09)
47 Among the contexts where on, tu and vous are interpreted as indefinite pronouns, t here are some fixed expressions requiring categorical use of either ON or TU/VOUS; discourse markers tu sais, tu vois, vous savez, vous voyez belong to this category of expressions, and were excluded from the multivariate analysis with GoldVarb. In total 113 tokens of the discourse markers were excluded. Expressions such as si tu veux, si vous voulez are frequently used with TU/VOUS, but it is also possible to say si on veut, hence they were kept in the data. Likewise, although one tends to say on verra, on va dire, on dirait, on appelle, TU has also been attested in those expressions (Laberge & Sankoff, 1980) so they were kept in the data as well. The generic ON, TU or VOUS appearing in false starts are also exclude d from coding and analysis, because the incompletion of utterance decreases the accuracy of interpretation of the context and might influence the analysis. If the same variant was repeated several times in a same utterance, only one token was counted and c oded. Examples 4J and 4K illustrate how ON and TU appeared in a false start and a repetition, respectively: ( 4J ) Et tu : fin TU peux balancer les copains mais quand TU sais pas a cest horrible. (PFL1#07) ( 4K ) TU casses les ufs tu TU fais les patates et tu et TU mlanges le tout, donc cest simple faire cest rapide. (PFL1#09) Finally, the variants ON, TU/VOUS used in reported discourse were not included in our data analysis either, because in this case, the generic pronouns were not really used by the speakers themselves, but were used by other people and were reported by the speakers. Four tokens of the variable were found in reported discourse, and they consisted of all four tokens of the VOUS variant found in the data. 4L is an example of this; the speaker was reporting the discourse from a commercial :
48 ( 4L ) (describing a commercial on bottled water) quand VOUS achetez cette bouteille VOUS aussi VOUS avez une femme nue qui vous rentre dans votre bouteille (PFL1#02) In sum, the tokens of on, tu and vous as definit e pronouns or address pronouns, the tokens of tu used in the discourse markers tu sais, tu vois, vous savez, vous voyez, and the tokens of ON, TU and VOUS appearing in false starts, repetitions and reported discourse are excluded from our data coding and analysis. Excluding these contex ts, 655 tokens of the ON TU variable were identified, extracted from the transcription files, and then copied and pasted to Excel worksheets for coding. Coding After identifying the tokens of the ON TU/VOUS variable under investigation, the linguistic factors and social factors of each token were analyzed and coded. Specifically, five factor groups and 29 factors were coded. A factor group refers to a variable, either dependent or independent. For GoldVarb, the first factor group represents the dependent variable, and the following ones represent the independent variables. Within each factor group, there could be multiple factors, potentially affecting the choice of variant. Table 4 3 displays the coding protocol of the prese nt study.
49 Table 4 3. Coding protocol Factor group 1(Variant) 2(Verb) 3(Syntax) 4(Pragmatics) 5(sex) Factor 1(On) a(appeler) p(presentative) s(situational insertion) F 2(Tu) d(dire) s(si clause) e(subjective evaluation/comments) M 3(Vous) v(voir) q( hypothetical quand clause) d(objective description or explanation/usual facts/ common knowledge) l(vouloir) i(other implicative clauses) c(combination of more than one pragmatic factor) s(savoir) o(subordination other than s, q, i) a(ambiguity) r(avoir) e(existential expressions, equivalent to il y a ) x(fixed expression, doesn't apply) p(prendre) j(coordinated and juxtaposed sentences without conjonctions that do not involve implication and existential expressions) o(other verbs) c(combin ation of more than one syntactic factor) x(fixed expression, doesn't apply)
50 Presentative clauses refer to expressions that tend to announce a general truth or a personal opinion that the speaker hopes to be shared and admitted, such as j e pens e que, je crois que, mon avis, daprs moi, il faut que, il est certain que, on voit que, on peut dire que, etc., as in Example 4M: ( 4M ) Evidemment on videmment ils ont pas rpondu, ce que je pense quON a pas le droit de ramener son assiette et ses couve rts. (P F L1#02) In these circumstances, one might tend to sound formal and this leads us to hypothesize the dominant use of the conservative variant ON. Implicative constructions refer to clauses involving a cause and a consequence, expressing supposition. An implicative construction can be headed by si quand, une fois que etc., which normally introduce a cause and an effect, for example : ( 4N ) C est si ON habite dans une grande ville ON tout est accessible.(PFL1#10) ( 4O ) Q uand TU la regardes, TU meurs 7 jours aprs.(PFL1#01) It may also be used in sentences where no explicit relative pronouns are used, but where supposition is still involved, as in Example 4P: ( 4P ) et puis plus TU fais des choses plus T (TU) as de la chance de rencontrer des gens (PFL1#07) The use of p lus in this example implies a supposition, the sentence can be reformulated using si, which clearly denotes a supposition, as in 4P#1: ( 4P ) # et puis si TU fais plus de choses, T (TU) as plus de chance de rencontrer des gens (PFL1#07) Implicative constructions other than those with a si clause or a quand clause are grouped into the factor coded by i. According to previous studies reviewed, the hypothetical nature of the 1 The # denotes an example that is created to illustrate a certain point.
51 implicative constructions reduces the ambiguity of TU/VOUS, and leads one to hypothesiz e that the generic TU/VOUS would be dominant in this context. One must presume that there may be other constructions where the generic subject cl itics may appear, which is the case in the present study, for instance, a relative clause with o, as illustr ated in Example 4 Q: ( 4Q ) a est un quartier o ON trouve qui est un peu normal comme nous sommes Paris (PFL1#01) For those clauses without presentative heads and implicative constructionsand which involve subordination, I categorize them into subordina tion. In addition, I have created a final syntactic frame, which I label juxtaposition and coordination. This frame refers to a simple sentence where only one verb phrase is present, as well as to juxtaposed or coordinated sentences where several indepe ndent sentences are linked by commas (juxtaposition) or by conjunctions mais, ou, et donc, or, ni, car (coordination).Examples 4R and 4 S illustrate this category: ( 4R ) (Juxtaposition) il y a des lyces payants, ON doit payer pour aller dans un bon lyce ( PFL1#06) ( 4S ) (Coordination) dans ce quart ier l il y en a srement mais ON sait pas comment quils sont en fait (PFL1#10) Readers cannot assume that the juxtaposition and coordination I label in the present study refers exactly to the generalization used by Laberge (1977) or the paratactic structures used by Ashby (1992), because Laberges (1977) definition of generalization was based on coreferential lexical and syntactic indices of generalization, instead of the syntactic structure where the variabl e appears. While the term paratactic structures means a series of phrases without the use of connecting words according to a dictionary (Neufeldt & D. B. Guralnik 1995) Ashby (1992) did not operati onalize and exemplify this definition. Within my label, juxtaposition and coordination, I group syntactic structures that do not fit into the other syntactic frames.
52 When more than one syntactic frame co occurs in one sentence where the variable appears, coding priority is given to the syntactic frame that is considered to be more salient than the others. The pragmatic nature of the utterance can help to interpret the salience of the syntactic frame, for example, in the case when je pense que and si cooccur, if the pragmatic characteristic of the utterance is evaluation, which means that the speaker is stating a moral or is evaluating somebody or something, the syntactic frame of je pense que is considered more salient than si. E xample 4T illus trates such a case: ( 4T ) je pense que ce ( ) cest si ON habite dans une grande ville on : tout est accessible tout accessible.(PFL1#10) Since S peaker #10 was evaluating the advantage of living in a big city, the effect of je pense que was analyzed as overri ding that of si, so presentative was assigned as the syntactic factor of this token instead of si. Three pragmatic categories are coded, namely, situational insertion, evaluation, and description/explanation. Situational insertion in the present st udy refers to the evocation of a hypothetical concrete situation. It often happens 1) when speakers are narrating a story, they are evoking the scene or situation of the story, or are inserting the psychological activities of the people in the story; or, 2 ) when speakers are giving an opinion or a comment, and then exemplify the argument, supporting the opinion. For example, ( 4U ) 1 : Quel est le nom ? 2 : THE RING. 1 :
53 ( 4V ) mais en fait jaime en fait jaime bien faire la cuisine, je sais pas cest relaxant, puis en plus ON se dit quON va bien manger aprs et ON se dit quON fait quelque chose pour des gens quON aime et tout a donc donc cest bien (PFL1#04) ( 4W ) 1 Oui, jai des amis qui disent que le voyage cest un bon moyen de rencontrer des gens ( ?) 2 Oui cest s r, fin aussi, si TU [4 W 1] fais un voyage deux ou trois mme mme l cest bien, mais si TU [4 W 2] fais un voyage en grand groupe, fin certainement ya tellement de chances que TU [4 W 3] restes dans ton groupe que TU [4W 4] rencontreras personne, mai s si TU [4 W 5] pars tout seul pour un certain temps dans un pays TU [4W 6] peux pas rester cloitr chez toi pendant trois mois, cest pas drle et mme partir du moment o TU [4W 7] vas aller faire fin cest vrai cest plus le cas fin cest vraiment le cas en France mais partir du cas TU [4 W 8] vas tous les jours dans un magasin, les gens ils vont te connaitre ( ?) mais sinon si TU [4 W 9] vas faire quoique ce soit mais de toute faon TU [4 W 10] es oblig dtre confront des gens de partout cest les alas de la vie, et cest vrai que cest, quand TU [4 W 11] voyages T(TU) [4W 12] as, cest plus inhabituel et donc cest plus tes sens sont plus en veil T(TU) [4 W 13] est plus oblig dtre actif ON [4w14] va dire donc je pense que TU [4W 15] rencontres beaucoup plus de gens (PFL1#07) In Example 4U, S peaker # 01 was telling the storyline in the film The Ring, and was evoking a scene where once one looks at the tape, one will die in seven days. In Example 4V, the speaker #04 was stating th e psychological activity of anybody who is cooking, based on her own thought. In Example 4W, from token [4W 1] to token [4W 10], Speaker 07 was exemplifying and elaborating his opinion Oui cest sr (que le voyage cest un bon moyen de rencontrer des gens ), by evoking situations of going abroad alone or in group, and of going to a store. The common point of these three examples is that the speakers were evoking a hypothetical yet concrete situation, which has an effect of increasing the vividness of th e discourse, engaging themselves as well as their interlocutor in the situation and hence increasing the degree of the interactive nature of a communication. Since the ON variant sounds more distant and conservative, it might not be as ideal as TU for maki ng the conversation more interactive. Conversely, since the TU variant shares the same surface form as the tu/vous as address
54 pronouns, one may think that if the speaker intends to engage his/her listeners into his/her narration or argumentation, he/she ma y take advantage of the ambiguity between the two meanings of tu. In comparison to ON, TU is of course closer to the implication of the tu denoting interlocutor. Therefore, I hypothesize that TU will be dominant in this pragmatic category (situational insertion). Again, however, one must not assume that the situational insertion that I apply in the present study is exactly the same as the one used by Laberge (1977) and Ashby (1992). As mentioned in the literature review chapter, the two studies do not converge on the definition of situational insertion, from which one can see that the interpretation of pragmatics of discourse varies from one investigator to another. Even though one attempts to use one of those definitions rigorously, one can still fi nd it difficult to apply to every token; for example, Ashby (1992) stated that situational insertion referred to generalizations based on personal experience and exposition referred to generalizations based on others experience or common knowledge. (P 151) However, it is difficult to know whether a generalization is based on the speaker himself/herself or some other, if the context does not give enough information. While coding, I found it very difficult to apply the definitions to my data, perhaps because it is always difficult to read the speakers minds and know what they really intend to express. In addition, sometimes my interpretation does not always match the definitions given by the previous studies. I finally decided to define and code some categories based on the interpretation of the previous studies, and define the others according to my own interpretation. Specifically, evaluation in the present study corresponds to that of Ashby (1992); it refers to the formulation of both morals and personal opinions. Tokens [4W 11] through [4W 15] illustrate evaluation. After exemplifying his point, the speaker comes to a conclusion by
55 evaluating the benefits of traveling. ON is assumed to be predominantly in this category because of the metaling uistic or formal nature of the discourse. When judging or evaluating, one tends to adopt a relatively formal sounding speech style. Unlike the previous studies, I added a new pragmatic category, description/explanation, which refers to objective description, presentation or narration of a person, a fact or a thing, and explanatory discourse. For example: ( 4X ) 1 : Le TOEIC ? 2 : Le TOEIC, ouai. 1 :
56 carrying out variable rule analysis and associated data manipulations and displays. The application has been successfully tested by numerous researchers. To run Gol dVarb, one must create or import a token file, a condition file and a cell file. Multivariate analysis is then performed and results can be displayed (D. Sankoff et al., 2005; Tagliamonte, 2006) In this chapter on the methodology of the present study, I have discussed the participants, my data collection procedures, my data collection techniques, the transcription process, and the data encoding process. Results will be discussed in the next chapter.
57 CHAPTER 5 RESULTS AND DISCUSSI ON ON FRENCH L1 DATA In this ch a pter concerni ng the resul ts and discussion of the coded L1 French data, an overview of the distribution of the variants of the variable ON TU/VOUS will be discussed, based on the raw data. Afterward, the results of the multivariante analysis of linguistic and extra linguistic factors generated by GoldVarb will be displayed and discussed. The results will be compared to those found in the previous studies. The research questions and hypotheses will also be addressed. Overview Before inputting the raw data and performing a multivaria te analysis with GoldVarb, an overview of the raw data on Excel worksheets allows one to observe some general patterns that are interesting to explore in more depth. As mentioned in the section on data encoding, we found 772 tokens of the variable ON TU/VOUS, of which 113 are in the discourse markers tu sais, tu vois and four appeared in reported discourse. The speakers use ON and TU as generic pronouns most of time. Only four tokens of the generic VOUS are found in the data and they all appeared in repor ted discourse, which was not really produced by the speakers themselves. Therefore, the four tokens of VOUS, as well as the discourse markers tu sais, tu vois that do not allow alternation with ON, were not included in the quantitative data analysis. Aft er this exclusion, Table 51 and Figure 5 1 displays the distribution of ON and TU in the raw data. Table 5 1. The distribution of ON and TU in the whole data Group Number of ON Percentage of ON(%) Number of TU Percentage of TU (%) Overall 362 55.3 293 44 .7
58 Figure 51. The distribution of ON and TU in the whole data In the data that was imported into GoldVarb, there are 362 tokens of ON (55.3%) and 293 tokens of TU (44.7%). The frequency of the second person variant is quite close to the 47.7% (TU/VOUS) found in Coveney (2003), and to the 42.6% (almost all VOUS) calculated by Ashby in his Tours corpus. It seems that overall, the frequency of second person pronouns remains stable from 1976 when Ashby constructed his Tours corpus to the present study. Ho wever, since I have only young middle class participants in my data, I will compare the percentage of ON or TU in my data with that of only the young middle class counterparts in the previous studies and see if different results will be revealed. These res ults concerning the linguistic change of the variable ON TU/VOUS will be discussed in more depth in the other subsection in this chapter. Variation is displayed both interpersonally and intra personally. Some participants alternated between ON and TU in th eir speech, while others used ON categorically. Some used ON predominantly while some preferred TU. As illustrated in Table 5 2, the distribution of ON and TU by sex shows that the female participants strongly preferred ON to TU, as opposed to their male counterparts, who used TU predominantly.
59 Table 5 2. The distribution of ON and TU by sex Group Number of ON Number of TU Percentage of ON(%) Females 238 11 95.6 Males 124 282 30.5 Figure 52. The distribution of ON and TU by sex In addition to sex, it appears that style might come into play with the choice of ON and TU, based on the overall impression of the data. That is, the participants who overwhelmingly used TU also produced considerably more discourse markers (tu sais, tu vois) than those w ho used ON most of the time. If the discourse markers reflect an informal speech style (Beeching, 2002; Bri ton, 1996) and the use of the generic TU correlates with that of the discourse markers, one might suggest that the generic TU is very likely to be an informal marker. More discussion concerning the stylistic variation will be provided later in the chapter. Finally, concerning the correspondence between the use of the address pronoun tu/vous and the use of generic pronouns TU and VOUS, although some participants used vous to address me and others used vous at the beginning of the conversation and switched t o tu shortly there
60 after, TU was the only form they used to express generic reference. This will also be discussed in this chapter. Real time comparison across corpora As mentioned earlier in the section proving an overview of my participants, I have only young middle class participants in the present study, but in the previous studies, the overall frequency was calculated based on participants from different age groups and social classes. So it is more meaningful that we compare the use of the ON TU/VOUS b y the participants in the present study to that of their counterparts in the previous studies. In order to conduct the comparison, I calculated the frequency of the generic ON by the young middle class participants in the two previous studies myself, based on the information provided in the tables or appendices of these studies (Ashby, 1992; Coveney, 2003) Figure 5 3 shows the frequencies of ON by only middle class p articipants below age 25 in Ashbys Tours corpus, Coveneys Picardy corpus and my Paris corpus. Figure 53. Frequency of generic ON by young middle class participants in three corpora From the table, one can see that the percentage of ON drops from higher than 70% attested in the previous studies to about 55% in the present study. Since the conversational interview is
61 the common speech context in all three corpora, I posit that the use of the generic ON might be in regression, giving place to TU/VOUS in spoken France French. One might posit that my suggestion would be more convincing if the three corpora were constructed based on participants from a same speech community and if I had older participants to make a prediction of linguistic change in apparent time (Meyerhoff, 2006) Nevertheless, while phonologica l variables are often geographically marked, grammatical variables display much less variation from region to region (Coveney, 1997) From this point of view, it is reasonable and meaningful to make a comparison among the three corpora and the results of this real time comparison do provide interesting information concerning the possible linguistic change of the variable ON TU/VOUS. If there was indeed a linguisti c change going on with the variable ON TU/VOUS in France French, one must wonder whether the men would be leading the change, as found in Montreal French. A comparison of the frequency of ON between the two sexes in the three corpora reveals a sex differen ce regarding the use of the variable, as illustrated in Table 5 6 and Figure 56. Table 5 3. Sex difference in the frequency of ON by young middle class participants in three corpora %ON 1976 Tours 1990 Picardy 2008 Paris Men 70 68 30.5 Women 79 79 95.6
62 Figure 54. Sex difference in the frequency of ON by young middle class participants in three corpus The comparison reveals a striking gender difference, in that while women increase their use of ON from the previous corpora to the present corpus, the relative frequency of ON significantly drops in mens speech (from around 70% to 30%). Thus, another hypothesis that can be made is that, as in Montreal French, men are the leaders of the ongoing linguistic change involving the ON TU/VOUS variable i n France French, and this change has not yet been adopted by other social groups. It is probably not surprising to find that men are the leaders of the change for this variable because the use of TU is associated with covert prestige, which is often valued by men. Concerning the motives of the linguistic change of the variable ON TU, in Montreal French, there has been an ongoing debate about whether the change is internally motivated or due to contact with English where you is used for generic reference (Blondeau, 2008; Chaudenson, Mougeon, & Beniak, 1993; Mougeon & Beniak, 1991; Thibault, 1991) While the contact situation in Quebec might lead one to postulate the possibility of a contact induced change, which is often accompanied by the processes of simplification and regularization, one might question the applicability of this hypothesis in France French, which has much less contact
63 with English except the general effect of globalization. The fact that there might be a linguistic change going on with the variable ON TU in France French supports the hypothesis that the change in Montreal French is very likely to be associated with the inherent variability of the language itself. In fact, according to Thibault (1991), this change is a process of restructuring of pronominal subsystems: with the first person singular pronoun on taking the place of nous the form on is gaining more functions, including those of personal pronoun and generic pronoun, therefore, in order to release some of the burden on on, the generic TU is chosen by speakers to substitute for the generic ON. In this way, the pronominal subsystems are changing and restructuring. This argument from the perspective of language restructuring provides an enlightening interpretation concerning the mechanis ms of linguistic change; however, it is interesting to go further and examine the cause of the linguistic change in France French. In the previous subsections of the present study, I have suggested that the use of TU might be associated with interaction, i nformality, and solidarity or camaraderie. The fact that men from lower social class background exhibit a higher ratio of TU/VOUS (Ashby, 1992) adds support to the argument, because informality and solidarity or camaraderie are typical characteristics of this social group. So, if over time, the linguistic behavior of this social group is adopted by other social groups, it will suggest that in a given situation, such as informal conversation between two people who do not know each other very well, people are interacting in a way which is more and more informal and interactive. And this change in interpersonal interaction could, inversely, infl uence the choice of ON TU, namely, people use TU in progression over time. Another piece of evidence that supports this argument is that in Montreal French, scholars find an increase in the use of the discourse marker tu sais / tu vois, along with that of the generic pronoun TU (Thibault, 1991)
64 Recall that the use of the discourse marker tu sais / tu vois denotes interaction and informality, thus the progression of the discourse marker in real time m ight reflect the evolution of interpersonal interaction, as well as the generic pronoun TU; in other words, I argue that the increasing use of the generic TU in real time is due to the change of pragmatic configurations in an informal interview situation. In sum, by comparing the distribution of ON by young middle class participants in the three corpora discussed above, I have attempted to suggest that, at least for men, there might be a linguistic change going on with the variable ON TU/VOUS; in other word s, men use TU more and more frequently in the place of ON over time. However, further strict replication studies should be conducted to verify the suggestion and shed light on the situation in real time. Concerning the motive of linguistic change involving ONTU, I suggested that it might be due to the change of interpersonal interaction or pragmatics over time. Linguistic variation After an overview of the raw data, a token file was created in GoldVarb and after generating a condition file and a cell file, the results concerning the distribution of ON and TU by each factor group (verb, syntax, and pragmatics) were displayed. After this step, the verbs appeler and prendre were eliminated because of a relative lack of tokens and categorical use with ON wi th these verbs. The data were thus recoded using a recode setup function of the software before performing a variable rule analysis. The results of a binominal one level analysis and a step up and stepdown analysis with the software show that all the factor groups in the present study are significant. In this subsection, the results concerning the constraints of verb, syntax and pragmatics will be shown and discussed.
65 Verbs The verbs initially examined include appeler, dire, voir, vouloir, savoir, avoi r, and prendre. Recall that in previous studies, appeler and dire are shown to favor the use of ON and vouloir the use of TU. The effect of the other verbs had not been examined. My hypothesis was that the results in the present study would confirm the effect of appeler, dire and vouloir on the use of the generic ON and TU. In addition, I posited that the verb voir would favor the use of ON, considering the frequent use of voir in the fixed expressions with ON such as on verra and on va voir; and that the use of the verb avoir would favor the use of TU for the same reason. Table 5 4. The count, frequency and factor weight of ON and TU according to different verbs Verbs # of ON % ON #of TU % TU Factor weight 1 (application value: ON) Dir e 40 95.2 2 4.8 0.93 Voir 28 71.8 11 28.2 0.69 Savoir 5 71.4 2 28.6 0.57 Avoir 42 36.2 74 63.8 0.36 Vouloir 7 30.4 16 69.6 0.32 appeler 10 100 0 0 N/A prendre 1 100 0 0 N/A Others 229 54.9 188 45.1 0.47 As shown in Table 54, dire and voir str ongly favor the use of ON. One reason is that in French, there exists a number of frozen expressions with these two verbs, such as on va dire, on dirait, on dit que, on dit, on verra, on va voir, on voit, on voit que, as illustrated in e xamples 5 A and 5B: ( 5A ) Honnt e ment je, mme mme ON va dire dans la cuisine franaise des : des choses assez simples je : je sais pas bien cuisiner du tout, du tout, du tout. (PFL1#01) ( 5B ) J espre vraiment que je vais pouvoir faire ce que jaime donc ON verra b ien (PFL1#5) 1 If the factor weight index is higher than 0.5, the factor is considered to favor the use of ON, and if the factor weight index is lower than 0.5, TU is used more frequently than ON under the effect of this factor.
66 It is very rare for French people to substitute ON with TU in these fixed expressions. TU was never used in these expressions in my data, and the participants who used TU predominantly nevertheless opted for ON when they used these expressions For example, speaker # 09 is a frequent user of the variant TU; however, he switched to ON whenever the fixed expressions with ON were involved, as illustrated in the following extract: Parce que en fait il y en a encore certains qui sont encore en DEA p arce qu e ils, vu quils ont pass a il y a trs longtemps ils savent pas trop ce que cest le Master 2me anne donc ON d it plutt DEA pour ceux qui savent pas trop (PFL1# 09) If there are fixed expressions with ON, one can imagine that there must exist s ome frozen forms with TU as well, and the verbs concerned in these contexts might favor the use of TU. This proved to be true with our data: the verb vouloir, which is used very frequently in expressions such as si tu veux and comme tu veux, favors t he use of TU. Even in regular expressions, the participants tend to choose TU when the verb vouloir is present. Example 5C shows the use of a fixed expression with TU and vouloir, and example 5D illustrates the use of TU with vouloir in regular spe ech. ( 5C ) en fait : en fait si TU veux, il y a d autres entreprises internet (PFL1#06) ( 5D ) cest toi qui choisis ton doma ine, en fonction de : de ce que TU veux faire (PFL1#09) It is interesting to note that the only token of TU that Participant #04 produced appea red with vouloir in regular speech. This adds an additional support to our suggestion. The influence of the verb avoir on the choice of ON and TU/VOUS has never been examined in the literature. In the present study, the results revealed that avoir wa s very frequently used with TU, although ON was also used by the participants in the same context. Most often, avoir appeared with TU when the participant was introducing a new referent or was describing a thing, and it normally appeared in existential c ompletive constructions. In this case, TU is equivalent to il y a (Ashby, 1992) for example:
67 ( 5E ) (describing the procedure of taking an art exam) T(TU) as des sances de nue et les personnes changent de position toutes les deux minutes (PFL1#02) The meaning of the sentence woul d not be changed if one replaced TU as with il y a, as shown in [5e]#: ( 5E ) # (describing the procedure of taking an art exam) Il y a des sances de nue et les personnes changent de position toutes les deux minutes (PFL1#02) In addition to the frequent use of TU in this type of construction, another interesting pattern observed is that TU was often reduced to T and linked to as. This supports the idea that informal variants tend to co occur in informal speech style (Ervin Tripp, 1972) Finally, concerning the verb savoir, I postulated that since savoir is often used in the expression on sait que, it would favor the use of ON in general speech. According to the results, although the tendency of choosing ON or TU is not very strong with the verb savoir, it does go slightly in the direction of ON. The weak tendency might be due to that fact that, in comparison to dire, appeler, voir, avoir, and vouloir, few fixed expressions involve savoir. In fact, the hierarchy of the factor weights of the verbs shows that while dire and voir strongly favor the use of ON and avoirand vouloir favor the generic TU, the verb savoir is in the middle of the hierarchy. This might suggest that the frequent use of fixed expressions does have an effect on the association of the subject pronoun and its verb. The results of the present study concerning the constraint of pronoun choice according to verb confirmed the findings of the previous studies that dire favors the use of ON and vouloir favors the use of TU (Laberge, 1977; Laberge & Sankoff, 1980) In addition, the results contribute to the previous studies by demonstrating that voir is frequently used with ON and avoir with TU.
68 Syntax The second factor group concerns the effect of syntax on the choice of ON or TU. I hypothesized earlier that presentative constructions would lead to frequent use of the generic ON, while implicative constructions like si clauses and quand clau ses would favor the use of TU. Table 5 5 displays the count and frequency of ON and TU according to syntactic factors, and the probability weight of ON in my data. Table 5 5. The count, frequency and factor weight of the syntactic factors Syntax # of ON # of TU % of ON Factor weight Presentatives 82 14 85.4 0.71 Si clause 24 29 45.3 0.68 Quand clause 52 65 44.4 0.57 Juxtaposition 82 69 54.3 0.45 Existential 29 51 36.2 0.38 From the table, one sees that the participants are highly likely to use ON in presentatives, as the factor weight of 0.71 shows. The following examples show the use of ON in different presentative constructions found in the data, namely il faut que cest que, cest vrai que, je pense que, je trouve que, parce que, je me dis que, on dit que, and on se dit que : ( 5F ) Il faut pas : il fa ut pas faire quelque chose quON aime pas l (PFL1 #05) ( 5G ) C est un perpetuel volution, cest a que jaime bien, cest quON se sennuie pas. (PFL1 #06) ( 5H ) Mais bon cest vrai que plus ON gra ndit aprs chacun a ses plans cot donc (PFL1 #03) ( 5I ) Dun cot par exemple dun point de vue syntaxique ou gr ammatical ou quoi je pense quON doi t faire des fautes normes et ON sen rend pas compte (PFL1 #03) ( 5J ) Je trouve que cest faux en part ie, parce que a dpend qui ON parle (PFL1 #02) ( 5K ) On grandit moins vite parce quON a pas besoin de faire deffort (PFL1 #02) ( 5L ) Moi je me dis quON peut vraiment trouver quelque chose ou chercher quelque chose (PFL1 #03)
69 ( 5M ) Plus ON va, ON dit quON se sent tout petit l bas tellement cest grand les immeubles (PFL1 #04) ( 5N ) Puis en plus ON se dit quON va bien manger aprs (PFL1 #04) Conversely, the existential structures, normally with the verb avoir, favor the use of TU, although both ON and TU can be found in this structure, for example : ( 5O ) ON a la Seine qui coule juste aux pieds de la de luniversit (PFL1 #05) ( 5P ) Cest la France, l T (TU)as lAngleterre et l T(TU)as la Sude (PFL1 #07) Si clauses and quand clauses, which belong to the implicative constructions, favored the use of ON in our data, which goes against our hypothesis. However, the percentages of ON in these two categories indicate that the participants produced TU more frequently than ON. The mismatch between the percentages and the factor weights might be due to the fact that the data is not very equally distributed in each cell, as a distributional analysis of Goldvarb indicates; in other words, a cross tabulation analysis using the software shows that there are large amounts of data in some cells and no data at all in the others (cf. Appendix D). Nevertheless, the percentages of 45.3% and 44.4%, respectively, do not seem to be strong evidence favoring our hypothesis that TU is significantly preferred in si clauses and quand clauses. The factor of juxtaposition and coordination only very slightly favors the choice of TU, as the weight index of 0.45 shows. The neutrality of this syntactic category explains the near equal effect of the two variants. Pragmatics In the methodology section, I hypothesized that the pa rticipants would prefer TU in situational insertion because this can create an effect of interaction, solidarity and informality. Conversely, I posited that the use of ON in evaluation would be dominant because people tend to be formal and conservative when evaluating. Finally, I hypothesized that the pragmatic
70 category objective description and explanation would situate in between the first two categories, because of its neutrality. As expected, pragmatic factors were also found to influence the use of the ON TU variable. Table 5 6 gives the information concerning the distribution of ON and TU according to the pragmatic factors examined, namely, situational insertion, evaluation, and objective description/explanation. Table 5 6. The count, frequency and factor weight of ON and TU according to different pragmatic factors Pragmatics # ON % ON # TU % TU Factor weight (application value: ON) Evaluation 144 69.6 63 30.4 0.69 Description/Explanation 95 50.5 93 49.5 0.57 Situational insertion 51 29. 8 120 70.2 0.13 As the factor weight of 0.13 shows, the participants have a strong tendency to use TU in situational insertion. Conversely, when they are making evaluations of a person or a thing, ON is their choice in most cases, as the factor weight of 0.69 shows. As for the objective description and explanation, the factor weight indicates that ON is only slightly preferred to TU, but that, as predicted, the choice between these two pronouns is roughly equivalent, with neither preferred overwhelmingl y. So, in general, our hypotheses concerning the choice of the variants under the effect of different pragmatic factors considered are confirmed. When the participants were evoking a hypothetical situation, supporting their opinions and arguments by giving a concrete example, or describing the psychological activities of a person, they tended to choose TU. Example 4W, provided in the subsection on data encoding, illustrates situational insertion: ( 4X ) 1 Oui, jai des amis qui disent que le voyage cest un bon moyen de rencontrer des gens ( ?) 2 Oui cest sr, fin aussi, si TU [4 W 1] fais un voyage deux ou trois mme mme l cest bien, mais si TU [4 W 2] fais un voyage en grand groupe, fin certainement ya
71 tellement de chances que TU [4 W 3] restes dans to n groupe que TU [4W 4] rencontreras personne, mais si TU [4 W 5] pars tout seul pour un certain temps dans un pays TU [4W 6] peux pas rester cloitr chez toi pendant trois mois, cest pas drle et mme partir du moment o TU [4W 7] vas aller faire fin cest vrai cest plus le cas fin cest vraiment le cas en France mais partir du cas TU [4 W 8] vas tous les jours dans un magasin, les gens ils vont te connaitre ( ?) mais sinon si TU [4 W 9] vas faire quoique ce soit mais de toute faon TU [4 W 10] es oblig dtre confront des gens de partout cest les alas de la vie, et cest vrai que cest, quand TU [4 W 11] voyages T(TU) [4W 12] as, cest plus inhabituel et donc cest plus tes sens sont plus en veil T(TU) [4 W 13] est plus oblig dtre acti f ON [4w14] va dire donc je pense que TU [4W 15] rencontres beaucoup plus de gens (PFL1#07) Concerning the interpretation of the choice of TU in this pragmatic context, the previous studies suggested that the hypothetical nature of the context decreases the ambiguity between the pronoun of address TU and the generic TU, and hence favors the use of TU (Laberge & Sankoff, 1980) However, one might wonder whether the same thing might also happen with ON, since the hypothetic si tuational insertion could also distinguish the definite pronoun ON from the generic pronoun ON, and ask why ON is not favored in this context? By defining and coding situational insertion differently from the previous studies, I have attempted to provide a different interpretation of the preference for TU in this context. Given the confirmation by the results from the present study, I suggest that, as mentioned earlier in the section on data coding, when the participants were inserting a situation, they i ntended to engage the interlocutor in their narration, elaboration, or argumentation. The use of TU creates an effect of interaction, since it is potentially tinged with its original meaning of TU as an address pronoun in France French. Another interpretat ion would be that when the participants are really involved in the situations they are evoking, it is possible that their interests and emotions override the atmosphere of an interview and an informal style is exhibited, if we can consider TU as an informa l marker. The question of style will be revisited later in this chapter. Our results concerning the effect of evaluations corroborate those found in the previous studies. When the participants were evaluating, judging, or formulating a moral or truism, the y
72 intended to be conservative or formal, and thus adopted the traditional variant ON. For example, one participant, # 06, who used TU much more frequently than ON in his speech, nevertheless switched to ON when he was stating morals, as illustrated in 5 Q and 5R: ( 5Q ) quand ON veut sortir de l ON sort (PFL1#06) ( 5R ) quand ON veut ON peut (PFL1#06) As for the pragmatic factor description/explanation, the factor weight of 0.57 indicates that in general, the use of ON patterns with that of TU, although ON is very sl ightly preferred in this context. This result confirmed our hypothesis that ON and TU would be relatively equally weighted in this context due to the neutrality of the discourse. The neutrality of the discourse in this context can be seen through a compari son with the other pragmatic factors situational insertion and evaluations. When the participants were giving information on something, explaining how to prepare a dish, or explaining the procedure of something, they were not as involved as in situati onal insertion in which the situation was linked with subjectivity. Similarly, they didnt intend to be as distant and formal sounding as in evaluations. The following extracts 5 S and 5 T demonstrate the use of ON or TU in objective description and ex planation: ( 5S ) Be n ON ramasse le raisin toute la journe, donc ON est l ON ramasse les petits les petits : les petits raisins, les grappes de raisins et puis ON les met dans les dans les grands paniers sur le camion et puis voil ON f ait a toute la journe donc il y en a qui prennent des gros coups de soleils mchants et puis ON est debout quoi toute la journe, fin cest dur, moi je suis debout aussi toute la journe PIMKY dans mon boulot mais mais je parle avec des gens cest ON coute de la musique ON se dtend de temps en temps l cest NON STOP (PFL1 #05) ( 5T ) cest simple faire q uand TU prends les aubergines TU les comment dire TU les vides et aprs TU fais comment ON appelle a ? TU fais pas bouillir mais en fait TU fais : TU fais chauffer quoi aprs TU : TU boues quoi le tout TU mets a en bouilli et aprs ben TU mlanges avec de lhuile et et TU manges (PFL1 #09)
73 If one compares th e examples provided above with 4W and 5Q, R one can notice the difference between the three pragmatic categories. The participants obviously do not equally engage themselves as well as their interlocutor in the examples 4 W, 5 S, T, and 5Q, R. If a continuum can be made to measure the informalityformality, or the interaction distance of the speech of the participants, situational insertion would be situated at the left part (informality/ intimacy/ high interaction) of the continuum, evaluations would be at the right part (formality/ distance/ low interaction), and objective description/explanation would be positio ned in the middle. Note that within each pole, there exists a range of degrees of interaction and distance. Not all the tokens convey the same degree of interaction and distance within each pragmatic factor. The continuum can be visualized in Figure 55: Figure 55. Continuum on the degree of formality and interaction of the pragmatic factors As the results show, TU is used less and less from the left side of the continuum to the right side, and conversely ON is used more and more from the left si de to the right side. This pattern supports, to a certain extent, our hypothesis that TU is very possibly associated with interaction and informality and ON is considered as distant and formal, although the topics of discourse or the correlation between the use of ON TU/VOUS and the use of other stylistic markers need to be examined in order to verify the stylistic nature of ON and TU. In sum, the linguistic factors examined in the present study confirmed in general the findings of the previous studies. In addition, several results that were revealed in the present study contribute to the existing literature: the verb voir favors the use of ON and the verb avoir favors the variant TU; the participants chose TU when they intended, maybe Situational insertion Description/Expla nation Evaluation High interaction/ I ntimacy/ Informality Low interaction/ Distance/formality
74 unconsciously, to engage themselves as well as their interlocutor into their discussion; the neutrality of objective description and explanation slightly favor the use of ON, but in general, the distribution of ON and TU is quite balanced in this pragmatic category. Gender variation Variation studies are not only interested in examining the constraints of linguistic factors on the use of a certain variable, but also the extralinguistic factors such as social and stylistic factors. Interpersonal variation, including variatio n across the sexes, is the focus of this subsection, and we will turn to style after this. As observed in my corpus, the use of ON and TU greatly varies from participant to participant, and from men to women. An interesting pattern is that in general, the participants were quite consistent in choosing between ON and TU as a generic pronoun. Three participants, #3, #5 and #10, did not display variation between ON and TU, whereas participants #1 and # 6 used both ON and TU about 50% of the time respectively. Although the other six participants did display variation, it is obvious that one of the variants overrode the other in their speech, namely, either he/she used ON predominantly, or he/she used TU overwhelmingly, as shown in Table 57. Table 5 7. Use of ON and TU by each participant in PFL1 Participant # Sex N. of ON N. of TU N. of ON+TU % ON 3 F 36 0 36 100 5 F 65 0 65 100 4 F 72 1 73 98.6 2 F 57 2 59 96.6 1 F 8 8 16 50 10 M 23 0 23 100 6 M 29 30 59 49.2 7 M 43 73 116 37.1 9 M 23 114 137 16.8 8 M 6 65 71 0.08
75 In addition, from Table 55 and Table 51 on the section of overview, one can find a significant sex difference regarding the choice of the variants. The females used ON 95.6% of the time, while the men used ON in 30.5% of the generic ref erent contexts. The relative weight of the sex factor in Table 58 also indicates a strikingly strong effect of sex on the use of the variable. Table 5 8. Sex difference in the use of ON TU Sex % ON Factor weight Females 95.6 0.95 Males 30.5 0.15 The sex difference in the use of ON TU is bigger than that attested in the previous studies, namely, the percentage of ON in the speech of females in the present study is higher than that attested in the previous studies and the rate of TU in the speech of mal es lower. In Laberge (1977), the females under 40 produced ON at a rate of 75.6% in the generic contexts in their speech, and men under 40 produced ON at a rate of 28.1%; in Ashby (1992), young males produced ON 45% of the time while young females produced ON 83% of the time. Since multiple factors could influence the use of a variable at the same time, it is understandable that the percentages in different studies vary to a certain extent. Linguistic factors or extra linguistic factors other than sex might be what cause the percentage variance in the different studies. Nevertheless, it is important to note that in all the studies the factor of sex proved to be highly significant, especially in the present study. Despite the general pattern of sex difference, one must notice that the female participant #01 used TU 50% of the time and the male participant #10 used ON categorically, which didnt conform to the general gendered pattern observed in the data as a whole. This leads one to think that there must be some other effects that play a role in influencing speakers choice of ON or
76 TU. One of these effects must be the linguistic environments and conditioning examined in the section on linguistic variation; namely, the linguistic conditioning of Participant #10 might strongly favor the use of ON and that of Participant #01 makes her use ON and TU interchangeably. Support of this hypothesis comes from the fact that the participant #10s tokens of ON are found in evaluations and presentative heads most of the tim e, while Participant #01 used TU in situational insertion with quand, whereas most of her tokens of ON are found in evaluations. Before turning to the discussion of the stylistic variation, one final remark concerns the magnitude of the effect of the diffe rent factors under quantitative examination. Since the range2 of the factor weight of the different factors in the group of sex is the largest, sex is considered to have the strongest effect on the choice of the different variants, followed by verbs, pragm atics and syntax. Stylistic variation In addition to the effect of gender and linguistic conditioning, one must wonder whether style can also account for the variation observed in the data. Two methods can be used to examine the question of style: first, a s linguistic features of a same style tend to co occur (Ervin Tripp, 1972) one can examine the correlation between the variable ON TU/VOUS and other stylistic markers to see whether TU tends to co occur with some other informal variants at different linguistic levels; secondly, the effect of topics on the choice of the variants can also be examined t o check whether formal topics will trigger more ON, and informal topics will trigger more TU/VOUS (Labov, 1972a) 2 The range refers to the difference of the highest factor weight and the lowest one in a factor group.
77 Co occurrence between the generic TU and the discourse markers tu sais and tu vois Coveney (2003) looked at the cooccurrence between the use of ON TU/VOUS and the use of the negative p article NE, and examined whether participants who favored the use of TU also deleted NE frequently. The results showed that there didnt seem to be a correlation between the two variables. Although I believe that informal markers tend to co occur in spoken speech, I suspect that in contemporary spoken French, the presence or absence of NE might not be as important a stylistic marker as it used to be. In France French, Ashby (1976; 2001) reported that in his 1995 corpus, the participants only retained 18% of NE. Moreover, in Montreal French, the rate of retention was found to be 0.5% in Sankoff & Vincent (1980). With such a high deletion rate in spoken French, one must posit that while the presence of NE indeed creates an effect of formality, the omission of NE might not necessarily indicate an informal speech style. One can still exhibit a relatively formal speech style even without using NE. In other words, I suggest that the absence of NE might not be sensitive enough to serve as a stylistic marker, at least in an informal interview situation, which is the situation in which the data of the present study and that of the previous studies were collected. This might be the reason why a correlation was not found between the use of the generic ON TU and that of N E. Then the research problem of whether or not the variable ON TU is a stylistic marker remains unanswered. There are, however, two ways to re approach this question. As mentioned earlier, one can examine the correlation of the use of the generic ON TU wit h other stylistic markers; alternatively, one can examine the topics of speech. In the present section, I attempt to propose a preliminary examination of the stylistic variation of the generic ON TU. First, while excluding the discourse marker tu sais / t u vois from statistical analysis, I noticed that there appeared to be a strong association of TU with this discourse marker in the data.
78 Table 5 9. Association between the ON TU and the tu sais / tu vois by each participant Participant # Sex N. of ON N. of TU % ON N. of Tu sais / tu vois 1 F 8 8 50 0 3 F 36 0 100 0 6 M 29 30 49.2 0 10 M 23 0 100 0 2 F 57 2 96.6 1 4 F 72 1 98.6 1 5 F 65 0 100 1 8 M 6 65 0.08 10 7 M 43 73 37.1 21 9 M 23 114 16.8 77 Figure 56. Association between the O N TU and the tu sais / tu vois by each participant As illustrated in Table 5 9 and Figure 56, the participants who used ON approximately 50% of the time or more ( #1, #2, #3, #4, #5, #6, #10) also produced very few tokens of the discourse marker tu sai s / tu vois; the discourse maker was almost absent in their speech. In contrast, the other three participants (#7, #8, #9), who were the highly frequent users of TU, produced the discourse marker much more frequently than the others. This apparent associa tion between the two linguistic features leads me to postulate that the generic TU reflects a
79 spontaneous and informal speech style, in which the discourse makers tu sais, tuvois are frequent. The stylistic characteristics of the discourse markers tu sais, tu vois had been studied in the previous literature. While I found few variationnist sociolinguistic studies on tu sais / tu vois, some qualitative analysis states that the use of discourse markers is considered to be associated with informality and to be stylistically stigmatized (Beeching, 2002; Briton, 1996) In addition, it is widely recognized that discourse markers function not only as textual markers, which serve to make the discourse coherent, but also as social monitors, which play an interactive role between interlocutors, as mentioned in Erman (2001): As textual monitors, pragmatic markers are in the main focused on the text, the speaker by using them turning sometimes fragmented pieces of discourse into a coherent text. As social monitors their principal function is to negotiate the meaning and management of discourse and to ensure that the ch annel is open between the interlocutors. At the social level, a speaker can use the discourse marker to secure the comprehension of his/her interlocutor, or to seek alignment from his/her interlocutor (Briton, 1996; Erman, 2001) These functions create an effect of interaction and solidarity. Therefore if the use of TU is positively correlated with that of the discourse marker tu sais / tu vois, I suggest that the use of TU reflects an informal speech style and increases the degree of interaction between interlocutors. Since the variant TU is excluded from f ormal writing and is not considered as standard language (Coveney, 2003) it is reasonable to associate it with informal speech style. The suggestion that TU reflects the interactive characteristic of spontaneous informal conversation is further supported by two facts found earlier in the chapter. On the one hand, TU is tinged with its original meaning of address pronoun, and a speaker might take advantage of this ambiguity to include his/her interlocutor into the apparent indefinite
80 reference and thus engage the interlocutor into the discussion. On the other hand, in the subsection on linguistic variation of ON TU, it is argued that TU is preferred in situational insertion where the participants intended to engage themselves, as well as their in terlocutor, into the situations evoked. The generic TU and the pronoun of address tu In addition to the cooccurrence between the generic TU and the discourse markers tu sais, tu vois, an examination of the correspondence between the generic TU/VOUS and the pronouns of address tu /vous will also help to shed light on the stylistic nature of the generic TU. Recall that in Montreal French, researchers found that the address pronouns tu/vous did not match the generic pronouns TU/VOUS; namely, some speakers w ho use VOUS to address the interviewer nevertheless opted for TU as generic pronoun, however, no speakers who used TU as pronoun of address changed to VOUS for generic reference. However, in the Picardy corpus of Coveney (2003), he found that this mismatch did not happen. All 25 of his participants who used tu or vous as a pronoun of address were consistent in selecting the same pronoun for generic reference. In my Paris corpus, one female participant (#03) used vous to address me and three participants (# 07, #09, #10) used vous at the very beginning of the interview but switched to tu shortly afterward, when they became more at ease. However, TU and ON are the only forms that the participants used to express generic reference; as mentioned earlier, VOUS was never really in this corpus. Participant #03, who used vous to address me, selected ON as a generic pronoun and never used TU. Among the three participants who used vous as an address pronoun at the beginning and switched to tu later, one (#10) used ON a ll the time, and the other two (#07 and #09) alternated between ON and TU, but never VOUS. Note that #07 and #09 switched to the
81 address pronoun tu shortly afterward they started to use the generic pronoun TU, as shown in 5U: ( 5U ) cest des botes prives c est de la recherche prive a dpend pas de ltat et aprs T(TU) as toutes les botes genre CNRS ou CEA j e sais pas si tu connais cest cest connu quoi (PFL1 # 09) Considering that Participants # 07 and # 09 started to use the generic TU and the address pronoun tu at approximately the same time, and were consistent in using tu as both address pronoun and generic pronoun for the rest of the interview, I suggest that overall, similar to what Coveney (2003) found in his corpus, the choice of generic pronouns matches that of address pronouns in my corpus. For my participants, the generic TU/VOUS are tinged with their original meaning as address pronoun; once the participants started to use one of them, they opted for the other to maintain consistency. It is interesting to note that the consistency of the use of the generic TU and the pronoun tu seem to indicate together the informality of some participants speech style. However, this consistency could not be interpreted as cooccurrence, in the sense that whe n one feature occurs, the occurrence of the other one is predictable, because in my data, some participants who addressed me as tu nevertheless chose the generic ON. Topic In addition to the examination of the cooccurrence of TU and other stylistic marker s, the topics of conversation also need to be examined in order to verify the stylistic nature of the generic TU in a more formal and traditional way. In this preliminary analysis of the stylistic variation of the generic ON TU, instead of carrying out a r igorous statistic analysis, I am more interested in an informal quantitative analysis and a qualitative discussion. The analysis is based only on Participants #7, #8 and #9, who produced a relatively high amount of the generic TU and who also produced the most tokens of the discourse marker tu sais / tu vois. Since these three
82 participants are frequent users of the generic TU and produced relatively few tokens of the generic ON, one might posit that if ON TU is stylistically marked, the tokens of the gene ric ON would appear more frequently in formal topics. To test this hypothesis preliminarily, I extracted the tokens of the generic ON appearing in these three participants speech samples and coded the topic in each token. Topics that are more formal include the economy, language, culture and philosophy of life, and informal topics cover travel and games. Finally, I consider the topic school, work, neighborhood as intermediary between the formal and the informal topics. A preliminary quantitative analysis in Table 5 10 and Figure 57 shows that the formal topics outweigh the informal ones, which gives a positive sign that a rigorous statistical analysis might prove our argument that the use of ON TU is stylistically marked. However, this is only a partial and informal analysis; one needs to look at the whole body of the data and carry out a statistical analysis in order to reach a final conclusion. Table 5 10. The distribution of topics in the tokens of ON among Participants #7,#8, and #9 Topics Ab solute number of tokens of ON Economy, language, culture, philosophy 19 School, work 12 Travel, game 6 Figure 57. The distribution of topics in the tokens of ON among Participants #7,#8, and #9
83 In summary, in this section concerning the socio styl istic variation of the ON TU variable, I found a significant sex difference in the use of the ON TU, namely, the variant TU is strongly preferred by the young men and ON by the young women. Additionally, positive evidence supports our hypotheses that ON TU is a stylistic marker because the use of TU is associated with the discourse marker tu sais / tu vois which is considered informal; and when Participants #7, #8, #9 used ON, they were talking about formal topics more often than informal ones. However, f urther statistical analysis is needed in order to test the hypotheses concerning the stylistic variation of the generic ON TU. Concluding remarks In this chapter, I have presented and discussed the results on the data of the French native participants. Asp ects that are covered include linguistic variation, socio stylistic variation, linguistic change and the association of the generic TU and the definite tu Verb, syntax, pragmatics, gender and style are found to influence the choice of ON or TU, among whic h the effect of the gender overrides the other factor groups, namely, verbs, pragmatics and syntax. It is also shown that the use of TU might create an effect of interaction and informality. In addition, a hypothesis has been suggested that there might be a linguistic change in progress concerning ONTU in France French, as well as in Montreal French.
84 CHAPTER 6 RESULTS AND DISCUSSI ON ON L2 LEARNERS D ATA After an analysis of the data on L1 French, I will turn to that of L2 learners speech data in this c hapter. As in the last chapter, I will first offer an overview based on the surface features of the data before I discuss and interpret the patterns in more details. A comparative approach based on the speech data of L1 French speakers and French L2 learne rs will be adopted. Comparison with the other studies on French L2 learners acquisition of sociolinguistic competence will also be made when necessary. General patterns of the use of ON, TU, LON and VOUS General P atterns To express indefinite reference, the French L2 learners have four generic subject clitics at their disposal: LON, ON, VOUS et TU. In total, 491 tokens of generic pronoun are found in the learners speech samples. As illustrated in Table 6 1, ON is the predominant variant in the learners data (92.87%), and only very few tokens of TU (4.28%), VOUS (0.81) and LON (2.04) are found. The overall distribution shows little variation in the choice of the different variants. Instead, it suggests that the variant ON serves as the default variant used by the learners. This result is perhaps not surprising, given the fact that ON is the only variant that is explicitly introduced in textbooks and taught by instructors. It might be considered by the learners to be the correct form since it is the only form that is allowed in formal written French. Therefore it is interesting to see whether the learners have the same linguistic repertoire for using this default variant as their French native speaking counterparts. This question will be examined later. Table 6 2 shows a comparison of the use of the variable between the French nativespeaking participants and the learners. In comparison with the PFL1 data, the learners in the
85 UFFL2 group used ON much more frequently than their French native speaking count erparts. The variant, LON, that is absent in the PFL1 data, occurred in the speech sample of one learner. Concerning the use of TU/VOUS, while the TU form is the only informal variant produced by the native speakers in the PFL1 data, the learners in the U FFL2 corpus used VOUS, as well as TU, to express generic references in their conversation. Nevertheless, their rate of usage of TU/VOUS is considerably lower than that in PFL1. Table 6 1. The count and frequency of LON, ON, VOUS, TU in the speech data of French L2 learners Variant Number of tokens Relative frequency (%) ON 449 91.45 TU 25 5.10 LON 10 2.04 VOUS 7 1.41 Total 491 100 Table 6 2. Comparison of the use of ON, TU, VOUS and LON between PFL1 and UFFL2 Number of tokens Frequency of ON PFL1 UFFL2 PFL1 UFFL2 ON 362 449 55.3 91.45 TU 293 25 44.7 5.10 VOUS 0 7 0 2.04 LON 0 10 0 1.41 Total 655 491 100 100
86 Figure 61. Comparison of the use of ON, TU, VOUS and LON between PFL1 and UFFL2 The general patterns revealed in the French L 2 data suggest that while variation is displayed in the French L1 data and linguistic and sociostylistic constraints are found to influence the choice of the different forms by the French native speaking participants, the French L2 data does not display t he same amount of variability, which could mean that for these learners, the different generic pronouns are not associated with particular linguistic contexts and social stylistic signification. Individual D ifferences Despite the general patterns revealed, the learners use of the variable shows interesting interpersonal variation. Table 63 shows the distribution of the variants among each L2 participant. Table 6 3. The distribution of ON, TU, VOUS, and LON in the UFFL2 data Particpant # Sex N.ON TU VOUS LON Total % ON 12 15 17 18 13 19 16 14 11 Total F F F F M F F F M 74 54 53 35 65 36 42 43 47 449 0 0 0 0 0 1 4 0 20 25 0 0 0 0 3 2 2 0 0 7 0 0 0 0 0 0 0 10 0 10 74 54 53 35 68 39 48 53 67 491 100 100 100 100 95.6 92.3 87.5 81.1 70 91.45
87 As the table shows, LON is only found in the speech sample of participant # 14. TU and VOUS are used rarely and sporadically by four of the learners (#11, #13, #16, and #19). While participants #11 and # 13 chose either TU or VOUS as an alternate of ON, participants #16 and #19 used both of them. Almost all of the tokens of TU come from participant #11. Although little variation between ON and TU/VOUS has been observed in the French L2 data, there are still several questions that must be addressed: do the learners possess the same linguistic repertoire as the French natives when making generalizations; why did they switch to the other forms occasionally; was the use of TU/VOUS due to L1 transfer; what can account for the use of LON by Participant #14; finally, what can these patterns concerning the use of generic subject clitics tell us about the acquisition of sociolinguistic competence by the learners and what pedagogical implications can be found based on the results? These questions will be the main focus of the res t of this chapter. The linguistic repertoire of expressing generic references in French L2 data In this subsection, I will discuss the linguistic repertoire of the French learners when making generalizations and compare it to their French counterparts. Fi xed E xpressions As illustrated in the following examples, the French learners used ON in the fixed expressions :on va dire, comment dit on en franais, and on peut dire. ( 6A ) Jaime beaucoup que le professeur guide la discussion, et quil : quil met les : les ides chez (sur) la table q uON va dire. (UFFL2 # 12) ( 6B ) Comment est ce quON dit GLASS ? ( UFFL2 # 14) ( 6C ) Tout le livre se passe dans hum la psycho analy ste sur le sofa ON peut dire. ( UFFL2 # 18) Two of these expressions on va dire and on peut dire, occurred also in the French L1 data, as shown in example 5A, given in the last chapter and repeated here, and in example 6D.
88 ( 5A ) Honnt e ment je, mme mme ON va dire dans la cuisine franaise des : des choses assez simples je : je sais pas bien cuisiner du tout, du tout, du tout. (PFL1#01) ( 6D ) C est des appa rtements pour deux personnes, ON peut dire avec deux chambres (PFL1 # 08) In both the French L1 and the French L2 data, the expressions on va dire and on peut dire serve as a hedging device to protect t he face of the speakers. By using these expressions, the speaker implies that what he/she is going to say or just said is essentially a common view and hence avoids losing face. In addition to these fixed expressions that occurred in both French L1 data and French L2 data, some frozen expressions, that appeared in the French L1 data, are absent in the French L2 data, for example, on verra, on va dire, and on dirait. Although the French L2 learners used fewer fixed expressions containing the generic pr onouns than the native speakers, their use of some of them shows that they not only acquired the surface forms but also their functions. S entence S tructures Recall that when the native French participants were making generalizations, they used a variety of sentences structures, such as the presentatives il faut que, cest que, cest vrai que, je pense que, je trouve que, parce que, je me dis que, on dit que, on se dit que, and the implicative constructions found in si clauses and quand c lauses. Most of these sentence structures also occurred in the French L2 data, as the following examples show: ( 6E ) Je pense que cest la preuve quON ne doit pas voter pour les rpublicains (UFFL2 # 13) ( 6F ) Je pense que si ON habite la campagne, pe ut tre dans une petit ville, ON a beaucoup plus de temps pour soi (UFFL2 #12) ( 6G ) Je crois que cest ncessaire cest trs : cest le : cest la meilleu re faon damliorer parce que LON a besoin vraiment utiliser le franais (UFFL2 # 14)
89 ( 6H ) Il faut : il faut : sorganiser pour quON puisse dormir la nuit (UFFL2 # 12) ( 6I ) ON dit que demander un linguiste combien de langues quil parle, cest comme demander un mdecin combien de maladies quil a (UFFL2 # 12) ( 6J ) La langue volue continue voluer mais il svolue pas autant quavant (1.moins rapidement) mo ins rapidement, mais et il y a des influences plus grandes parce qu ON peut monter dcendre lchelle sociale. (UFFL2 # 11) ( 6K ) Si ON ne le fait pas au : au temps correct, ON a presque la note mauvaise, ON a perd (on perd) une ( un) point comme a. ( UFFL2 # 17) ( 6L ) ON doit tre en classe ON doit parler en classe, mais quand ON rentrer (rentre) chez soi ON doit regarder le texte. (UFFL2 # 14) ( 6M ) Quand ON voit qqch quON ne connat pas, ON doit le mettre dans notre mannire de penser, ou ON ne peut pas comprendre ( UFFL2 # 15) One sees that a variety of sentence structures are at the disposal of the French L2 learners, as well as the native speakers. However, although the learners acquired the structural repertoire to express generic refer ences, their choice of variant showed no variation. The linguistic factors that influence the use of the variable in the French L1 data did not constrain the use of the variable in the French L2 data of the present study. Pragmatic Dimension Of The G eneralizations Concerning the pragmatic dimension when making generalizations, I concluded, in the last chapter, that the French natives made generalizations in three pragmatic categories, namely, situational insertion description/explanation, and evaluations. Like the French natives, when the learners were exemplifying an argument, describing a fact, explaining a procedure, or evaluating a thing, they made generalizations very frequently. The examples 6N through 6S illustrate this. ( 6N ) Bon le gouvernement peut fa ire ce que les gens veulent pour ces buts mais pour les autres choses cest pas les rles de gouvernement dintervenir mon avis et beaucoup : il y a beaucoup de gouvernement qui interviennent dans les affaires des peuples parce que : mon avis ils ont peur, parce que quand ON a pas le pouvoir ON a peu : mme dans la vie si Tas pas le pouvoir Tas : par exemple si Tas une voiture Taime bien la voiture, TU ne
90 veux pas il y a un ami il a pas de voiture il a besoin de( ) quelque ch ose il a dit : ah est ce que je peux prter (emprunter) la voiture ? TU dis : ouai ok TU dis : ah cest un(une) bon(bonne) chose de laisser la voiture je sais il peut conduire mais Tas pas le pouvoir, TU peux rien faire et Tas un peu peur parce que Tas pas le pouvoir ( UFFL2 # 11) ( 6O ) je fais la cuisine un peu bizarre, mais une chose jaime bien faire cest CURRY et je fais : cest trs simple ON coupe des vgetaux (lgumes), ON fait une pole ON mettre (met) sur le feu et ON mettre (met) de lhuile, ON fait chauffer, aprs dix minutes non pas dix minutes, ON mettre (met) les vgetaux (lgumes) dedans le pole mais pas dans une : pas trs haut le feu parce que a fait ( ) les vgtaux (lgumes) et lhuile aussi, ON mettre (met) un couvert sur les pole, et ON reste : O N le reste : le ( ) rester pendant dix cinq minutes et jaime bien ( ) avec a, donc ON a fait ( ) de leau, ON fait ( ) dedans, ON lave pendant seulement 5 minutes. ( UFFL2 # 15) ( 6P ) et si VOUS avez des rves, suivez les (UFFL2 # 16) ( 6Q ) oui mme si TU ne fais pa s des : des stages trs professionnels je pense que cest important davoir au moins un travail un job pour avoir de lexprience. (UFFL2 # 16) ( 6R ) I l faut vraiment penser v os propres langues, lorsque VOUS traduisez, une : une autre et mon avis, par exemple, je ne dirais pas que je parle franais couramment ( ), mais je peux comprendre nimporte quoi, mais hum, je ne peux pas dire nimporte quoi, et donc si on si ON t udie la traduction entre le franais et langlais, et langlais franais par exemple, en particulier pour moi, entre langlais et le franais, il faut vraiment penser comment dire qqch (UFFL2 # 13) ( 6S ) Dans les tudes du chinois, la culture ma appris que cest important dtre la meilleur quON peut :, par exemple chaque jour, si ON ne fait r ien votre vie ntre pas utile ou trs, VOUS navez pas beaucoup de satisfaction avec votre vie, alors cest important de chaque jour de amlio rer votre identit ou pas votre identit ou vos penses etc., dessayer de amliorer votre soi, etc. (UFFL2 # 19) In 6N, participant # 11 was elaborating and exemplifying his argument of Si on a pas de pouvoir, on a peur. In 6 O, participant # 15 was explaining the procedure of cooking curry, and examples from 6 P to 6 S illustrate the stating of morals by the participants. In all those examples, one sees that they used ON or TU to make generalizations in the three pragmatic categories observed in the French L1 data.
91 It is interesting to note that it seems the pragmatic factor situational insertion has an effe ct on the choice of the generic pronouns in 6N, since participant # 11 used TU in this context. The use of TU/VOUS by the participants will be the focus of the next section. In sum, the use of the generic pronouns by the French L2 learners shows that although the linguistic resources examined do not have an effect on the choice between the different generic pronouns, they are nevertheless adopted by them to express generic references. The learners possess the same structures as the French native participan ts when making generalizations, but their use of the variable did not show much variation. The use of TU/VOUS in the French L2 data As mentioned earlier in the chapter, unlike the French native speaking participants, the French L2 learners used the generic TU and VOUS very rarely. I will base this section on the speech samples of participants # 11, 13, 16, and 19, where TU and VOUS occurred, to discuss their patterns of usage. Before examining the use of TU and VOUS in the data, two hypotheses can be made r egarding the scarcity and the use of the variants: 1) It could be due to L1 transfer, namely, the French L2 learners transfer the English generic pronoun you to the French TU/VOUS. Recall that in English, the second person pronoun you is also used as generic pronoun (Hyman, 2004) As TU/VOUS is the literal translation of you in French, it is possible that the learners were influen ced by their L1 when using the generic pronouns. 2) The second hypothesis would be that the learners are starting to acquire the variation pattern of ON and TU/VOUS as displayed in the French L1 data. They might be at the beginning point of acquiring the s ociolinguistic variation of ON TU/VOUS. To verify whether the use of TU/VOUS in the French L2 data is due to English transfer or fits the L1 French pattern, one needs to compare the use of TU/VOUS in the two data sets.
92 In the French L1 data, I noticed that occurrences of TU occur successively in sequence instead of alternating between ON and TU, as illustrated in 6 Q, and in 4W which is given earlier in the chapter on methodology. ( 6 T ) Un hamea u cest un, un endroit de la campagne oTas trois ou quatre maisons ; cest tout c'est dire y a pas de ( ), y a pas de cachet, y a pas de poste, y a pas de, y a rien cest que de vieilles maisons. Ben faut aller ; si TU veux faire des courses, si TU veux chercher le pain, faut prendre la voiture et rouler quinze ou vingt kilomtres au village le plus proche. Cest mme pas une ville, cest un village. Donc TU vas au village, o y a, TU peux aller chez le boucher, TU peux aller au centre commercial, enfin des petits centres, ben voil (PFL1 # 08) ( 4W ) 1 Oui, jai des amis qui disent que le voyage cest un bon moyen de rencontrer des gens ( ?) 2 Oui cest sr, fin aussi, si TU [4 W 1] fais un voyage deux ou trois mme mme l cest bien, mais si TU [4 W 2] fais un voyage en grand groupe, fin certainement ya tellement d e chances que TU [4 W 3] restes dans ton groupe que TU [4W 4] rencontreras personne, mais si TU [4 W 5] pars tout seul pour un certain temps dans un pays TU [4W 6] peux pas rester cloitr chez toi pendant trois mois, cest pas drle et mme partir du m oment o TU [4W 7] vas aller faire fin cest vrai cest plus le cas fin cest vraiment le cas en France mais partir du cas TU [4 W 8] vas tous les jours dans un magasin, les gens ils vont te connaitre ( ?) mais sinon si TU [4 W 9] vas faire quoique ce s oit mais de toute faon TU [4W 10] es oblig dtre confront des gens de partout cest les alas de la vie, et cest vrai que cest, quand TU [4 W 11] voyages T(TU) [4W 12] as, cest plus inhabituel et donc cest plus tes sens sont plus en veil T(T U) [4 W 13] est plus oblig dtre actif ON [4w14] va dire donc je pense que TU [4W 15] rencontres beaucoup plus de gens (PFL1#07) This pattern corroborates Sankoff and Laberges finding about three pronominal variables nous on, ontu/vous and onil s, and Coveneys finding about the variable ils elles (Coveney, 2004b; D. Sankoff & Laberge, 1978a) As mentioned in Coveney (2004b), the succession of occurrences may be due to three reasons: to av oid causing any misunderstanding and confusion; to keep textual coherence; to keep stylistic homogeneity during a passage of interview. If the use of TU is the French L1 data shows a regular pattern of succession of occurrences, this is not exactly the cas e in the French L2 data. Most of the learners used TU or VOUS either
93 sporadically in isolated sentences or in alternation with ON in the same sequence, as illustrated in examples 6 P through 6S. Only participant # 11 ever used TU in a sequence, as shown i n 6N. From examples 6 P and 6Q, one sees that participant # 6 used TU/VOUS sporadically in her speech sample, while using ON most of the time. In examples [6r] and [6s], when participant # 13 and the participant # 19 were talking about a same topic in a sequence, they switched between VOUS and ON instead of using one of them successively. Despite the fact that most of the French L2 learners didnt use TU or VOUS in sequence, participant # 11 used TU successively when he was exemplifying his argument about power and fear. Participant # 11 is the one who used most of the tokens of TU found in the data (20 out of 25 tokens are exhibited by the participant # 11). In addition, as mentioned earlier, although not all the time, he did use TU in sequence and in the pragmatic context situational insertion, which has been proven to favor the use of TU in the French L1 data. His use of TU suggests a positive sign of the acquisition of this informal variant, which is not explicitly taught in the classroom. The point of the acquisition of sociolinguistic variation by the learners will be discussed later in the chapter. In addition to the fact that learners #13, 16 and19 did not use TU or VOUS in a sequence, the participants # 16 and # 19 didnt use either TU or VOUS consistently in their speech sample, which is another pattern that did not match the use of TU/VOUS in the French L1 data. In the latter, the participants used TU consistently during their interviews. This shows that these learners have not assimilated the na tive patterns of the language. Finally, recall that in the French L1 data, the generic TU is tinged with its original meaning of address pronoun. Participants used TU as both address pronoun and generic pronoun. In the French L2 data, as shown in Table 6 4, of the four participants who used TU/VOUS, participant
94 # 11 used TU to address me and he also used TU as generic pronoun consistently, since he used a number of generic TU, it is suggested that he acquired the correspondence of address pronoun and generi c pronoun; participant # 13 used VOUS as both address pronoun and generic pronoun, however, given the scarcity of tokens of VOUS in her speech sample (3 tokens), I am cautious of interpreting it as acquisition of the correspondence between the two; although participant # 16 used TU to address me, she switched between TU and VOUS as generic pronouns; and it is impossible to examine the correspondence of address pronoun and generic pronoun in participant # 19s speech sample since he did not use any address p ronouns. Table 6 4. The correspondence between address pronouns and generic pronouns in participants #11, 13, 16 and 19 Participant # Address pronouns Generic pronouns 11 tu TU 13 vous VOUS 16 tu TU/VOUS 19 N/A TU/VOUS In sum, the comparison of the use of TU/VOUS between the two sets of data revealed a learner difference: one the one hand, the use of the variants by participants #13, #16 and #19 showed a random pattern because they did not use it in sequence, switched between TU and VOUS, or did not use the address pronoun and generic pronoun in correspondence; on the other hand, participant #11 showed a positive sign of the acquisition of the use of TU because he used TU in situational insertion, in a sequence, he used TU consistently during the in terview, and he used TU as both address pronoun and generic pronoun. Based on the comparison, while it can be suggested that participant#11 is at the starting point of assimilating the sociolinguistic patterns in native French, it is difficult to prove tha t the randomness of the use of the variable in the other learners speech samples is actually due to L1 transfer. In general, the use of TU/VOUS as
95 generic pronouns by the French L2 learners shows that they differ from their native counterparts by conformi ng to the norms of the target language to a much higher degree. The use of LON As mentioned in the section on general patterns of the use of generic pronouns, one must notice that a new form LON, which was not used by the French native speakers, was, how ever, adopted by one of the learners to serve the generic function. Although the use of LON is absent in the data of the French native participants in the present study, this variant does exist in spoken French, as demonstrated in Coveney (2004a). In Cove neys study, the author also found that almost all of the tokens of LON in the Picardy corpus appeared after que, and he concluded that LON, which is used very marginally in informal spoken French (1.2% in the Picardy corpus), indicates a conservative and highly formal speech style (Coveney, 2004a) Based on the findings of Coveney, three possible interpretations can account for the use of this variant by the learners: 1) if LON is a formal variant and is normally associated with literary language (Goosse, 1959; Sandfeld, 1965) one may hypothes ize that the learners spoken French could be influenced by her formal written French, which means that she doesnt quite know that LON belongs to formal written French in general; 2) it is also possible that the learner might indeed intend to use this hyperstyle variant to express formality; 3) however, one cannot exclude a third possibility: the learner was indeed aware of the fact that LON is used occasionally in spoken French, but she hasnt acquired the stylistic difference of LON and ON, which wo uld result in a random alternation between LON and ON in her speech sample. The three interpretations differ concerning whether or not the participant acquired the stylistic signification of LON and ON, as well as the stylistic difference between spoken French and oral French. An examination of the patterns of the use of LON by this participant will help to shed light on the situation. The participant exhibited 10 tokens of LON out of a total of 53 generic
96 pronouns, which means that she used LON nearly 20% of the time. This rate is much higher than that of the highest in the previous literature, which is 8.4%, found in the participants of the Picardy corpus in Coveney (2004a). If the participant indeed used LON for the purpose of being formal (as we wi ll demonstrate later in this paragraph), her overuse of the variant might signal that she perceived the situation as more formal than her French counterparts. This is understandable since the learner speaks French as a second language; she would certainly be more aware of her linguistic performance than French native speakers (Coveney, 1998) Concerning the effect of the linguistic contexts in which LON is preferred I found that all ten of the 10 tokens of LON in this participants speech are preceded by que, which generally conforms to the result in Coveney (2004a). As shown in 6U, the only difference is that while the participant used LON after que as a con junction, the participants in the Picardy corpus used LON more frequently when it is preceded by que as a relative pronoun. ( 6U ) 1) lavantage cest que LON peut aller aux restaurants diffrents tous les soirs (UFFL2 # 14) 2) le truc qui / que lon met de dans (Coveney, 2004a) Nevertheless, the participant generally followed the linguistic constraint on the use of this variant. Finally, I notice d that whenever the participant used LON, she was evaluating something or stating a moral, as illustrated in the following examples: ( 6V ) je crois que cest ncessaire cest trs cest le (silence) cest la meilleure faon damliorer parce que LON a besoin vraiment utiliser le franais (UFFL2 # 14) ( 6W ) jai appris d es principaux qui a dit que LON doit partager avec le monde pour tre plus catholique (UFFL2 # 14) I n the learners speech sample, since LON only ever occurs in this pragmatic context, which is asso ciated with formality (refer to last chapter and figure 5 1), I suggest that the learner was indeed using LON to sound formal or conservative. In other words, she acquired the
97 stylistic signification of LON in her spoken French. Overall, based on the exa mination of the patterns of the use of LON by the participant #14, it is suggested that in general, the participant patterns with French native speakers concerning the effect of the linguistic contexts and formality, although she overused the LON variant This finding contrasts with that of Howard (2006), which examined variable liaison and found that the learners in his study used variable liaison including hyperformal liaison very limitedly. To end this section, it must be noted that although I argue towards an acquisition of the stylistic variation between LON and ON by this learner, I am not suggesting that the participant acquired the stylistic variation between the three variants LON, ON and TU/VOUS, since the latter is not found in her speech samp le and we cannot prove that she acquired the informal variants TU/VOUS and possessed the same stylistic repertoire as her French counterparts. The acquisition of the sociolinguistic variation of the variable ON TU/VOUS After an examination of the use of th e different variants by the learners, I will provide answers to the research questions concerning the acquisition of the sociolinguistic variation of the variable ON TU/VOUS in this section; namely, do the learners possess the same linguistic and stylistic repertoire when using the variable and what factors influence the use of the different variants? From the previous section, I have demonstrated that the French L2 learners made generalizations in the same syntactic and pragmatic categories as their French counterparts. Concerning the stylistic repertoire, since all of the learners used the formal variant ON most of the time, and even participant #14 used the hyperstyle variant LON, it is argued that they acquired the formal style of making generic reference. The scarcity of the tokens of TU/VOUS leads to the suggestion that most of the learners have not yet internalized these informal variants and were lacking the informal repertoire of making generalizations. One may question the
98 validity of this conclusi on for the following reasons: 1) since seven out of the nine learners in the present study are females, and the predominant use of ON is associated with females, one may wonder whether the predominance of ON in the French L2 data is due to the fact that th e number of females learners outweighs that of male learners. 2) If the variable ON TU/VOUS has been demonstrated to be a stylistic variable, then it is possible that the learners perceived the context of the sociolinguistic interview as more formal than t he French native speakers, since for the first, French is not their mother language and they might be more nervous in this kind of situation, and thus might have exhibited a more formal speech style than the French native speaking participants. While one c annot directly demonstrate, based exclusively on the linguistic behavior of the ON TU/VOUS variable, which interpretation is more pertinent to explain the scarcity of the tokens of TU/VOUS in the French L2 data, one can still look at other parts of the dat a to see whether or not the learners exhibited an informal style overall. While transcribing, I noticed that while the French native speakers used a lot of informal forms such as the deletion of l and the negative particle ne or even the deletion of th e impersonal subject il from il y a, the French L2 learners seldom dropped these elements, as illustrated in the examples below. This overall impression leads me to suggest towards a lack of acquisition of the informal speech style. This finding corroborates that of the previous studies on the acquisition of sociolinguistic variables, which has shown that learners with a formal instructional background make minimal use of informal variants (Mougeon et al., 2002; Mougeon et al., 2004; Nadasdi, 2005; Rehner & Mougeon, 1999; Rehner et al., 2003; Uritescu et al., 2004) ( 6X ) y a plus de m ixits on va dire dans la fac. (PFL1 # 03) ( 6Y ) il y a un une classe dhonneur, je ne sais pas si on peut dire a, dans le dpartement hum danglais (UF F L2 # 18)
99 Despite the general pattern of the lack of variation between ON and TU/VOUS, one of the learners, participant # 11, displayed a use of TU that was close to the French L1 pattern. This learner difference is not surprising if one analyzes his proficiency level and his contact with native speakers. The proficiency level of Participant # 11 was classified as advanced high earlier in the methodology section, according to his performance in the interview. In addition, according to the information given in the questionnaire and in the interview, he had extensive contact with French native speakers when he was in France because he lived with a French family and had dinners and other activities regularly with the family members. When he is in Gainesville, he maintains regular interaction with French native speakers since he is a French teaching assistant, works r egularly with French speakers, and is part of the social network of his French colleagues and friends. His high proficiency level is very likely to play a role in his correct pattern of the use of TU and is unlikely to lead one to associate his use of TU w ith L1 transfer. His regular and extensive contact with French native speakers in informal situations provides more opportunities for him to assimilate the pattern of ON TU/VOUS variation and hence increases his use of the informal variants TU. From this p erspective, one may wonder whether the scarcity of TU in the other learners speech samples is due to their lower proficiency level and degree of contact with native speakers. Based on the general linguistic background of the participants (cf. Table 42), the information on their contact with native speakers, and their performance during the interview, participants #13 to #19 do display a gap from participant #11, either their proficiency level is lower, or their contact with the native speakers is not as h igh as that of participant #11. However, one must notice that participant #12 was also classified as an advanced high learner and she had spent 2 years in France as a high school student; in addition, she keeps regular contact with her French
100 friends throu gh online chatting, and she also indicated on the questionnaire, that she has French friends and they get together sometimes. Then how to explain the absence of TU in this learners speech sample? Note that participant #12 is a female learner; since ON is clearly a female variant, like most of the females in the PFL1 data, it might not surprising that she would have produced ON categorically, even if she had acquired the informal variants TU/VOUS. More speech data from female learners, who share similar lan guage background with participant #12, are needed in order to verify this interpretation. The development of the acquisition of the informal variants TU/VOUS is very likely to follow a U shaped curve, as visualized in Figure 61. In my teaching experience with firstyear French learners, I noticed that as a result of L1 transfer, the beginners used TU/VOUS frequently when they made generic reference; their use of TU/VOUS is located somewhere in the A section of the U curve. As their proficiency levels rise, their L1 transfer reduces, and so does the number of TU/VOUS. The bottom of the U curve represents the period when the least L1 transfer is found. Participants #13 to #19 in the present study would be placed in this section. Finally, as learners language competence continues to rise, they start to realize that TU/VOUS is actually allowed in spoken French and that it alternates with ON; at this time (section C), they start to use these variants, which were once corrected by their instructors in their writt en French, to make generic references. Participants #11 and #12 are at the beginning part of this section. Although learners use TU/VOUS substantially in both period A and period C, the cause and the pattern of the use are different. In period A, due to L1 transfer, learners might not be able make generalizations using the linguistic resources demonstrated earlier in the chapter and the variation pattern might be very different from the French L1 pattern. In contrast, in period C, learners not only possess the same linguistic repertoire of expressing generic references as
101 French native speakers, but also start to alternate ON and TU/VOUS with a pattern that is close to the French L1 pattern. They will also generally follow the linguistic and sociostylistic constraints on the variation of the ON TU/VOUS variable. Previous studies, which are interested in examining French L2 learners in a natural context where they are in daily contact with French native speakers and learned French informally (Blondeau, 2008; Blondeau & Nagy, 1998; Blondeau et al., 2002; G. Sankoff et al., 1997) have shown results that correspond to the acquisit ion of sociolinguistic variables in period C. Further investigation and empirical evidence concerning French beginners are needed to demonstrate the validity of this model concerning the development of the acquisition of the sociolinguistic variation of the ONTU/VOUS variable. Figure 61. U course of the development of the acquisition of TU/VOUS After examining the use and the development of this variable in the French L2 participants in the present study, it is now necessary to discuss what factors influence the acquisition of the sociolinguistic variation of the variable. In addition to proficiency level and contact with native speakers, which were discussed earlier, what other independent variables can influence the acquisit ion of the informal variants TU/VOUS? While contact with native speakers has been demonstrated to play a significant role in the acquisition of sociolinguistic variables (Blondeau & Nagy, 1998; Blondeau et al., 2002; A B C
102 Mougeon et al., 2002; Mougeon et al., 2004; Nadasdi, 2005; Rehner & Mougeon, 1999; Rehner et al., 2003; G. Sankoff et al., 1997; Uritescu et al., 2004) one cannot ignore the effect of explicit instruction; namely, it is important to at lea st introduce explicitly the variation phenomenon to students and make them aware of the fact that informal variants like TU/VOUS are sometimes frequently used by French native speakers in informal spoken French. The avoidance of using TU/VOUS by most of th e learners in this study is very likely due to the fact that they consider the form as incorrect or ungrammatical, even in spoken French. This can be seen by the self correction that occurred in their speech samples, as illustrated in the following example, when participant #19 produced ce que tu veux dire before she self corrected to ce quon veut dire. ( 6Z ) En gnral, hum, il y a un peu comprendre ce que TU veux dire, ce quON veut dire hum, si ON utilise la faon de parler quON a appris dans le cours, mais pour tre dans ( ) pour tre naturel, je pense quON doit rester ou vivre dans ( ) de sa langue pendant quelque temps ( UFFL2 # 19) If we are to encourage students to develop towards the right hand side of the U curve, we need to let them know that the use of TU/VOUS is acceptable in informal contexts. To end this section, it is suggested the learners in the present study are developing in a positive direction towards the acquisition of the sociolinguistic variation of the variable ON TU/VOUS, as illus trated in Figure 6 1. Concerning the independent variables that influence the acquisition, I suggested that proficiency level, contact with native speakers, and explicit instruction that would help the learners notice this variation, all favor the acquisition of the variation pattern of sociolinguistic variables.
103 Concluding remarks In this chapter, I have attempted to reveal the patterns of the use of the different generic pronouns in the French L2 data before I compared them to the French L1 data and gave the interpretations of these patterns. I discussed, after that, the acquisition of the sociolinguistic variation of the variable and the independent variables that influence it. In general, the similarities and differences of the use of the variable in th e two sets of data can be summarized as below: Similarity: the French learners, as well as the French native speakers, possess a variety of sentence structures and pragmatic contexts to make generic references. Difference: while the French native speakers alternate ON and TU according to different linguistic and sociostylistic constraints, the French L2 learners show little variation. For most of them, the use of ON TU/VOUS does not really signal any speech style; the fact, that the most fluent Frenchspeaking learners #11 and #12 used ON TU/VOUS in a different way and belong to different gender, gave a positive sign that learners would be able to express gender identity once they achieve at a higher level of proficiency and obtain more exposure to French. Nevertheless, it is demonstrated that participant #11s use of TU is similar to that of French native speakers. If we can consider the learners as a French L2 speech community, then leaded by the male learner participant # 11, the learners might be about t o develop towards the right part of the U course which represents the development of the acquisition of the sociolinguistic variation of the variable. Concerning the independent variables that influence the acquisition of this variable, I suggested that pr oficiency level, contact with French native speakers and the explicit instruction that would help to raise the level of awareness of learners, all favor the acquisition of the sociolinguistic variation of ON TU/VOUS.
104 CHAPTER 7 CONCLUSION In this thesis I h ave attempted to examine and compare the linguistic and sociolinguistic variation of the generic subject clitics ON TU/VOUS in L1 French and L2 French. In the L1 French data, I have first provided a synchronic analysis of the variation of the generic subje ct clitics ON TU. I have compared the rate of their usage across three sociolinguistic corpora collected over different periods of time, providing a diachronic perspective. The comparison in real time suggested a possible linguistic change involving the va riable; namely, TU is in progression and ON is in regression in Paris French. I posited that if this is the case, the linguistic change was motivated by a diachronic evolution of pragmatic configurations in interpersonal interaction, namely, people use the informal TU more frequently because interpersonal interaction is more and more informal and interactive, at least in the informal interview context in which the data of the corpora under comparison were collected. Since previous studies concerning the linguistic change of ON TU/VOUS has been focused on Montreal French; this finding of the present study provides evidence of linguistic change concerning ON TU/VOUS in France French and hence complemented existing French L1 literature on this topic. It also co ntributes to the research question concerning the motive of the linguistic change in Montreal French by showing that contact with English did not cause the change. In addition to the analysis in real time, I have focused on a synchronic analysis of the lin guistic and sociostylistic variation of the variable under investigation. Linguistic constraints examined include verbs, syntactic frames, and pragmatic categories; the extralinguistic factors considered are gender and style. Some results concerning the linguistic variation confirm in general the findings of the previous studies, whereas others complement the existing literature.
105 Specifically, the verbs dire and voir favor the use of ON and vouloir and avoir favor the use of TU; presentatives are found to favor the choice of ON whereas existential sentences favor that of TU; finally the participants used ON predominantly in evaluations and TU in situational insertion, while the description and explanation is a neutral context where the use of ON and TU was parallel. As for gender, a highly significant sex difference has been confirmed regarding the choice of the variable. Since ON is considered to be the standard form and TU the informal form, it is not surprising that women, who are often found t o prefer prestigious language forms, favor ON and men, who often value covert prestige, prefer TU. In fact, the sex difference found in the present study is even a lot bigger than that of previous studies. I have focused on the question of stylistic variat ion of the variable in my discussion: based on an examination of 1) the correlation between the generic TU and the discourse markers tu sais / tu vois which denote familiarity; and 2) topics of speech, it has been argued that the ON TU variable is stylis tically marked, namely, ON is associated with formal speech style while TU signals an informal speech style. This finding provides empirical evidence to the study of the stylistic variation of ON TU/VOUS, which had not been successfully demonstrated before the present study. In the L2 French data, little variation between ON, TU, and VOUS has been observed. The learners used ON by default and switched to other forms of generic pronoun TU/VOUS/LON only occasionally. Although the occurrences of TU/VOUS could result from transfer of English you, data from beginners are needed to verify. According to the U shape curve which represents the development of the acquisition of the variable, most of the learners in the present study are placed in the bottom of the U shaped curve, which means that they use TU minimally,
106 where the L1 transfer is also the least. However, based on the fact that one learner, participant #11 in the present study who had much contact with French native speakers, followed the native pattern s when using TU, it is suggested that the learners are on the verge of developing towards the right side of the U shape curve, where an increase in the use of TU/VOUS will be observed, but this time not due to the L1 transfer but as a result of assimilatio n of the native patterns of the variation. After an analysis of L2 French patterns on the use of the variable, I have turned to discuss the independent variables that influence the use of the different variants. It is suggested that proficiency level in Fr ench, contact with French native speakers and explicit instruction each play a role in the use of the variants. This finding corroborates with that of previous studies based on different sociolinguistic variables (Mougeon et al., 2004; Nadasdi, 2005) As language proficiency levels rises, learners will naturally be more at ease with grammar forms and formulation of language, and hence be able to pay attention to the soc iolinguistic patterns which are beyond the level of just being understandable and grammatical. To express speech formality or to signal social identity must be a sociolinguistic competence that is above grammar competence; participant # 11s frequent and c orrect use of TU suggested that being part of a social network which involves French plays a role in acquiring native patterns of sociolinguistic variation. Finally, it is argued that the scarcity of informal variants in the learners speech samples was pa rtially due to the lack of explicit instruction concerning the variation between ONTU/VOUS in informal spoken French. The learners might not be aware of the fact that the forms TU/VOUS are used by native speakers in informal speech. These findings concerning French L2 conform to that of the existing literature on the acquisition of sociolinguistic competence, which have found that instructed learners overused formal variants and only use
107 informal or vernacular variants at a much lower rate than native speakers (Bayley & Regan, 2004) The generic ON is often translated to one or people in English by educators. While ON and one might be considere d to be equivalent in formal writing, they are not exactly equivalent in informal speech. When some of the learners were asked to answer in English some of the same questions that I had asked in French, the English form that was used in place of ON turned out to be you, as expected. Thus it is empirically demonstrated that you is the English equivalent form of ON, at least in informal speech. It is possible that the hyperstyle form LON could be considered as one, although empirical evidence needs t o be found. Further investigation on the variation of the ON TU/VOUS variable in L1 French might want to look at other social groups such as older people and teenagers. If an age difference can be found, it would support my hypothesis of an ongoing linguis tic change with the variable. Speech samples in other interactional situations, such as intimate conversation among close friends, would allow one to collect more vernacular speech data, which will be helpful in stylistic analysis. Similarly, group convers ations with peers (without the presence of a researcher) might be a context in which L2 learners feel the most comfortable when speaking French and would thus allow one to collect their most naturally occurring speech data. More Data from male learners, be ginners and superior learners would help to shed light on many questions that were mentioned in the discussion, for example, the question concerning L1 transfer. In French SLA, it is interesting to look at the effect of instruction on the acquisition of sociolinguistic variation and develop tasks and materials to teach sociolinguistic variation.
108 APPENDIX A INFORMED CONSENT Informed Consent Protocol Title: Comparative study of pronominal variation in French as a first language and by advanced American lear ners of French as a second language Please read this consent document carefully before you decide to participate in this study. Purpose of the research study: The purpose of this study is to examine and compare the linguistic and sociolinguistic variati on of pronouns in spoken French as a first language and by advanced American learners of French as a second language What you will be asked to do in the study: If you choose to participate in this study, you will be asked to do an informal conversation c oncerning topics such as your personal experience, preferred activities in life, family, friends, pets, childhood, hometown, study, work, French language, general aspects and features of France culture, economy, society and politics. No specific judges wil l be concerned to any social groups or members. You are free to refuse to talk about any of the topics listed above or to answer any of my questions concerning the topics. Time required: 45 minutes to one hour Risks and Benefits: There are no risks and b enefits associated with this study. Compensation: You will be paid 10 US dollars per hour for participating in this study. Confidentiality: Your identity will be kept confidential to the extent provided by law. Your information will be assigned a code number. The list connecting your name to this number will be kept in a locked file in my office. When the study is completed and the data have been analyzed, the list will be destroyed. Your name will not be used in any report. Voluntary participation: You r participation in this study is completely voluntary. There is no penalty for not participating. Right to withdraw from the study:
109 You have the right to withdraw from the study at anytime without consequence. Whom to contact if you have questions about the study: Jingya Zhong, Department of Romance languages and Literatures, University of Florida. Office location: 3C Basement Dauer Hall. Email: firstname.lastname@example.org or email@example.com. Ph: (352)8716805, or (352) 3922016 ext. 270 ( department). Whom to contact about your rights as a research participant in the study: IRB02 Office, Box 112250, University of Florida, Gainesville, FL 326112250; phone 3920433. Agreement: I have read the procedure described above. I voluntarily agree to participate in the procedure and I have received a copy of this description. Participant: ___________________________________________ Date: _________________ Principal Investigator: __________________________________ Date: _______________ Formulaire de consentement Titre de la recherche : Etude comparative de la variation pronominale en franais langue premire et chez les apprenants amricains avancs du franais langue seconde Veuillez lire attentivement ce document de consen tement avant de decider de participer cette recherche. Objectifs de la recherche: Ce projet vise examiner et comparer la variation linguistique et sociolinguistique des pronoms en franais langue premire et chez les apprenants amricains du franais au niveau avanc. Participation la recherche: La participation cette recherche consiste rencontrer lagent de recherche pour une entrevue de 45 minutes une heure un moment et dans un lieu que vous choisirez. Cette entrevue portera sur votre e xprience personnelle en milieu scolaire ou professionnel, vos activits prfres, votre famille, vos amis, vos tudes, votre travail, et des aspects gnrales de la culture, lconomie, la socit et la politique franaise. Vous tes libre refuser de p arler des sujets ci dessus ou de rpondre mes questions concernant des sujets ci dessus. Lentrevue sera enregistre, puis transcrite. Dure: 45 minutes une heure Risques et bnfices:
110 Les participants ne recevront aucune risque et recevront 10 eur os pour la participation. Compensation: Il ny aura pas de compensation spciale pour participer cette tude. Confidentialit: Les renseignements que vous nous donnerez demeureront confidentiels. Les entrevues seront transcrites et les enregistremen ts effacs. Chaque participant la recherche se verra attribuer un numro et seul le chercheur principal et/ou la personne mandate cet effet auront la liste des participants et des numros qui leur auront t attribus. De plus, les renseignements sero nt conservs dans un classeur sous cl situ dans un bureau ferm. Aucune information permettant de vous identifier dune faon ou dune autre ne sera publie. Ces renseignements personnels seront dtruits aprs la fin du projet. Votre nom ne sera pas util is dans aucun raport. Participation Volontaire: Votre participation cette etude est compltement volontaire. Il ny a pas dindemnit pour ne pas participer cette tdue. Droit de retrait: Vous tes libre de vous retirer en tout temps sur simple avis verbal, sans prjudice et sans devoir justifier votre dcision. Si vous vous retirez de la recherche, les renseignements qui auront t recueillis au moment de votre retrait seront dtruits. Contacts si vous avez des questions sur cette recherche: Jingya Zhong, 170 Dauer Hall, Department of Romance languages and Literatures, University of Florida. USA. 32611. Tl: 1 352 871 6805, ou 1 352 392 2016 ext. 270. Courriel: firstname.lastname@example.org ou email@example.com. Contact s sur vos droits comme participant de recherche dans cette tdue: IRB 02 Office, Box 112250, University of Florida, Gainesville, FL, USA. 326112250; Tl: 1 352 392 0433. Courriel: firstname.lastname@example.org. Consentement: Je dcl are avoir pris connaissance des informations ci dessus. Je consens volontairement participer cette tude. Jai reu une copie de cette description. Participant: ___________________________________________ Date: _________________
111 APPENDIX B QUESTIO NNAIRE Questionnaire You may skip any questions you do not feel comfortable answering. I. Personal information Name : Sex : F M Year of birth : Fathers profession: Mothers profession: Firs t language(s): Other languages you speak and proficiency level (Beginning, intermediate, or advanced) ____________________________________________________________ ____________________________________________________________ II. French study 1, H ow long have you been studying French ? _______________________________ 2, What do you think of your oral proficiency in French? A, Very fluent B, Fluent C, Not very fluent, but I can communicate in French without difficult ies 3, Have you ever been in France? A, Yes. B, No. (Please skip to no.7) 4, When and how long have you stayed in France? ____________________________________________________________________ _____________________________________________________________________ _____________________________________________________________________ 5, Did you take classes there? A. Yes B. No. If no, please specify your purpose of stay in France: _____________________________________________________________________
112 6, When you were in France, what were your activities with French speakers out of class and how often? ______________________________________________________________________ ______________________________________________________________________ _______________________________________________________________________ 7, In Gainesville, did you spend time with French native speakers last month? If yes, please specify: A, I have stayed in a French speaking family B, I made a trip with French native speakers C, I have French friends and we get together sometimes. D, Others, please specify: _____________________________________________________________________ _____________________________________________________________________ If no, please skip to no.8. 8, Over the last two weeks approximately how many hours did you spend on listening to French radio or watchi ng French TV (online)? _____________________________________________________________________ 9, Over the last two weeks approximately how many hours did you spend on reading French books, magazines or newspapers? _____________________________________________________________________ 10,When you took your class of Beginning French, what was the title of the textbook? ______________________________________________________________________ III. Contact Information Email:_________________________________________________________________ Phone number:__________________________________________________________ Merci beaucoup
113 APPENDIX C INTERVIEW QUESTIONS Some of the questions I a sked during the interview with p ar ticipant # 4, a female Parisi an: 1. Tu peux me parler un peu tes tudes? 2. Tu peux me parler de ton stage ? 3. Oui, est ce que tu penses que lEspagne est trs diffrente de la France, au niveau de la culture ? 4. Tu as t en Espagne, et tu as t dans dautres endroits ? 5. Toi tu aime plutt habi ter dans une grande ville ? 6. Quest ce qui tintresse le plus ( New York) ? 7. Normalement comment est ce que tu cherches un JOB ? 8. Tu comprends bien le franais qubcois, le franais dAfrique, (et le franais) de dautres pays francophones ? 9. Quelle est sa meilleure recette ? 10. Tu sais cuisiner quelque chose ? 11. Comment faire le gratin dauphinois ? 12. Tu as dautres projets aujourdhui ? Some of the questions I a sked during the interview with p articipant # 11, a male American learner of French: 1. Tu peux me dcrire un film que que tu as vu et lhistoire dans le film, quest ce qui sest pass dans le film ? 2. Tu peux imaginer comment a sera ta vie dans dix ans ? 3. Tu peux me dcrire un endroit que tu connais trs bien, o est ce que tu vas souvent ? 4. Tu peux me dire com ment y aller? 5. Est ce que tu peux comparer les adantages et les dsavantages dhabiter dans une grande ville et une petite ville ? 6. Tu peux me parler un peu des cours que tu suis ce semestre ? 7. En gnral quelle sorte du cours est ce que tu prfres ? 8. Tu peux partager avec moi un peu ton exprience dapprendre le franais ? 9. Est ce que tu sors souvent avec tes amis franais tes camarades de classe franais du cours? 10. Comment amliorer lacquisition dune langue selon toi? 11. Est ce que tu sais comment faire un plat franais ou amricain ? 12. Est ce que tu as t dans une situation trs dangereuse ? 13. Est ce que tu peux comparer les avantages et les dsavantages pour soutenir la partie dmocratique ?
114 APPENDIX D GOLDVARB DATA Cross tabulation_Group1+2
115 Cross tabula tion_Group2+3
116 Cross tabulation_Group 2+4
117 LIST OF REFERENCES Armstrong, N. (2002). Variable deletion of French ne : a cross stylistic perspective. Language Sciences, 24 153173. Ashby, W. J. (1981). The Loss of the Negative Particle ne in French: a Syntactic Change in Progress. Language 57 (3), 674687. Ashby, W. J. (1992). The variable use of on versus tu/vous for indefinite reference for spoken French. Journal of French Language Studies, 2, 135157. Ashby, W. J. (2001). Un nouveau regard sur la chute du ne en franais parl tourangeau: s'agit il d'un changement en cours? Journal of French Language Studies, 11, 122. Bayley, R. (2000). Second language acquisition and variationist linguistics. American Speech, 75(3), 288290. Bayley, R., & Regan, V. (2004). The acquisition of sociolinguistic competence. Journal of Sociolinguistics, 8 (3), 323338. Beeching, K. (2002). Gender, Politeness and Pragmatic Particles in French. Amsterdam: Benjamins. Blondeau, H. ( 2008). The dynamics of pronouns in the Qubec languages in contact dynamics. In M. Meyerhoff & N. Naomi (Eds.), Social Lives in Language Sociolinguistics and multilingual speech communities: Celebrating the work of Gillian Sankoff Amsterdam: John Benjam ins Publishing Company. Blondeau, H., & Nagy, N. (1998). Double marquage du sujet dans le franais parl par les jeunes anglomontralais. In J. J. a. G. V. Herk (Ed.), Actes du congre' s annuel de l'Association canadienne de linguistique (pp. 5970). Ott awa, Canada: Cahiers Linguistique d'Ottawa. Blondeau, H., Naomi, N., Sankoff, G., & Thibault, P. (2002). La couleur locale du franais L2 des AngloMontralais. Acquisition et Interaction en Langue Etrangre, 17, 73100. Briton, L. (1996). Pragmatic Mark ers in English. Grammaticalization and Discourse Functions. Berlin: Mouton de Gruyter. Chaudenson, R., Mougeon, R., & Beniak, E. (1993). Vers une approche panlectale de la variation du franais Paris: Didier Erudition. Coveney, A. (2003). Anything you can do, tu can do better: tu and vous as substitutes for indefinite on in French. Journal of Sociolinguistics, 7(2), 164191. Coveney, A. (1996). Variability in spoken French Exeter: Elm Bank Publications.
118 Coveney, A. (1997). Lapproche variationniste et la description de la grammaire du franais : le cas des interrogatives. Langue franaise, 115, 88 100. Coveney, A. (1998). Awareness of linguistic constraints on variable ne omission. Journal of French Language Studies, 8, 159188. Coveney, A. (2003) Anything you can do, tu can do better: tu and vous as substitutes for indefinite on in French. Journal of Sociolinguistics, 7(2), 164191. Coveney, A. (2004a). The alternation between l'on and on in spoken French. Journal of French Language Studies, 14, 91 112. Coveney, A. (2004b). La variation entre elles et ils en franais parl. In A. Coveney & A.M. Hintz (Eds.), Variation et francophonie (pp. 301 329). Paris: L'Harmattan. Dewaele, J. M. (1992). L'omission du ne dans deux styles oraux d'interlangue franaise. Journal of Applied Linguistics, 7, 3 17. Dewaele, J. M. (2004a). Retention or omission of the "ne" in advanced French IL: The variable effect of extralinguistic factors. Journal of Sociolinguistics, 8, 433450. Dewaele, J. M., & Regan, V. ( 2002). Matriser la norme sociolinguistique en interlangue franaise: le cas de l'omission variable de ne. Journal of French Language Studies, 12 123148. Erman, B. (2001). Pragmatic markers revisited with a focus on you know in adult and adolescent talk. Journal of Pragmatics, 33, 13371359. Ervin Tripp, S. (1972). On sociolinguistic rules: Alternation and cooccurences. In J. G. a. D. Hymes (Ed.), Directions in Sociolinguistics (pp. 213250). New York: Holt, Rinehart and Winston. Gadet, F. (1992). Vari ation et htrognit. Langues, 108, 515. Goosse, A. (1959). Les emplois modernes de l'on Zeitschrift fur Romanische Philologie, 75, 269305. Hansen, A. B. (1994). Etude du E caduc : stabilisation en cours et variations lexicales. Journal of French L anguage Studies, 4(1), 25 54. Hansen, A. B. (2000). Le E caduc interconsonnantique en tant que variable sociolinguistique --une tude en rgion parisienne. LINX, 42 (1), 4548. Howard, M. (2006). Variation in Advanced French Interlanguage: A Comparison of Three (Socio)linguistic Variables. The Canadian Modern Language Review / La revue canadienne des langues vivantes, 62 (3), 379400.
119 Hyman, E. (2004). The indefinite you English Studies, 2, 161 176. Laberge, S. (1977). Etude de la variation des pronoms sujets dfinis et indfinis dans le franais parl Montral. Unpublished Ph.D. dissertation, Universit de Montral. Laberge, S. (1980). The changing distribution of indefinite pronouns in discourse. In R. W. S. a. A. Schnukal (Ed.), Language Use and the Uses of language (pp. 7687). Washington: Georgetown University Press. Laberge, S., & Sankoff, G. (1980). Anything You Can Do In G. Sankoff (Ed.), The Social Life of Language Philadelphia: University of Pennsylvania Press. Labov, W. (1972a). Sociolinguistic Patterns Philadelphia: University of Pennsylvania Press. Labov, W. (1978). Methodology. In W. O. Dingwall (Ed.), A Survey of Linguistic Science (pp. 413497). Stamford, Conn. : Greylock Publishers. Meyerhoff, M. (2006). Introduction to Sociolinguistics. London and New York: Routledge. Milroy, L. (1980). Studying Language in the Community: The Fieldworker and the Social Network. In Language and Social Networks (pp. 4169). Baltimore: University Park Press. Mitchell, R., & Myles, F. (2004). Second Language Learning Theories : Arnold Publishers. Mougeon, R., & Beniak, E. (1991). Linguistic consequences of language contact and restriction: The case of French in Ontario. Oxford: Oxford University Press. Mougeon, R., Nadasdi, T., & Rehner, K. (2002). Etat de la recherche sur l'appropriation de la variation par les apprenants avancs du FL2 ou FLE. Acquisition et Interaction en Langue Etrangre, 17, 17 50. Mougeon, R., Rehner, K., & Nadasdi, T. (2004). The learning of spoken French variation by immersi on students fromToronto, Canada. Journal of Sociolinguistics, 8(3), 408432. Nadasdi, T., Mougeon, R., Rehner, K. (2005). Learning to speak Everyday (Canadian) French. The Canadian Modern Language Review / La revue canadienne des langues vivantes, 61(4), 543561. Neufeldt & D. B. Guralnik (Eds.) (1995), Webster's New World College Dictionary (Third Edition ed., pp. 981). New York: Macmillan. Omaggio, A. (2000). Teaching Language in Context Boston, MA: Heinle & Heinle Publishers. Paolillo, J. C. (2001). Analyzing Linguistic Variation : Statistical Models and Methods Stanford, CA: Center for the Study of Language and Information.
120 Preston, D. R., & Bayley, R. (2008). Variationist Linguistics and Second Language Acquisition. In W. R. a. T. Bhatia (Ed.), Th e new handbook of second language acquisition. Bingley: Emerald Group Publishing Ltd. Regan, V. (1995). The acquisition of sociolinguistic native speech norms: Effects of ayear abroad on second language learners of French. In B. Freed (Ed.), Second Language Acquisition in a StudyAbroad Context (pp. 245267). Amsterdam, The Netherlands: John Benjamins. Regan, V. (1996). Variation in French interlanguage: A longitudinal study of sociolinguistic competence. In R. B. a. D. R. Preston (Ed.), Second Language Ac quisition and Linguistic Variation (pp. 177201). Philadelphia, Pennsylvania: Benjamins. Regan, V. (1997). Les apprenants avancs, la lexicalisation et l'acquisition de la comptence sociolinguistique: une approche variationniste. AILE (Acquisition et Int eraction en Langue Etrangre), 9, 193210. Rehner, K., & Mougeon, R. (1999). Variation in the spoken French of immersion students: To ne or not to ne that is the sociolinguistic question. Canadian Modern Language Review, 56, 124154. Rehner, K., Mougeon, R., & Nadasdi, T. (2003). The learning of sociolinguistic variation by advanced FSL learners: The case of nous versus on in immersion French. Studies in Second LanguageAcquisition, 25, 127 156. Sandfeld, K. (1965). Syntaxe du franais contemporain. Tome 1: Les pronoms Paris: Champion. Sankoff, D. (1988). Sociolinguistics and syntactic variation. In F. Newmeyer (Ed.), Linguistics: the Cambridge survey, tome 4: Language: the sociocultural context (pp. 140161). Cambridge: Cambridge University Press. Sa nkoff, D., & Laberge, S. (1978a). Statistical dependence among successive occurrences of a variable in discourse. In D. Sankoff (Ed.), Linguistic variation: models and methods (pp. 119126). New York: Academic Press. Sankoff, D., Tagliamonte, S., & Smith, E. (2005). Goldvarb X: A variable rule application for Macintosh and Windows: Department of Linguistics, University of Toronto. Sankoff, G., Thibault, P., Naomi, N., Blondeau, H., Fonollosa, M.O., & Gagnon, L. (1997). Variation in the use of discourse m arkers in a language contact situation. LanguageVariation and Change, 9, 191217. Sankoff, G., & Vincent, D. (1980). The Productive Use of ne in Spoken Montral French. In G. Sankoff (Ed.), The social life of language Pennsylvania: University of Pennsylvania Press.
121 Sax, K. (1999). Acquisition of stylistic variation by American learners of French:The case of ne deletion, New Ways of AnalyzingVariation. Toronto, Canada. Sax, K. (2000). Acquisition of stylistic variation byAmerican learners of French: /l/ e lision in the subject pronouns il and ils., Second Language Research Forum Indiana University, Bloomington, Indiana. Selinker, L. (1972). Interlanguage. International Review of Applied Linguistics, 10, 209241. Staels, E. H. (2004). The indefinite you English Studies, 2 161176. Stansfield, C. W., & Kenyon, D. (1996). Simulated Oral Proficiency Interviews: An Update. In ERIC Digest (Vol. May 1996). Washington, DC: Center for Applied Linguistics. Tagliamonte, S. (2006). Analysing Sociolinguistic Variation Cambridge: Cambridge University Press. Tarone, E. (2007). Sociolinguistic Approaches to Second Language Acquisition Research 19972007. The Modern Language Journal, 91, 837848. Thibault, P. (1991). La langue en mouvement: Simplification, rgularis ation, restructuration. LINX, 25 7992. Thibault, P., & Develuy, M. (1989). Quelques traces du passage du temps dans le parler des Montralais, 19711984. LanguageVariation and Change, 1, 1945. Thibault, P., & Vincent, D. (1990). Un corpus de franais parl. Montral 84: historique, mthodes et perspectives de recherche. Qubec: Dpartement de langues et linguistique, Universit de Laval. Thomas, A. (2002). La variation phontique en franais langue seconde au niveau universitaire avanc. Acquisition e t Interaction en Langue Etrangre (17), 101121. Uritescu, D., Mougeon, R., Rehner, K., & Nadasdi, T. (2004). Acquisition of the internal and external constraints of variable schwa deletion by French immersion students. IRAL, 42 349364.
122 BIOGRAPHICAL SKETCH Jingya Zhong received her B.A. degree in French from Xiamen University in China and has been an exchange student in University of Franche Comte in France. She has been pursuing her M.A. degree in French and Linguistics and she has been teaching Fre nch at the University of Florida.