. u: . 2:. 9 £ ‘9. H . - z - mafh‘fil'mmi’sg- 5 )si s;‘3}.é§x! .5 IV. 573 :1 .lvl -- "Iota .\ muss l llllllulllullll‘llllllll MICHIGAN TATE “w “W 3 1293 01410 186 l This is to certify that the dissertation entitled INTERLANGUAGE VARIABILITY OF THE SPANISH SUBJUNCTIVE MOOD presented by Margaret Lubbers Quesada has been accepted towards fulfillment of the requirements for Ph.D. degree in Linguistics firm/W Major professor Date [AV Ail/ff; MS U i: an Affirmative Action/Equal Opportunity Institution 042771 LIBRARY Michigan State Universlty PLACE IN RETURN BOX to remove thle checkout from your record. TO AVOID FINES return on or bdore date due. DA'l'E DUE DATE DUE DATE DUE $1 A 00—--.'-o.. II I l MSU le An Affirmative Action/Equal Opportunity lnetltwon Mt ABSTRACT INTERLANGUAGE VARIABILITY OF THE SPANISH SUBJUNCTIVE MOOD By Margaret Lubbers Quesada This dissertation is a study of the interlanguage variability' of a group of jyoung' English-speaking' adults studying Spanish as a second language in. Mexico. Specifically, it analyzes how morphological and syntactic linguistic forms interact with semantic and pragmatic features .to constrain the use of the subjunctive mood in learners' speech. Two sets of oral interviews of the learners were recorded, transcribed. and analyzed for the study; The frameword adopted for analysis is that of Huebner (1985) who claims that it is important to decipher in what ways semantic and discourse-pragmatic functions are encoded in interlanguage Systems. Furthermore, this study draws upon work in sYl'lchronic descriptions of the Spanish subjunctive, most notably Lavandera (1982, 1933) and Lunn (1989a. 1989b). which claim that syntactic, semantic and pragmatic factors interact to constrain native speakers' use of the subjunctive. Finally, in order to define how different features of the subjunctive interact to inhibit or favor subjunctive use, I draw upon the work in variation theory of Adamson (1988, 1990) and set up a prototype schema of the Spanish subjunctive that appears to favor subjunctive use among learners. It was found that there is considerable variability in these learners' use of the present subjunctive but that this variability is systematic. .Analysis of the 'variability revealed that learners' use of the present subjunctive is constrained and caused by an interaction among syntactic, morphological and semantic features. It is suggested that for the subjunctive, learners construct a prototype schema which conforms to only a limited facet of the native speaker's schema of this structure. In addition, this study found that adult learners, like child. L1 learners, acquire certain features of the Spanish mood system.before others. But unlike -children, these adults had no difficulty producing complex sentences. The analysis lends insight into both the nature of the subjunctive mood in Spanish and the acquisition processes of complex structures such as the subjunctive. Copyright by MARGARET LUBBERS QUESADA 1995 ," (3. hi. I“. ACKNOWLEDGMENTS This work would have never been completed without the help, guidance and support of so many throughout the years. First of all, I acknowledge and thank Dr. Grover Hudson, my dissertation director, who always remained cheerful and encouraging even after reading numerous, tedious drafts and for giving clear, concrete advice for helping to shape this work into 'more like a dissertation.‘ I deeply appreciate Dr. .Ruth Brend, my advisor from the beginning of my studies in linguistics and dissertation director until her retirement from the department (but not from linguistics!) Without her .encouragement at difficult and critical moments, I would never have gotten to this point. I thank Dr. Patricia Lunn for getting me interested and excited about difficult areas of Spanish (and for also good—naturedly reading numerous drafts). I thank Dr. Julia Falk, who first introduced me to the fascinating study of language acquisition and who insisted that I be, and helped me to become, a better scholar. To all of them I am grateful for their efforts to make this a scholarly work. Of course, any errors or oversights that remain in the work are my sole responsibility. I thank Drs. George Peters and Patricia Pausell for channeling financial support to me so that I could stay in V Michigan and finish writing. I also thank Dr. Dennis Preston for lending books and giving me ideas, Mr. Gary Cook of the English Language Center for helping with the statistical analysis, and Dr. Scott Whiteford of the Center for Latin American and Caribbean Studies for encouragement and office space. From Mexico, I am forever indebted to M. en I. José Alfredo Zepeda, Rector of the Universidad Autonoma de Querétaro, for his valuable support and enthusiasm for this project from its beginning to its end. I also thank numerous other Mexican colleagues and friends, too many to number here, who provided me with help and support to carry out this project. I thank my students from the 'Licenciatura' in Modern Languages who collaborated in the project in many ways. I thank Mrs. Guadalupe Castorena and other colleagues of the Centro de Estudios Lingfiisticos y Literarios for their encouragement and help. Finally, I would not have ever completed this project without the love, patience and sacrifices of my family. To my mother and father, Jean and John Bloemendaal, I am eternally grateful. They gave me the confidence to persevere. For Sergio, Sarah, Sergio José and Amanda there are no words to express what they mean to me. I owe everything to them and they are everything to me. vi TABLE OF CONTENTS Chapter One: Introduction Chapter Two: A Review of Studies on Variability in Interlanguage Introduction The Classification of Interlanguage Variability Studies Interlanguage Studies and Attention to Form Interlanguage and Linguistic Environment Interlanguage and the Relationship between Form.and Meaning and/or Function variation and Prototype Theory Conclusion Chapter Three: Traditional and Pedagogical Descriptions w h WWW UNI-4 of the Spanish Mood System Introduction Traditional Grammars A Textbook Description of the Spanish Subjunctive Conclusion Chapter Four: Syntactic, Semantic and Pragmatic bub NH 4 .3. Explanations of the Spanish Mood System Introduction Semantic Studies 4.2.1. Bolinger 4.2.2. Rivero 4.2.3. Terrell and Hooper/Terrell 4.2.4. Garcia and Terrell and Lantolf 4.2.5. Goldin and Bell 4.2.6. Takagaki and Reider Semantic/Pragmatic Studies 4.3.1. Lavandera 4.3.2. Lunn 4.3.3. Krakusin and Cedeno 4.3.4. Mejias-Bikandi vii 15 15 17 29 34 39 47 49 52 52 54 60 71 73 73 74 74 77 80 85 88 92 98 98 102 109 110 4. 4. Conclusions Chapter Five: The Acquisition of the Spanish UlU'IU'I WNH Subjunctive Introduction Review and Discussion of the Studies Conclusion Chapter Six: Data Collection and Methods of Analysis mmmmmmmm (DQQUIDUJNH Introduction Data Collection Subjects Setting Language Program in Queretaro Interviews Journals Methods of Analysis Chapter Seven: Results and Discussion 7. 7 7 7. 7 1. 2. .3. .5. Introduction Error Analysis and Grouping According to Syntactic Frame Subjunctive-Marking Vs. Other Verb Formr Marking 7.3.1. Interview One 7.3.2. Indicative Clause Analysis and Chi-Square Test 7.3.3. Discussion 7.3.4. Interview Two 7.3.5. Indicative Clause Analysis and Chi-Square Test 7. .6. Discussion Prototype Features of Syntactic Frames and Complement Verbs 7.4.1. Interview One 7.4.2. Interview Two Conclusions Chapter Eight: Implications for Teaching and Future (D0 oooooo who: NH Research Summary Insights into the Nature of the Subjunctive Mood and Acquisition Implications for Teaching Directions for Future Research Conclusion viii 112 116 116 117 125 127 127 127 128 129 130 132 133 134 138 138 141 146 146 150 154 167 172 174 181 181 186 193 197 197 200 202 205 207 Appendix A: The Interview Formats Appendix B: Occasions Requiring or Allowing Present Subjunctive by Syntactic Frame Appendix C: Occasions Requiring or Allowing Present Indicative by Syntactic Frame Appendix D: Prototype Features List of References ix 208 211 229 249 264 Table Table Table Table Table Table Table Table LIST OF TABLES Subjunctive (S) and Other (0) Marking According to Syntactic Frame - Interview One Subjunctive (S) and Other (0) Marking According to Syntactic Frame - Interview Two Forms Used for Present Subjunctive by Syntactic Frame - Interview One Forms Used for Present Indicative by Syntactic Frame - Interview One Tabulated Statistics of Chi-Square Test - Interview One Forme Used_for Present Subjunctive by Syntactic Frame - Interview Two Forms Used for Present Indicative by Syntactic Frame - Interview Two Tabulated Statistics of Chi-Square Test - Interview Two 142 143 147 151 155 168 173 175 Figure 1. Figure 2. LI ST OF FIGURES Cross Products Chart: Four features yielding sixteen environments for subjunctive marking - Interview One 184 Cross Products Chart: Three features yielding eight environments for subjunctive marking — Interview Two 188 xi 3‘5 CA ”v CHAPTER ONE : INTRODUCTION This dissertation is a study of the interlanguage variability of a group of young English—speaking adults studying Spanish as a second language in Mexico. It was found that there is considerable variability in these learners' use of the present subjunctive mood but that this variability is systematic. Analysis of this variability shows that the choices learners make to either mark or not mark for the present subjunctive mood are constrained and caused by an interaction between multiple linguistic features of the target language structure, (including syntax, morphology and semantics), and universal learning processes (such as L1 transfer, overgeneralization and hypothesis formation). It is suggested that variability in the use of the subjunctive and . other verb forms may play an important role in the process of the acquisition of the Spanish subjunctive mood. It is seen that for structures which are complex both syntactically and semantically, such as the subjunctive, learners construct prototype schema which include very specific features of the structure. This prototype schema conforms to only a limited facet of the native speaker's schema of the same structure. As acquisition develops, the schema, which includes syntactic, morphological, semantic, and, late in the acquisition process, pragmatic features, drops some features and adds others as it adjusts to the native speaker schema. By identifying what linguistic 1 feazur' .. .u U an" C S A F pan\a‘ i. Hub-— .5 “ 5U ’19 beds, ‘4 5“: I.‘ v J. the L“ n . C"- .‘N‘.‘ 5" O“ '- 'n ..~Se ‘I i i. 'l i a 'e e 4e“, 2 features, or combination of features, inhibit or promote subjunctive marking, we can come to an understanding of what causes interlanguage variability. In addition, this study found similarities and differences in the acquisition of the Spanish subjunctive between child L1 and adult L2 learners. It has been suggested that children's late acquisition of the subjunctive may be due to the fact that they acquire complex sentences late. It was found in the present study that these adult learners had no difficulty producing complex sentences and that subordination was not a feature promoting or inhibiting subjunctive use. .However, both child L1 and adult L2 learners acquire certain features of the Spanish mood system before others. The analysis of these features and how they affect the variability in the use of the subjunctive lends insights into the nature of the Spanish mood system, That variability exists in second language learner's interlanguage speech is undisputed. What is disputed is whether or not a study of interlanguage variation is necessary or interesting or can help us to understand anything about language in general and language acquisition processes in particular (Gregg 1990; Ellis 1990; Tarone 1990). Even among those investigating interlanguage variation, there is a wide variety of approaches and purposes. Early variability studies in second language acquisition have two characteristics in common. The first is that they use as their models variability studies in adult native 3 language or sociolinguistic studies. However, these studies maintain that variability is due to, or caused by social contextual factors such as setting, topic, interlocutor, etc.' These external factors influence the degree to which the speaker focuses on the way that s/he speaks, producing a more or less formal or careful style of speech. The present study adopts the views of Hulstijn (1989) and Preston (1989) that second language learners, especially in the beginning stages, do not have sufficient linguistic competence in the second language to be sensitive to many relevant external social contextual factors; specifically, they do not have the competence to manipulate a variety of styles ranging from informal to formal. Variability is therefore, due to other factors. The second characteristic of early studies in variability in second language acquisition was a focus primarily on phonological variation, although recent studies (e.g., Young 1988, 1989, 1991, 1993) analyzing phonological variation take into account multiple causes (e.g. phonological and syntactic environment, function of the morpheme and of the clause in which the structure is required, and external social factors). The present research is the study of the variability in the use of a structure which is complex syntactically, morphologically and semantically. WOlfram.(1989) claims that several levels of language organization need to be examined in order to explain systematic variation in such structures as tense marking. Tarone (1988:69) mentions that 'certainly any ,_.. e 0-. w o -04 R‘v- H 5‘44 sus- p) 0 1 (T) . rug; “an‘ . "?A C“ e 4 study seeking to establish constraints on variation at any one level must bear in mind the possibility of such interaction with constraints at other levels.‘ By investigating both linguistic form and how it interacts with meaning/function we can better understand the acquisition process. The Spanish subjunctive has been difficult to analyze perhaps because it is so complex. It occurs mainly in subordinate clauses and within certain syntactic frames; there are different forms for regular and irregular verbs; as with all Spanish verbs, there must be concordance in person and number between the subject and the verb; and its use is often dictated by subtle meaning differences controlled by the speaker. In addition, there is often considerable variability in its use among native speakers from speaker to speaker and across dialects. Traditional and pedagogical grammars do not take these factors into account when describing the Spanish mood system“ They have had difficulty in accounting for its. use mostly because they fail to point out the interaction of syntactic, morphological and semantic complexity. Synchronic/theoretical studies of the Spanish subjunctive have attempted to describe and explain the use of the structure taking into account these complexities, but in the 1990's, most researchers are still refining Bolinger's 1968 work, which launched the semantically-based analyses of recent years. Because of the complexity of this structure, it is not surprising that the subjunctive mood is considered by Spanish ‘ ' teacuers a, learn or a Son 53:62 1 acaisiticr Stckes 1955 is evidenCE ,angu‘ge a: second lax; ::e 5'43““ but the CIE came late Lee 1987)’ ..a. in: cat .qai‘fe’q C :55 a var“ language 9" features 0 in \ mu" ' 15mm 5 teachers and learners alike to be a difficult structure to learn or acquire. Researchers confirm that it is a late— learned structure in both first (Blake 1991; Floyd 1990; Gonzalez 1975; and Gili y Gaya 1972) and second. language acquisition (Collentine 1995; Al-Kasey 1993; Studerus 1992; Stokes 1988, 1990; Terrell, et a1 1987; and Lee 1987). There is evidence that both children learning Spanish as their first language and children and adults learning Spanish as their second language use 'chunks' or memorized formulas containing the subjunctive (e.g. 'Que te vaya bien' or 'Ojalé llueva'). but the creative use of the subjunctive in subordinate clauses ‘comes late in the language acquisition process (Floyd 1990; Lee 1987), especially for certain uses. Adults will use complex sentences containing embedded clauses which children's speech does not display. The present research found that adults have no difficultly producing complex sentences where the indicative is required, but where the subjunctive mood is required, or normally used by native speakers, these learners use a variety of forms. In addition, it was found that second language acquisition of the more subtle semantic and pragmatic features of the subjunctive mirror child language acquisition in some ways. The present research project is a study of variability in the use of both the subjunctive and alternative forms. Although there is a great deal of variability in this use, it is mostly systematic and constrained by multiple linguistic features and processing strategies. A detailed analysis of this variability and the features and strategies 6 that cause it in the use of Spanish by young adult English speakers can give us an increased understanding of this systematicity and offer insights into both the nature of the subjunctive mood and the acquisition processes of complex structures such as the subjunctive. The data that are the object of the analysis in this study are two sets of recorded oral interviews of sixteen American university students studying Spanish in an intensive twelve-week language, culture, and arts program in Mexico. To analyze the data, the study adopts a function-form framework which has been assumed in studies of the second language acquisition of English (Huebner 1983, 1985; Tarone 1985a, 1985b, 1989; Schachter 1986; Bardovi-Harlig 1992; and Young 1988, 1993). Although these researchers have defined the framework in slightly different ways, they all seek to link the process of acquisition of the linguistic structure under study with its meaning and pragmatic function in the learners' speech. The specific framework this study adopts is that of Huebner, who claims that in the analysis of interlanguage variation it is important to decipher in what ways semantic and discourse-pragmatic functions are encoded and to what extent these relationships are systematic (Huebner 1985: 155). In addition, like the Huebner study, this study will analyze interlanguage to see not only when learners do not use the target language form but to understand when and why they use the forms which they do use. 7 Furthermore, this study draws upon work in synchronic theoretical descriptions of the Spanish subjunctive which claim that syntactic, semantic and pragmatic factors interact to constrain native speakers' use of the subjunctive (Terrell and Hooper 1974; Terrell 1976; Lavandera 1982, 1983; Lunn 1989a, 1989b; and Mejias-Bikandi 1994). Finally, in order to attempt to define and explain how the syntactic, morphological and semantic/pragmatic features of the subjunctive interact to inhibit or favor subjunctive use, I draw upon the work in variation theory of Adamson (1988, 1989 and 1990) and set up a prototype schema of the .Spanish subjunctive that appears to favor subjunctive use. The absence of one or more of the features of the schema inhibits subjunctive use. The prototype schema identifies certain features which promote subjunctive usage in the first set of interviews but have no influence in the second set, and which appear not to influence subjunctive usage in the first interview but strongly influence this in the second. Specifically, this study attempts to describe how morphological and syntactic linguistic forms interact with semantic/pragmatic features to constrain the acquisition and use of the Spanish subjunctive by one group of learners. The study addresses the following questions: 1. Do English-speaking adults learning Spanish use the subjunctive in the early stages or not? Is there variation in their use? If so, is this variation systematic? 2.. Is there a tendency for the subjunctive to be used in certain syntactic frames and not in others? What are the features of these syntactic frames? Are certain morphological forms more likely to be candidates for subjunctive-marking than others? What features do these forms share? 3. Is there a tendency for the subjunctive to be used for certain meanings and functions and not for others? What features do these meanings/functions have? 44. When the subjunctive is not used, which verb forms are used? What is the use of these forms due to: interference from English, overgeneralization of target language rules, universal processing strategies? 5. Is the use of the subjunctive correlated to the use of complex sentences, as in child L1 acquisition? Or are adult L2 learners capable of producing complex sentences? The following hypotheses are set up in order to attempt to answer these questions: H1: Young English—speaking adults do use the subjunctive in the upper-beginning/lower- intermediate stages of language learning. This use, although limited to few linguistic environments, is systematically variable. H2: The subjunctive will only be used in linguistic environments where (1) the syntactic frame is overtly produced, (2) the morphology is highly salient and (3) the meaning/function of the utterance is to signal either a possible future action or event or a desire on the part of the subject of the main clause. H3: The subjunctive will not be used in linguistic enviromnents which are lacking either one or more of these features (listed in H2). H4: There will be variability in the use of alternative forms used when the subjunctive is not used. H5: Young English-speaking adults in the upper- beginning/lower-intermediate stages of language learning do produce complex sentences and there is no correlation between this production and 10 subjunctive use. In order to address the hypotheses, using the data provided by the thirty—two interviews, several methods of analysis were employed. To ascertain the amount of subjunctive use and non-use, an error analysis of the data was performed. It was seen that in the first interview the subjunctive was used very seldom but increased dramatically during the second interview. Following the error analysis, the utterances in which the subjunctive is not used were analyzed in order to ascertain _the forms which the subjects use instead of the subjunctive. As was expected, indicative forms were those most frequently used, but they were not the only ones used. variation in the use of infinitives, gerunds, past preterite, future, and invented forms for present subjunctive all indicate different strategies and levels of processing by these subjects. From this analysis we can see structures which closely resemble L1 (possibly due to L1 transfer), structures which resemble neither L1 nor L2 (or perhaps resemble both), and structures which are very much 'L2—like'. It appears that in very early stages of second language acquisition, the transfer of structures from L1 to L2 is a strategy used for communicating. Later, universal processing strategies such as overgeneralization and hypothesis testing of the L2 rules go on. In order to identify the features mentioned in Hypotheses 11 Two and Three and to determine their influence on the usage of the subjunctive, a prototype schema of the type outlined by Adamson is set up and displayed on a cross products chart. From. this kind. of chart, two kinds of information are immediately available: (1) features or a combination of features which favor or inhibit subjunctive marking, and (2) areas of systematic or non-systematic variability. As a result of the analysis, it is seen that the use of the subjunctive is very limited in the first set of interviews. The subjunctive is used when the syntactic environment contains overtly produced syntactic frames cuando and querer que and salient subjunctive morphology (as in irregular verbs), and.when the function of the subjunctive is to express either a possible future action or event or a desire on the part of the speaker. Other overtly produced syntactic frames are not candidates for subjunctive usage by the subjects under study. In addition, when cuando and querer are followed by complement clauses containing regular verbs, these verbs are not marked as often for subjunctive mood. There is no use of subjunctive when the meaning of the syntactic frame is other than possible future action or event, or desire. Furthermore, in these data there is evidence for systematic variation; there is a great deal of variability in the use of fonms other than the subjunctive for three of the syntactic frames. In order to test Hypothesis Five and to see if a similar . .. a nun itweb V en‘vrxna; Jun“ ' e s a. Opel. : 12 pattern of variability is seen for indicative clauses, all of the complex sentences which allow or require present indicative are analyzed in the same way the subjunctive clauses are. The same pattern of variability is not seen for the indicative mood. It is seen that the present indicative has been acquired by these subjects. In addition, the abundance of subordinate clauses in which both the present indicative and subjunctive are allowed or required provides evidence that adult learners have no difficulty producing complex sentences. The data from the second set of interviews showed a _simdlar pattern of use. However, here the production of an overt syntactic frame appears to neither promote nor inhibit subjunctive use; salient verb morphology does. The syntactic frames expressing futurity and/or desire are the ones that show the greatest amount of subjunctive use in their complement verbs and also the greatest amount of variability in the use of other verb forms. For both sets of data, then, it appears that increased subjunctive usage is accompanied by increased variability or vice versa. An overview of the dissertation is as follows. Chapter Two is a review of the studies of interlanguage variability which concludes that these kinds of studies are necessary to our understanding of SLA acquisition and that those studies which have come the closest to our understanding of this process are those 'which. have adopted the function-form approach, specifically those which have studied the Fe Co "1 5. 2‘ she 13 interaction of form, meaning and function. It is concluded that even those studies whidh did not adopt this approach revealed upon later analysis that variability in the IL could only be accounted for by analyzing the meaning and function of the forms under study and how they were used by the subjects. Chapter Three is a review and discussion of traditional and textbook descriptions of the subjunctive mood. It is concluded that these descriptions neither capture the nature of the subjunctive mood nor do texts help students of Spanish as a second language because of their attempt to describe the subjunctive as syntactically as possible. Chapter Four details the synchronic/theoretical studies of the Spanish subjunctive. These studies lead to a conclusion supported by the present study: the nature of the Spanish subjunctive cannot be accounted for any other way than by taking into account its semantic and pragmatic features. Chapter Five discusses studies of the acquisition of the Spanish subjunctive by children and adults. It was seen that children have difficulties acquiring the subjunctive in the same way the adults in this study did. Again, the conclusion drawn from the discussion of these studies is that there is a need to analyze data from.a function-form.approach. Chapter Six presents the research methodology including the subjects, setting, data collection and methods of analysis of the present study. It is pointed out that although the studies of interlanguage variability reviewed in Chapter Two were based on the analysis of data which was collected in many ii A F CV u a... (16.5: the l (I) I I 'L) 't 1 Stud. concj Fina: 14 different settings, the present research is based on the analysis of data collected. where the learners are :more concerned with meaning than with grammatical accuracy. The data are analyzed using several methods of analysis, including error analysis, analysis of the variability in the use of verb forms, a chi-square test and analysis of a prototype schema. Chapter Seven presents and discusses the results of this research in terms of how these confirm or partially confirm the hypotheses set forth in Chapter One and how these results support or reject the claims made by researchers in the studies discussed in Chapters Two, Three, Four and Five. Chapter Eight summarizes the results and sets forth some concluding remarks. Implications for the teaching of languages in general and Spanish in particular are discussed. Finally, directions for future research are suggested. L ‘45 Y" CHAPTER 'IWO: A REVIEW OF STUDIES ON VARIABILITY IN INTERLANGUAGE 2.1. Introduction In this chapter we will examine a number of studies which are relevant to the present study in the following ways. First, the claim that variation studies are valid and even necessary to our understanding of the language acquisition process is defended. In a review of how variability studies have traditionally been classified, it is pointed out that these studies are not merely performance studies, but rather, when they focus on understanding what characteristics or .features of the linguistic system constrain or cause variability, they contribute to competence theories and help to explain how certain structures are acquired. In addition, it is concluded in this section that regardless of how studies .are classified, variability in interlanguage will only be understood by taking into account a variety of both linguistic and non-linguistic factors. Secondly, it should be recalled that Hypotheses Two and Three of the present research state that subjunctive marking for the learners in this study is caused by an interaction of syntactic, morphological and semantic factors related to the structure. In the subsequent section of this chapter it is made evident that the majoridy of the earlier studies are inadequate because they claim.that the cause of interlanguage variability is external, social/contextual factors - factors 15 r1 16 which second language learners may not be focusing on in early stages of acquisition. Later studies analyzed the linguistic system itself in order to understand how features of the structure under study caused variability. These studies brought us closer to an understanding of how the linguistic system constrains variability but many researchers began to realize that interlanguage variability was due to multiple causes. More recent studies have linked the causes of variability with the relationship of form, syntactic and morphological, and meaning and/or function. The claimtmade in Hypothesis Four of the present research is that there is .variability in the use of alternative forms. This section concludes that although interlanguage studies have increasingly offered more interesting hypotheses about the second language acquisition process and the linguistic constraints causing interlanguage variabity, there is a need to analyze not just when and why learners do not use the language structure under study, but also to account for what alternative structures they produce in place of that structure. Finally, we discuss how variation theory and prototype schema provide a tool which can help to identify which linguistic environments or features of linguistic structures or combinations of features, such as those listed in Hypothesis Two of the present study, cause variability. n? (I) or: 1 Rah \ "Ll 17 2.2. The Classification of Interlanguage variability Studies It has been a question in the literature whether studies which examine the variability in the interlanguage (IL) of second language (L2) learners are valid and important and even necessary to our understanding of the second language acquisition (SLA) process. There are those who believe that an examination, description and explanation of the variability which L2 learner language exhibits are essential to a theory of SLA (Tarone 1982, 1983, 1985a, 1985b; Ellis 1985, 1988, 1990; Crookes 1989; and Young 1988, 1989). Others admit that at least it is a question deserving attention in research "(Tarone 1989, 1990; Huebner 1985; Beebe 1988; Preston 1989, 1993; Hulstijn 1989; Welfran11989; Schachter 1986; and Larsen- Freeman. and. Long 1991). Recently, Gregg (1989, 1990), however, has argued that the study of IL variability is the study of perfonmance and has no place in a theory of SLA which should deal only with a description and explanation of a language learner's competence. And, although they do not openly refute the value of the study of variability in IL, others choose to study only learners' intuitions. The latter studies attempt to tap learners' competence, mostly to ascertain adult L2 learners' accessibility to the principles of Universal Grammar and the role that parameter setting has on the process of SLA (Schachter 1989; Flynn 1989; Liceras 1986, 1989; and White 1987, 1989), or to ascertain the role of typological, processing and other language universals in SLA 18 (Gass 1984, 1989; Rutherford 1982, 1989; Eckman 1984; and Bardovi-Harlig 1987). Gregg (1989, 1990), the strongest critic of variation studies, distinguishes between two kinds of studies in SLA: those which examine linguistic ability or performance; and those which examine linguistic knowledge or competence. He feels there is a great deal of confusion in the literature between competence and performance, and using words like '1inguistic ability' only clouds the picture. He claims that 'humans have knowledge of language quite apart from their ability to use that knowledge... [and although] our linguistic .ability rests primarily...on [that] knowledge...it is not identical to that knowledge' (1989:19). He maintains that the domain of SLA studies, like that of first language acquisition (FLA) studies, should be language knowledge, either acquired or innate, and not speech (which reflects linguistic ability). Studying the linguistic behavior of learners, although interesting in and of itself, should not be the goal of SLA research (1989:18). Based on this distinction, he rejects the validity of studies on variation in IL, and states that 'if we are careful to establish the domain of a theory of second language acquisition so that it is confined to the acquisition of linguistic competence, then we will not be compelled to account for those data on variability as far as that theory is concerned' (1989:22). Gregg feels the goal of SLA research should be 'to 19 explain the.acaniaition of linguistic knowledge' (1990: 374). He states that it is important to distinguish explanation from description, acquisition from use, competence from behavior and that the majority of studies in variation in SLA to date have focused on the description of the use or production of L2's interlanguage. Therefore, he feels these studies cannot be seen as contributing to a theory of SLA (1990:365). What Gregg ultimately is arguing for is the superiority of using a Universal Grammar framework as a base for the study of SLA (1989, 1990). He argues that such a perspective can give clarity, explanatory power and a general guiding purpose 'to SLA research and avoid such pitfalls such as using ad hoc, confusing or vague terminology. He steadfastly adheres to the idea that a theory of SLA must explain the grammatical competence of the L2 learner and not how the learner uses this knowledge in production. Gregg's insistence on clarity in the use of terminology and the distinction he makes between competence and performance studies are valid. However, he sees the support for one kind of study as the necessary rejection of the other. It is important to determine just what is the 'linguistic knowledge' that Gregg is talking about. All linguistic knowledge is either acquired or innate. What other kind of knowledge is there? What Gregg seems to want to do is to separate syntactic knowledge from other kinds of linguistic knowledge, such as morphological, semantic, pragmatic, sociolinguistic, etc. But in part the question is whether or q R Nu, ‘51.:‘ E6 st Pb . "‘1‘ '- 5-. “l s) 20 not these other kinds of linguistic knowledge form part of a speaker's competence. The present research shows that syntactic, morphological, semantic and pragmatic kinds of knowledge all form.part of a speaker's linguistic competence, and that these kinds of knowledge interact to affect how structures are acquired. Ellis, in his response to Gregg's criticisms of his studies of interlanguage variation, notes that it is important to be clear in this distinction of terms, and he welcomes the dialogue between 'those, like Chomsky, who believe that language can best be explained in terms of abstract formal .properties which exist independently of the way it is used in communication... (and) those, like Halliday, who believe that the formal properties of language reflect and derive from.its communicative uses' (1990: 384). He suggests that the style of research one elects will directly depend upon one's view of language and the research questions that one is interested in. He states: A variabilist theory, I would argue, is of greater relevance than the kind of competence theory Gregg advocates when it comes to issues related to the study of classroom L2 and language pedagogy. Such a theory however, may be less relevant if one's aim: is to build an abstract model of language (386). Ellis is interested in understanding how learners develop the ability to use competence in different kinds of language situations. He also believes that by analyzing what learners 21 do systematically we can infer a great deal about their competence, both grammatical and communicative. In this sense, output that displays variability, both systematic and unsystematic, can give us clues as to the competence and developing competence of language learners. Ultimately, Ellis' stance is that SLA research needs to address both linguistic form and how it interrelates with function in order to fully explain the SLA process. Preston (1993) , while advocating more communication among different subfields of linguistics and among researchers working within different theoretical paradigms, believes that .if second language acquisition researchers focus only on the issues and goals of what he calls the 'dominant paradigm' , or the Universal Grammar research program, then, they 'will miss many linguistic boats' (158) . SLA researchers have been interested in far more issues and linguistic structures than UG's narrow list of syntactic structures. He mentions that his 'concern is that UG may occupy too much of our time and may prove to be unhelpful due to its own limited research focus' (156). However, he does conclude that researchers in both fields could benefit from each other's work. Bialystok (1990) also sees the two different focuses as Complementing each other, rather than competing for validity, though she suggests classifying such studies in a slightly different way. She explains that theories of SLA have traditionally been distinguished by whether they have as their goals the description of knowledge of rules for linguistic n‘n‘ NEE: al‘venu' a; UL LL “.931 ‘ (3)—1:an 22 structure, as in competence theories, or descriptions of language use, as in processing models. This distinction comes from Chomsky, who has given only competence theories his attention. Like Ellis, Bialystok believes that the distinction between the two types of theory is important because 'commitment to one or the other . . . has important consequences for the type of questions pursued, research conducted, and evidence accumulated' (1990: 636). However, like Preston, she recognizes that both types of studies are necessary in order to produce a 'coherent description for second language (acquisition' (637). The distinction. as .traditionally viewed, she points out, carries with it a value judgement in terms of relative merit and this is unproductive. What is needed is a reclassification of the distinction in objective terms so that both approaches can become 'tenable ’avenues of study' (637). Based on the traditional distinction, the study of the IL of L2 learners and descriptions of the variability inherent in them.would be regarded as the domain of a processing theory. However, Bialystok does not accept this traditional view. She maintains that the study of the variability in an L2 learner's IL can contribute to a description of the competence of that learner's system. She states: The exclusion of incomplete knowledge and variability in language use from.the jurisdiction of competence theories... does not necessarily follow from. a stricter interpretation of competence . First , there is an important 23 difference between linguistic knowledge that is static and linguistic knowledge that is stable. The claim (from the processing point of view) is not that interlanguage is static, but that it is stable. At a given point in time, language learners have structured representations of their knowledge of the language and these representations are used systematically to produce utterances... These premises form the basis for a competence theory at the centre of which is some construct that may be described as interlanguage (638). She goes on to mention that there are two kinds of variability manifested in IL. The first is synchronic, which is the alternation in the use of forms at one point in time due to social and contextual features of the situation. This 'type of variability is the domain of processing theories or, perhaps, even of production models. Many current studies on variability have taken this approach using sociolinguistic models such as Labov's to explain variability in IL (Tarone 1982, 1983; Ellis 1985). The second kind of variability is diachronic, which is the gradual change over time in the learner's use of certain forms and which shows that the learner's system is evolving. This type of variability, although studied by examining the performance of the learner, ultimately needs to be explained by a competence theory. She explains: . . .the criteria for determining whether a model of second language acquisition is providing an explanation at the level of a competence theory or a processing theory are vague and contentious. [C]ompetence theories are overly restricted by excluding incomplete knowledge systems and variability. Consequently, traditional distinctions 24 in the literature in which theories are classified in this way are likely to be incorrect (638). She rejects the notion that production, which is the use of the knowledge of mental structures (consisting of rules and representations), be relegated to the domain of performance theories. One interpretation of this distinction, such as Gregg's, is that production, because it may not describe the linguistic representations possessed by the ideal speaker/bearer, is uninteresting and unimportant as the object of study. However, Bialystok naintains, the use of such knowledge is part of a speaker's pragmatic competence, which 'along with grammatical competence, constitutes the realm of competence theories. She states that even: Chomsky accepts, at least in principle, the importance of some forms of production data and the potential relevance of processing models, at least to the extent that they are descriptions of pragmatic competence (641). Finally, Bialystok suggests categorizing theories, and the contribution that studies make to these theories, based on other criteria. She suggests that competence theories be characterized first, by their attempt at representing mental structures and, secondly, by how they describe the way these structures are used under optimal conditions. Although she admits that the functioning of the system is an idealization and may not accurately describe actual performance, she 25 maintains that such an idealization is necessary in order 'to create a coherent model' (642). She cites Macnamara, who 'argues that theories need to be idealizations in order to have productive consequences' (643). From this it follows that although performance may not represent the grammars described in competence models because of production limitations, in no way should this limitation prevent descriptions of performance from.being idealizations. It may be that performance can be seen as an actualization of intuitions on the part of speakers. Performance or processing theories should be _characterized by neutrality with regard to the nature of the structure of the mental representations underlying performance and by their synchronic descriptions of performance. By these criteria, Tarone's variable Competence Model, and any other sociolinguistic model of variability, fall into the realm.of a processing model which explains variability at one point in time due to external, contextual factors: attention to form, demand of task, etc. Any description of the development of the competence of linguistic structures, however, would fall into the realm.of competence theories. Preston (1989, 1993) advocates a sociolinguistic model for analyzing variation in interlanguage, but at the same time, admits that social factors alone cannot account for this ‘variability. He mentions that in as much as '...a developing interlanguage is a natural language, it will require the sorts 0f descriptive apparatuses made available by a variationist .- u U1 ‘ V 4. Stir, late en a" . 26 grammar' (1989:198), but that the factors involved in native speaker (NS) variation are not necessarily the same for a non- native speaker (NNS). Most importantly he suggests that an exchange of knowledge from sociolinguistics and second language acquisition would be beneficial to both fields. Specifically, he points out that SLA would benefit from some of the procedures and understandings of sociolinguistics and that this field, in turn, would profit from the findings of SLA research (1993). Preston proposes analyzing interlanguage variation at different levels (phonological, syntactic, lexical, etc.) from _a variety of sociolinguistic issues: social context, task, topic, interlocutors, attention to task or form, ethnolinguistic background, etc. and how these factors interact with purely linguistic factors, such as linguistic environment (phonological or syntactic) and function-form relationships. In reviewing data from other studies he concludes, however, that at the early stages of second language acquisition '1inguistic rather than social environments are more powerful influences' (1989:256). Hulstijn (1989) makes a similar distinction among interlanguage variability studies but based on other criteria. He suggests a distinction between sociolinguistic approaches such as Tarone's and a cognitive, information-processing approach, which views language acquisition as the gradual transition from controlled to automatic processing. Hulstijn feels that 'style itself does not determine the degree of 27 attention to all produced forms alike; ... [and that] beginning language learners (due to their restricted language skills) are likely' to jpay :more attention to individual elements than more advanced learners' (20). He argues that for second language learners 'even in the vernacular (most casual) style, attention to form.is seldom altogether absent' (22). Beyond the question of attention to form, he feels researchers 'must find out if and how social, psychological, and educational factors are differentially associated with speech styles and attention paid to form! (22). Sociolinguistic studies show that L1 speakers manipulate _a variety of styles ranging from vernacular to careful (formal) styles based on whether or not they are attending to form, which, in turn, accounts for variability in their production. However, as Preston mentions, 'the variation of emerging forms cannot...be explained by the sorts of social factors which stand behind much NS variation' (1989: 33). Hulstijn also agrees that L2 speakers do not control all the styles in their L2 as an L1 speaker does. Learners nay control just one or, at most, a limited number of styles even at the advanced stages. A learner may be speaking in the vernacular, because that is the only style she controls, and yet be careful of how she produces that style. In other words, attention to form. for an L2 speaker does not necessarily signify a formal style for that speaker. Hulstijn mentions that '... attention to form is likely to vary within just as well as between speech styles' (19) . St 1" 1 . cité 28 Therefore, because the L2 learner is not capable of manipulating the same variety of styles that the L1 speaker is, the social model may not be appropriate for examining variability in interlanguage. Attention to form is not sufficient to account for variability in L2 production. Ellis (1986), points out that attention to form has, at times, produced.less target-like forms when the learner makes use of an L1 variant in careful speech. Ellis recognizes that the study of interlanguage and variability has greatly benefited from.using sociolinguistic approaches and feels it necessary to take social context into account, as opposed to the 'nmmerous studies which have analyzed interlanguage outside the context in which the data were collected (such as the morpheme studies). However, he is careful to distinguish between linguistic context and social context, and suggests that both interact to produce variability. Tarone mentions that variability is due to a complex interaction of factors: task, necessity for clarity, cohesiveness, linguistic environment, psychosocial factors and as Hulstijn suggests, knowledge of rules, and requirements of task as well. To Hulstijn's and Tarone's list of interacting factors several key linguistic factors must be added; in fact, these factors must be central to any study of interlanguage variability. These are the phonological, morphological and syntactic form of the structure and its semantic and/or functional complexity. 29 2.3. Interlanguage Studies and Attention to Form This section reviews the earlier interlanguage studies which attempted to link learners' variable production of a linguistic form with social factors such as setting, topic, interlocutor or task, factors which are suggested to cause the learner to attend to form. It is seen, however, that in many of these studies the researchers found that it was internal linguistic factors which seemed to best explain the source of the interlanguage variability. Early studies on interlanguage variability dealt primarily with phonology and how pronunciation shifted in response to differing tasks. _Dickerson (1975) and Dickerson and Dickerson (1977) found that their Japanese speakers improved in accuracy in the pronunciation of English /r/ as they' moved from. tasks requiring less to more attention (from free speech to reading ' word lists). What they also found in these studies was that their L2 speakers were not only sensitive to the demands of the task, which defines style, but also to the phonetic environment. Their subjects showed greater accuracy in pronunciation when the /r/ preceded a mid vowel than when it preceded a high vowel. Preston naintains that although external factors may cause variation, the variation is constrained by linguistic factors (1989). Ellis suggests that 'the effects of the linguistic and situational context interact to influence jointly the learner's use of interlanguage forms'(1986:84). Therefore, researchers should expect greater accuracy in the use of some forms both in 30 certain linguistic environments and when there is greater attention to form. However, 'attention to form} has been a difficult term.to define and has caused a great deal of discussion in the literature. One theory of second language acquisition is based on it (Krashen 1981, 1982, 1985). Tarone (1985a), who uses this concept as a variable in her earlier studies, admits it is difficult for researchers to control or :measure. Assuming that one can control for it by giving subjects tasks which we presume cause them to attend to form in varying degrees, researchers have not always seen a relationship _between accuracy and attention to form. Crookes (1989) found that providing learners with time to plan their utterances in oral descriptive tasks resulted in slightly more complex interlanguage production but did not result in greater accuracy. In his study, he looked at the variability and target-likeness of interlanguage by comparing it with interlanguage speech produced with no planning time. Crookes claims: Results showed that under conditions of planning, NNSs produced a greater variety of lexis (on one task), but not of syntax (in the test case of verb phrase). On both tasks, non—native speakers produced significantly more complex language, as .measured in terms of words per utterance, number of subordinate clauses per utterance, and S-nodes per utterance. General measures of accuracy ... did not show significant differences between conditions (377-378). he 9' 11 A‘I 31 Crookes points out that several factors may explain why he did not find greater differentiation between the two tasks. His Japanese subjects may have exhibited 'culture-specific patterns of language use' (379). In other words, Japanese may always carefully plan what they say, whether given time to do so or not. The social context of the data collection method may have affected the results of the study. In addition, he points out, it may be that the more complex the language produced (as was the planned production), the less likely it is to be accurate. It may have been that the subjects were attending to form _to an equal extent for both tasks, taking into account the cultural factor and the social context in which the tasks were carried out. This study points out, again, the difficulty in controlling for 'attention to form' and for making claims about a relationship between accuracy and 'attention to formm' Tarone (1985b, 1989) rejects that the factor 'attention to form' predicts accuracy in the use of certain morphemes in interlanguage. From a study she carried out with Japanese and Arabic speakers learning English, she concludes that both her model of interlanguage variation and that of Krashen, The Monitor Model (1981, 1982), fail to account for and capture the complexity of the variation of these speakers. She looked at the production of four grammatical morphemes: third person singular present tense -s, noun plural -s, articles a/an and the, and direct object pronouns in three different elicitation tasks: a written sentence-level grammaticality judgment task, FL bl)» CUT. 32 an oral interview, and an oral narration task. It was asswmed that each task would require increasingly less attention to language form, thereby causing less accuracy in the use of TL forms. The Monitor Model (Krashen 1981, 1982, 1985) claims that the learner possesses two independent systems of knowledge of the second language: conscious and unconscious. Conscious knowledge is accessible only when the learner is attending to form.and is available only as a 'monitor' to edit output. In this approach the only factor which accounts for variability is the presence or absence of monitoring - that is, whether the learner is attending to form or not. By the .same token, this approach would predict that there would be greater accuracy in the use of TL forms when monitoring. Like the Monitor Model, Tarone's Variable Competence Model also predicts greater accuracy when there is greater attention to form. The difference is that the Variable Competence Model accounts for a continuum.of styles. What Tarone found was that both groups of speakers significantly improved in accuracy in the use of articles and direct object pronouns on the narrative tasks, which supposedly required increasingly less attention to form. There was no significant difference in accuracy for either group among the three tasks for noun plural -s, nor for the Japanese speakers for third person singular present tense'-s. Only the Arabic speakers showed a significant decrease in accuracy in use of third person singular present tense -s from the grammar task to the oral narrative, as was expected by 33 Tarone. The results of her study failed to confirm the predictions of either the Monitor Model or the Variable Competence Model. Tarone says: ... the failure of both. models to accurately predict the complex pattern found may be due to the simplistic assumptions which both seem to share about the attention to form as the cause of style- shifting(1989:9). She concludes that it is '... the nature of the discourse which the tasks required and the degree of communicative .pressure which the tasks brought to bear upon the learner' (13) that account for the variability in these learners' interlanguage. She argues that, in narrative tasks, the need to be cohesive and clear caused the learners to use articles and the direct object pronoun it more carefully because these words are important in establishing cohesiveness and maintaining clarity in this kind of discourse. Third person singular present tense -s in this kind of discourse is redundant, and that is perhaps why it decreased in accuracy. Tarone (1989) concludes that variation in interlanguage nay be due to an interrelation among several causes: the function of the forms studied, the linguistic context of the forms, the identity and role of the interlocutor and the topic of discourse. It is evident that it is more than attention to form.that causes interlanguage variability. The linguistic structure 34 interacts at different levels and in different ways to produce variation in the learner's production. The next section discusses studies which have examined some of these linguistic features which may cause interlanguage variability. 2.4. Interlanguage and Linguistic Environment Later studies recognized that the linguistic system itself was an important factor affecting interlanguage variability and began to look at what features of the target language structure were responsible for this variability. This section examines studies by (1) wolfram (1989) who found, .as the present research also confirms for subjunctive marking, that irregular verbs are more likely to be marked for past tense; (2) Ellis (1988) who also discovered that the form.of the subject, whether it was a noun or a pronoun, affected both third person singular -s marking and use of the singular copula; and (3) Young (1988, 1993) who found that multiple linguistic and non-linguistic factors influenced the use of the plural suffix.—s for nouns. Young determined that mostly phonological and.syntactic factors affected variability in the use of this structure for lower proficiency learners. Wolfram (1989) looked at systematic variability in second language tense marking and found that there was evidence to suggest that tense marking in English L2 speakers is a highly variable phenomenon but that there is a set of linguistic constraints that affect tense marking in a systematic way. He found that 'both the phonetic composition of the past tense 35 form.and the phonological context in which it is found affect the relative frequency of past tense marking in a systematic way' (191). In his study he found that irregularity favored marking, in that regular forms of past tense, regardless of the phonetic form, were less likely to be marked for tense than irregular forms. Also, among the regular verbs, there tended to be less marking of past tense when a form.marked for past tense would end in a consonant cluster (e.g. /mIst/ -> /mIs/ 'mdssed') than when it would end in a single consonant (/peyd/ -> /pey/ 'paid'). Additionally, tense was more likely to be ‘marked when the following item began with a vowel (e.g. 'missed autumn') than a consonant ('missed school')(190). Among the irregular verbs he found that replacive forms (e.g. have/had, make/made) were less likely to be marked for tense than suppletive forms (e.g., go/went, am/was) (191). Wolfram attributes this to the 'principle of perceptual saliency' according to which the closer phonetically the past tense form is to the present tense the less likely it is to be marked for past tense. He also found that verbs which occur frequently (e.g. have/had, come/came, do{nt}/did{nt}) tended to be marked more often for past tense than less frequently occurring verbs, although this constraint was not as strong as the others. Wolfram further comments that the relative frequency of tense marking cannot be absolutely predictable based on these constraints, but that the 'likelihggfi of marking is systematically affected by these linguistic factors' (192). 36 Ellis (1988) also found that grammatical variation in the interlanguage of his subjects was sensitive to linguistic environment. In his two-year longitudinal study of two pre- puberty children and one puberty—aged child learning English as a second language, he looked at the development of third person singular -s marking and the use of singular copula, both the full (is) form and the contracted ('5) form. For third person singular -s, he found that target-like performance '...is more likely when the preceding element is a pronoun than when it is a noun' (263) and that there was a strong preference for using the contracted singular copula fwhen the preceding element (was) a pronoun, but for full copula when the subject (contained) a noun' (266) for at least two of his subjects. He could find no evidence in his data that the phonological environment affected the use of third ' person -s nor that its use was linked to specific verb forms. He believes that this study provides evidence that 'the distribution of grammatical variants in learner speech is sensitive to linguistic context' (269) and that 'it may also be that systematic variation occurs only when a certain stage of development has been reached' (269). He also found that 'for some learners, at least, the performance of a target variant proves easier in some linguistic contexts than others' (269). Young (1988, 1993) carried out studies where he looked at multiple linguistic and non-linguistic factors that could influence the varying usage of the plural suffix (-s) to mark 37 plural nouns among a group of Chinese speakers (1988) and two groups of Chinese and Czech/Slovak speakers (1993) learning English as a second language. He feels that 'most previous studies (in interlanguage variability) have been limited in their methodology to considering the influence of only one independent variable, and by their analysis of the dependent variable in terms of whether it conforms or deviates from.the form.required in the target language' (1988:283). He feels these kinds of analyses have given us an incomplete picture of the systematicity and its causes in the variable output of second language learners. From his (1988) study he predicted ‘that the degree of (-s) plural marking would be a complex function of factors deriving from four major influences: (a) the situational context of the interview, (b) the informants' overall proficiency in English, (c) the semantic and syntactic features of the NP and the phonological environment of the (s) plural marker, and (d) a tendency to eliminate redundant marking of number in the same clause (285—286). In short, he decided to analyze all the different factors that had been identified as possible influences in the variability in interlanguage. Using a multivariate procedure for data analysis, he came up with some interesting results. Basically he found that the tendency to mark for plural (s) was due to a complex interaction of the multiple factors he looked at. th surprisingly, he found that speakers from.the high proficiency group tended to mark for plural -s more 38 accurately than those from the low proficiency group. Other findings included that prenominal position and adverbial and complement noun phrases all strongly favored plural marking, but subject and object noun phrases were not marked for plural. In the phonological environments that he examined, he found that preceding vowels, non-sibilant fricatives and stops all favored plural marking. Preceding sibilants, nasals and laterals were found to inhibit plural marking. Following vocalic segments favored -s plurals and following consonantal segments inhibited them. His most interesting finding was that 'Redundant plural marking within the NP...did reveal a .highly significant effect...(;) the most salient markers of plural number - numerals and these/those - appear to strongly favor redundant marking of plural on the head noun, whereas if number is not marked anywhere else in the NP, there is a very good chance that it will not be marked on the head noun either' (295). Surprisingly, for this factor he found little difference between his high and low proficiency groups. However, there were differences between the two groups in terms of the influence of other factors. He found that the phonological environment had a greater effect during the early stages of acquisition, whereas factors such as the situational context of the interview, specifically the social convergence with a native speaker, tended to influence accuracy at later stages. Young concludes that 'the factors that influence variation in interlanguage change as the learner's system develops' (296). Finally, he believes that a diverse number 39 of factors --linguistic, developmental and contextual—— influence variability, and only by taking into account and controlling for as many variables as possible when analyzing interlanguage can we hope to explain this variability. 2.5. Interlanguage and the Relationship between Form.and Meaning and/or Function This section discusses three studies which also analyzed interlanguage variability by examining features of the linguistic system. and how this influenced variability. However, these studies conclude that the linguistic form interacts with the meaning or the function of the structure. '(1) Schachter (1986) found that what appeared to be syntactic free variation in negation development was actually quite systematic when viewed in terms of functions; (2) Huebner (1982, 1985) discovered that his subject's interlanguage variability was semantically and functionally determined; and (3) Rutherford found that in some cases learners are more likely’ to transfer jpragmatic features of their L1 than syntactic features which reveals that the SLA process is an interaction of a number of features of both the L1 and the L2 and processing strategies. In Schachter's (1986) study, she reexamines data from.a study done by Cazden, Cancino, Rosansky and Schumann (1975). The original study looked at the development of several syntactic structures, including sentential negation, by six second language learners of English: two children, two f) (I) I) 9’ r4 40 adolescents and two adults. The original authors concluded that Jorge, the twelve year—old Spanish speaker from.Colombia, showed considerable free variation in his development of negation in English. The variation could not be attributed to the varying situational contexts since the data were always collected under the same conditions: an informal interview between the researchers and the subjects. Schachter found that Jorge exhibited a great deal of functional systematicity in his use of negation. He tended 'to associate with each function a very limited set of syntactic forms and to associate with each syntactic form a _very limited set of functions' (131). One of the factors that helped Schachter to come to this conclusion is that which she defined as the 'onset' of analyzed use of a form/structure occurred. In the original study, Cazden, et al claimed that analyzed don't V'and analyzed Aux occurred early in the taping sessions in free variation with no V; Schachter claims that since the utterance, I don't understand occurred only once each in tapes 1, 2, and 3, the use of don't V’is an unanalyzed chunk. Regular, analyzed use doesn't occur until later, after tape 5, when no V'is dropped. She asks if onset should be considered to occur when the form/structure first appears, or when it begins to appear productively. She claims that it is the latter; onset occurs 'where the structure/form occurs across different lexical items and with some (unquantifiable) regularity' (127). She further suggests 41 Certainly these isolated early occurrences of a structure should not be ignored; they no doubt serve as indicators of what is to come. However, they should not lead one into claiming onset has occurred before it actually has (127). Ellis (1985, 1986) claims that when a new form enters into use in a learner's interlanguage it is used in free variation. If this is the case, then free variation of a form should alert the researcher that development is occurring. However, I agree with Schachter that this type of isolated or limited use of a form may signal onset of acquisition, but does not always do so. When the form occurs merely as a .memorized chunk, the form may appear accurately. This type of 'accurate' use does not signal onset of acquisition. A better indicator of onset is the attempt by the speaker to use the form in meaningful or functionally significant contexts. These attempts to use the form will show less accuracy than when the form is used as a memorized chunk and will show a greater degree of variability. Preston points out that 'the variability of an SLA system could be attributed to the emerging states of form-function correlations' (1989:33) . Schachter points out that many researchers claim that variability in the use of structures can be attributed to situational variability or to task differences, but that 'the learner is more limited [than the native speaker] in ability to juggle the various situations, meanings and forms at a given time. In fact, one might view proficiency in this juggling act as the ultimate goal of the language learner' 42 (121). This, of course, means that the ultimate goal of second language learning is to attain not only grammatical competence but semantic and pragmatic competence as well. In this sense, an account of a learner's acquisition of semantic/pragmatic competence is just as important as an account of grammatical competence. Another study which could make no sense of the variation in the interlanguage of a second language learner until the form/function relationship was considered is that of Huebner (1983, 1985). Specifically, Huebner studied the development of an adult Hmong speaker, a refugee from Cambodia residing in .Hawaii, who was acquiring English in an informal environment. The original study (1983) was a longitudinal study in which the researcher collected data approximately every three weeks through informal, one-hour, taped conversations. The participants, the setting and the tasks for data collection were held constant throughout the study. A subsequent study (1985) examines new data collected from the same subject twenty months after the end of the initial study. In the initial study, the data collected after one year were transcribed and analyzed for word order, topic markers, da (the subject's interlanguage article form) versus zero marking (0) for articles, and pronoun versus zero anaphora. Huebner reports that initially da was used with all referential definite noun phrases, except when used as a topic marker, when the NP was marked with zero. After six weeks, the informant 'flooded' all noun phrase environments with da 43 and manifested a more SVO word order over the previously presupposed-asserted word order. Around the let week of data collection, da was dropped from nonreferential environments and around the 27th week, da was dropped from referential indefinite environments. From.that point on, da, appeared to function very similarly to standard English the. In the subsequent study, Huebner found that the use of da was virtually the same as it was at the end of the original study; however, he found that the use of a had changed. In the original study, a was used either as a phonological variant of 0 or a hesitation phenomenon, whereas in the 1985 .study it was used primarily for referential indefinite noun phrases, mainly singular count nouns. Huebner feels these findings 'show positive correlations between the distribution of syntactic devices such as da and a and semantic and discourse-pragmatic distinctions' (1985:155). Although the use of these devices is not necessarily target-like, it is, nonetheless, fairly regular. Although this subject does not always mark the referential indefinite noun for the singular/plural distinction (the standard English. plural morpheme -s is not used), Huebner believes that 'the a/0 opposition is sufficient to perfonm the function of marking number' (156). What appeared. to Huebner at first as unsystematic variation was, in fact, highly systematic when viewed in terms of semantic and discourse-pragmatic functions. Huebner states that in order for his subject to be able to communicate it was important to express semantic and 44 functional distinctions and that [t]hese distinctions must be made, at any given point in time, through a fairly systematic use of linguistic devices. A totally random marking of these distinctions would make it impossible for the listener to decode correctly the intended message (1985: 155). Huebner concludes that A careful analysis of the ways in which these semantic and discourse-pragmatic functions are encoded in early varieties of interlanguage will yield insights into the internal organization of those interlanguage varieties [and will] contribute to the general body of knowledge about the organization of language in general (1985: 155). Rutherford (1989) also believes there is a relationship between the form and function of grammatical structures and this relationship affects the interlanguage production of L2 students. In examining over 300 compositions written by Japanese, Spanish and Arabic—speaking students of ESL, he found. that the Spanish. and..Arabic-speaking students' IL violated to a great degree the canonical SVO word order of English, even though these two languages also have canonical SVO word order. In the Japanese speakers' English compositions there was no evidence of violation of English canonical SVO word order. Rutherford explains these phenomena as occurring because, although Spanish and Arabic have canonical SVO, this word 45 order is flexible based on the traits of pragmatic constraints. And although Japanese, which is basically a SOV language, may permute the subject and object order according to pragmatic constraints, the verb is rigidly anchored in final position. In other words, Spanish and Arabic are languages which manifest pragmatic word order traits but have SVO canonical word order. Japanese is a language which manifests grammatical word order traits and has a non-SVO canonical word order. Rutherford maintains that this is support for his claim that there is pragmatic transfer from the L1 and not direct, syntactic transfer. Rutherford is interested in more than just the surface features of 11 variability. In his study, he applies the concepts of Universal Grammar in order to explain the features of IL production. Rutherford explains that Studies have moved away from. straightforward comparisons of a second language learner's IL and native language 'with respect to some strictly surface feature, and have instead begun investigating the possibility of more subtle influences of the native language (L1) on the shape of the IL . . . influences, for example, of L1 discourse function, of lexical features, of syntactic processes, of abstract organization, and so forth (165-166). Rutherford concludes that the IL word order he detected among his subjects is not due to a single principle of language behavior but rather results from the complex interaction of a number of principles related to the way in 46 which languages are naturally organized, how they are processed in use, and to how they are learned. Rutherford admits to the value of examining the natural production of second language learners stating that 'interlanguage by its very nature can provide a means for better understanding the forces engaged in that interplay' (178). These studies all reveal that the variability in the use of syntactic forms or syntactic word order is influenced in some way by both meaning and function. In the Schachter and Huebner studies it was seen that although the form their subjects' used and the way in which it was used did not ‘conform to native speakers' form and use, these learners' IL variability was systematic when viewed in terms of meaning and function; In the Rutherford study it was seen that although features of the L1 do influence IL variability, it is not always syntactic features but it some cases may be pragmatic features which transfer. Therefore, if we do not study how grammatical competence develops in relationship to semantic/pragmatic competence we are neglecting perhaps the most important aspects of second language acquisition. What is needed are studies that examine the features of the form/meaning/function relationship of grammatical structures and how these develop and vary in interlanguage. The next section presents a tool from variation theory that is helpful in determining what these features are. 47 2.6. Variation and Prototype Theory The present research study sets forth the hypothesis that syntactic and.morphological form.and semantics and pragmatics interact to constrain interlanguage variation. Therefore, it is necessary to distinguish what those linguistic features are for the structure under study. Adamson (1988, 1989 and 1990) proposes investigating the possibility that learners construct prototype schema for certain complex syntactic structures and for attempting to identify the features of these schema in the analysis of interlanguage variability. A prototype schema is a conceptual category which is not .clear cut but rather is a network of essential properties or features. Members of a conceptual category do not have to have any properties in common but will share properties in common with a prototype, which is the most typical or central member of the category. Adamson suggests that: if conceptual categories such as 'bird' and 'lie' are mentally organized around prototypical members, it is natural to wonder whether other linguistic structures such as passive sentence and noun phrase have this organization as well (1990:5). Adamson suggests that in a theory of language use a linguistic rule may be regarded as a type of schema in that it consists of knowledge regarding the structure and its use. Learners construct prototype schema of structures and apply this knowledge in the processing (production and reception) of the language they are acquiring. That learners may store 48 linguistic information in the form of a prototype schema has been investigated in second language acquisition by Adamson (1989) and Gass (1987), and in child language acquisition by de Villiers (1980). Gass found that both Italian native speakers learning English and English native speakers learning Italian construct a prototype schema of the subject-object relationship of different types of sentences. Her findings suggest that for both groups the schema for identifying a subject in sentences with the word order pattern of NVN is an animate noun acting ion an inanimate object and that this word order pattern was 'the strongest constraint. For word orders other than NVN, the native speakers of Italian relied more on animacy cues and the native English speakers on word order to identify the subject. In the DeVilliers (1980) study, it was hypothesized that children construct a prototype schema of the English passive rule based on the passive sentences they hear. This schema contains the features of an animate surface subject and an active verb. In an experiment DeVilliers found that children who did not produce passives were able to learn sentences that contained these prototypical features more easily than those which. did. not. This suggests that :members of certain grammatical categories are more central than others to the prototype schema. For children, action verbs like hit are more prototypical than stative verbs like know or experiential verbs like feel. Those members that are more central to the category are used.more often and more correctly in grammatical 49 constructions. As studies have become progressively more concerned with identifying the linguistic factors which constrain and cause variation in L2 learners' interlanguage, it can be seen that prototype theory can provide a tool with which to identify these factors and offer insight into how they interact or combine to cause variation in learners' interlanguage. 2.7. Conclusion Studies on variability and interlanguage have moved away from the emphasis on looking at a single external, social- .contextual factor to account for the variability inherent in learners' interlanguage to a focus on studying how multiple linguistic factors concerning phonology, morphology, syntax, semantics and pragmatics interact with each other and also with other' psycho-social factors to cause interlanguage variability. Ellis claims that interlanguage production can give us clues as to the developing competence of L2 learners' language and that research needs to address both linguistic form.and how it interrelates with function in order to explain the SLA process. Schachter found that apparent syntactic free variation in negation development, when viewed in terms of functions, was systematic. Huebner revealed that what appeared to be unsystematic variation, when viewed in terms of semantic and discourse-pragmatic functions, was also highly systematic. Rutherford points out that among Spanish, Arabic 50 and Japanese speakers, pragmatic word order transfer was more evident than syntactic grammatical word order transfer. In his study, it is apparent that pragmatics, for certain functions, takes precedence over syntax. Young, who looked at multiple features, both social contextual and linguistic, found that phonological, syntactic and semantic features interact to account for interlanguage systematicity and variability. Finally, Adamson provides us with a valuable tool for the identification of distinct linguistic features of structures and an analysis of how these features combine to constrain interlanguage variation of certain structures. It is of primary importance to look at the linguistic system of the target language itself and to attempt to identify which of these factors might cause interlanguage variation. However, it is equally necessary to analyze the ' system of the interlanguage that is produced by learners and try to understand how this might affect their production of forms. We must keep in mind that interlanguage forms may not always correspond to target language forms, nor do meanings and functions in the interlanguage always correspond to those of the native speaker. The only study that emphasizes this point is that of Huebner, who found that the morphological form of da does not correspond to the target form the. And certainly, the function or functions for which this form was used were not target-like throughout the time-period of the study. His study provided important insights into how language is organized in general and, specifically, how 51 linguistic form and meaning/function relations interact in the second language acquisition process. The present research study maintains that it is important to analyze what linguistic factors cause interlanguage variation, and to attempt to identify the features of a target-language structure and how these favor or inhibit marking of the structure; but it is believed that it is equally necessary to analyze the forms that are used as alternatives to the structure. We must analyze the interlanguage system in its own right. Of course, this claim does not diminish the necessity of analyzing the target .language as well. The researcher must be acquainted with the target language. variation studies therefore, need to combine analyses of the target language with analyses of the learner's interlanguage. It is in this way that an analysis of interlanguage variation can lend insight into the complex interaction of the linguistic system, the developing competence and the second language acquisition process. CHAPTER THREE: TRADITIONAL AND PEDAGOGICAL DESCRIPTIONS OF THE SPANISH MOOD SYSTEM 3.1. Introduction What is the target language which the learner aspires to acquire? For the present research study, the most relevant question at hand is, what is the nature of the present subjunctive mood in Spanish? The researcher like the language learner, needs to know specifically: (1) the morphological form of the structure under study, (2) the syntactic structures in which it is used, (3) its meaning, (4) how and why it is used, and (5) how it contrasts with other forms .(in this case, the present indicative mood). Most learners have their first encounter with the target language in the language classroom and through some kind of textbook. Until relatively recently, with the appearance of texts such as Lunn and DeCesaris's (1992), which require students, usually advanced learners, to analyze the language in meaningful ways, the majority of texts are based on traditional structural grammars which answer the first two questions posed above, to a lesser extent the third, and almost never the fourth and fifth. This chapter reviews some of the more 'accepted' traditional grammars' accounts of Spanish mood in general and the present subjunctive mood in particular, and one language learning text's presentation of the present subjunctive in order to show that these kinds of grammars are insufficient 52 53 for answering the most difficult, but most essential, questions learners need to answer in order for full acquisition of the Spanish subjunctive to develop. The Spanish grammars selected for review are (1) La Gramatica de la Real Academia ESpafiola (Onieva Morales 1993), (2) Gramatica de la Lengua castellana (Bello 19841), (3) Manual de Gramética Espafiola (Seco 19892), and (4) Curso Superior de Suntaxis ESpafiola (Gili y Gaya 1969). All of the grammars are prescriptive descriptions of formal, educated Spanish. The Onieva Morales, Bello and Gili y Gaya grammars are based on the Castillian dialect and the Seco grammar, although strongly ‘influenced by Castillian, was written specifically for Latin American speakers. The text reviewed is Mundo Hispano: Lengua y Culture (Olivella de Castells 1981). The text attempts to present every day, educated Spanish, both the European and Latin American varieties, but basically presents the most formal style. This text was used in the Spanish language program. in which the subjects in this study participated. The review demonstrates that these traditional grammars and the textbook these learners used all fail to note factors about the nature of the Spanish subjunctive which appear to play a role in the interlanguage revealed in the present study. Most importantly, these grammars and the text fail to point out the importance of how semantics and pragmatics interact with. syntax and. morphology in understanding and using the Spanish subjunctive. 54 3.2. Traditional Grammars Probably considered the foremost authority on Spanish grammar in both the Spanish- and non—Spanish-speaking world is La Gramatica de la Real Academia Espaflola (Onieva Morales 1993). In the introduction it is stated from the beginning that the purpose of the manual is to 'divulgar la filtima...doctrina gramatical de la Real Academia' (13). Later, it is contended that despite criticisms on the part of certain linguists, the grammatical doctrine presented in the volume represents accepted points of view, both normative and descriptive. It is not mentioned by whom this doctrine is ‘accepted; certainly not by 'certain linguists'. The Academia's chapter 37, 'Significado y Uso de los Tiempos del Subjuntivo' ('Significance and Use of the Tenses of the Subjunctive'), is neither significant nor useful for second language learners. It mentions that it is the unreal character of the action expressed with the different forms of the subjunctive that makes the temporal relations of its tenses much less clear than those of the indicative. In addition, it is affirmed that 'el valor temporal de las formas del subjuntivo es muy impreciso y...en todo caso depende del contexto' (277), an affirmation that is not untrue but hardly enlightening. The present indicative is compared with the present subjunctive and it is pointed out that for clauses which are dependents of verbs of perception or declaration, the indicative fonm is used as in creo que viene JUan. ('I believe 55 John is coming.') and Creo que vendra JUan. ('I believe John will come.'). In the indicative the present and future are distinguished in the form of the verb 'to come'. However, when the two actions of coming (both present and future) are expressed as unreal, there is no distinction of present and future and the verb is expressed in the subjunctive as in Deseo que venga JUan. ('I hope John comes.') The Academia states that in order to express present or future desire in subordinate sentences, there is but one tense in Castillian: the present subjunctive, which is present and future at the same time (278). [MLQ's translation] and that When we say Nb creo que llegue a tiempo ['I don't believe that he arrives on time'], we can refer just as much. to the act of arriving that is occurring right now as to that which will occur in the future (278). [MLQ's translation] There are several problems with the Academia's description of the use of the present subjunctive. The first is that it does not discuss the fact that creer, when it is negated as in the above example, can be followed by either a subjunctive or an indicative verb in its complement clause with subtle changes in meaning. The second problem, is that for sentences such as.Nb creo que llegue a tiempo,:most native speakers would agree that the action of arriving certainly 56 refers to a future action. If the action were in the present, most native speakers would state it in the present indicative, No creo que llega JUan. or No creo que esté llegando JUan. This doesn't deny the fact that the present subjunctive can express actions that are occurring at the time of utterance in sentences such as,.Me alegro que estés aqui, or El hecho'de que Juan llegue tarde no me molesta, but then these actions cannot refer to future actions 'at the same time'. Another problem with the Academia's 'adopted point of view' is that it does not take the different syntactic 'environments into account in pointing out how these may constrain mood choice and meaning. Finally, even if the Academia's description of the use and description of the present subjunctive mood were a reflection of the way native speakers use the subjunctive, it makes no attempt to explain why the subjunctive is used in this way. In fact, the 1993 edition of La Gramatica de la Real Academia Espafiola, answers none of the five issues raised at the beginning of this chapter (page 50). It is the unreal character of the Academia's description of the Spanish mood that makes it much less clear than most Spanish grammars. In Bello, the leading authority in the 19th century, the description of the subjunctive is also brief and very prescriptive in nature; he is more concerned with explaining the syntactic rules followed in using the subjunctive than on any pragmatic or semantic explanation. if 51 Si 61 57 Basically, Bello's definition of subjunctive mood is a syntactic one: They are called moods, those inflections of the verb which come from the influence or the regimen of a word or phrase which is or can be subordinated (158). [MLQ's translation] and mentions that the name of the mood is 'subjunctive, because it figures often in subjunct propositions, that is, subordinated ones' (159) [MLQ's translation]. Bello points out that forms of the subjunctive mood are subordinated or can be subordinated to the verbs 'to doubt' ‘and 'to desire.‘ He contrasts this with the forms of the indicative mood which is the mood which serves to indicate affirmative or negative judgements whether these are on the _ part of the person speaking or another person indicated in the proposition that depends on the verb. He also mentions that the subjunctive mood is used in words or subordinate 'phrases' which denote uncertainty or doubt or some kind of emotion, even those indirectly affirming the object or cause of the emotional state. In the sentence, Dudamos que vivas contento, aunque todo contribuye a que lo estés ('we doubt that you live happily, although everything contributes to the fact that you do'), he mentions that the rule that asks for the subjunctive following verbs expressing emotional states prevails over the rule which assigns indicative to judgements and that this explains why vivir is expressed in the subjunctive although 58 immediately after, in the next clause, this action is confirmed. For his day, Bello gives an extraordinary amount of detail and description as to how the subjunctive is used. His account, however, is basically a syntactic one: the subjunctive is used in subordinate clauses and the terminology he assigns, no longer conventional, is purely descriptive. Seco, Latin America's early twentieth century authority, explains very briefly that the different moods in Spanish, the indicative, the subjunctive and the imperative, represent three different points of view. The indicative is the objective mood, without the speaker imposing any judgement .about what he is saying. The subjunctive mood exhibits a purely subjective point of view. In the example, Yo no creo que Pedro venga boy ('I don't believe that Peter comes today'), he explains that ...in no case is it affirmed that Peter comes or not, but rather that this fact of coming exists only in the mind of the speaker as a certain subjective disposition of his. This is, in general, the meaning of the subjunctive mood, to which the form. venga corresponds (67). [MLQ's translation] Seco points out that while the subjunctive expresses that} i'essence' of a fact, the indicative expresses its 'existence'; 5 the indicative has the sense of ' affirmation ' while the -. {subjunctive has the sense of 'the suspension of affirmation.‘ ( ' i‘ H\ Seco, although.he makes no attempt to explain why these set of" facts should be so, at least comes closer to answering some of 59 the important questions about the nature of the Spanish mood system.that a researcher and learner would need to know. Gili y Gaya (1969), a modern grammarian interested in the child's language acquisition of Spanish, refines Seco's 'description of 'essence.‘ He claims that the subjunctive is the mood which expresses one' s subjective point of vieW’. The Subjunctive is used in both independent and subordinate clauses, although it is essentially subordinate and depends upon another verb which expresses some matrix of irreality. W~-‘___ w __-. \ He suggests that in independent sentences such as inala llueva! ('I hope it rains!) and Quizas no volvamos a verle _(Perhaps we won't see each other again), the verbs, which grammatically do not depend upon a main verb, are in the subjunctive because they deal with.mental subordinations which psychologically involve the judgment of the person speaking (132). The subjunctive in subordinate clauses is employed w”~‘\~ when the main verb expresses a doubtful, possible, necessary ‘1‘ .‘\-‘-.-.< _ ... or desired action (133). He maintains that mood is a linguistic signal for both the speaker and the listener which expresses affirmation or negation. He suggests that 'los modos reflejan las mas leves oscilaciones de la duda y constituyen un medio de expresién extremadamente sensible' (136). Again, in the latter half of the twentieth century, apart from.the Academia Real, there is an attempt on the part of leading grammarians to account for the significance and use of the Spanish subjunctive in terms of the interaction of syntax 60 and semantics but none is able to get beyond the question of mere description. It is of no surprise, then, that textbooks designed to teach Spanish to non-native speakers are not able to get beyond this point as well. 3.3. A Textbook Description of the Spanish Subjunctive The textbook description of the Spanish subjunctive is taken from Olivella de Castells' second.year textbook, Mhndo Hispano: Lengua y Culture, which the subjects in this study used during their study of Spanish in the Querétaro program. Olivella de Castells introduces the present subjunctive .in the seventh of twelve chapters. She contrasts it with the indicative and presents the forms of the present subjunctive, pointing out that the spelling changes and the forms of the irregular verbs manifest the same irregularities as the USted and UStedes commands. The explanation she gives in contrasting the two moods is partially syntactic, partially functional. She mentions that the indicative is used in both main and subordinate clauses Kain,“ “av-“1...... we and expresses events that happened or are happening or will K\I_ ”Iniiggfig_igfigw_ _ - happen in the future with a some degree of certainty. The Rwy.“ .... " subjunct1ve, on the other hand, she states, is almost always 'W used in subord1nate clauses and generally refers to events I _. «rm-u...“ A that one does not know have occurred or will occur. In cases where the subjunct1ve mood refers to an event which has n "H. 9. * occurred or which is occurring at the moment of speaking, the -..W Wv“ u. 11.. v. verb in the main clause (which is in the indicative) expresses 61 an emotional state or personal point of view. This difference ...—...— nafik_ , is then made clearer by the use of numerous examples contrasting the indicative and the subjunctive in the following format: main clause subordinate clause (1) Yo sé que él trabaja en esa oficina. 'I know (that) he works in that office'. main clause subordinate clause (2) Yo quiero que él trabaje en otra oficina. 'I want him to work in another office'. In both. examples, the 'verb in the :main clause is expressed in the indicative. Olivella de Castells points out that, in (1), the verb in the subordinate clause is expressed in the indicative because what is expressed in that clause is a real fact and the speaker is not expressing any emotional state or personal point of view in the main clause. In (2), the verb in the subordinate clause is expressed in the subjunctive because it is not a true fact that is expressed but rather a desire on the part of the speaker. It is mentioned that the verb in the subordinate clause may be in the indicative or the subjunctive mood depending on what the speaker wants to express. It is evident that the author at least attempts to make the student aware of the semantic and pragmatic differences between the use of the two moods and 62 doesn't depend on a purely syntactic explanation for teaching the subjunctive. This author continues by pointing out that the subjunctive is used in the subordinate clause when the subject of the main clause tries to influence in some way the action on the part of the subject (a different subject) of the subordinate clause, as in the following examples: (3) Yo espero que vengas mafiana. 'I hope that you come tomorrow.‘ (4) Nosotros le exigimos que entregue el dinero ahora. 'We demand that you hand over the money.now.' (5) Te prohiben que fumes. 'They forbid.you to smoke.' (195) It is mentioned that this influence on the part of the main clause subject can range from a simple desire to a demand or a prohibition on the part of the speaker. In addition, when the action of the subject of the subordinate clause influences the subject of the main clause (and they are not the same subject) producing a psychological reaction (happiness, fear, pain, emotion, etc.) then the verb of the subordinate clause is expressed in the subjunctive, as in the following: (6) Yo siento mucho que no puedas venir. 63 'I am very sorry that you can't come.' (7) Nos alegramos (de) que estén satisfechos con el,pedido. 'we are glad. that jyou. are satisfied. with the order.‘ (198-99) However, it can be seen that in these examples, the directionality of the influence between clauses is reversed (as opposed to examples 2—5 above). Since the subordinate clause is influencing the main clause, the action has to be real, although it is expressed in the subjunctive. The text .fails to point this out and in these cases, an analysis of the subjunctive as the mood of irreality does not hold. The text needs to make this clear to students. Another use of the subjunctive that this chapter mentions is the expression of doubt. If the_speaker has doubts, then the subjunctive is used; if he doesn't have doubts, or these w“ doubts are :minimal, he uses the indicative, as in the -b.hl-JA(¢'\ -.u-‘f fallowing examples: (8) Dudo que el pedido salga hey. 'I doubt that the order will go out today.' (9) .No creo que el pedido salga/sale hey. 'I don't think the order will go out today.' (10) Creo que el pedido sale hey. 'I think the order will go out today.' (199) 64 In (8), the author explains, the speaker expresses doubt and the verb in the subordinate clause reflects this (although it may be redundant since doubt is expressed in the main verb). In (9), the subordinate verb may be either in the indicative or in the subjunctive depending on whether the speaker wants to express considerable doubt (indicated by the subjunctive) or minimal doubt or certainty (indicated by the indicative). In (10), the speaker wants to express certainty, and the subordinate verb expresses this by use of the indicative. It is mentioned in the text that doubt or certainty can also be expressed by impersonal expressions, 'which reflect personal opinions or judgments; in these cases the subjunctive confirns the doubt and the indicative the certainty of the speaker, as in: (11) Es verdad que tenemos una fotocopia. 'It's true that we have a photocopy.‘ (12) Nb es seguro que necesiten/necesitan .mas empleados. 'It's not sure that they need more employees.‘ (200) It is explained that in (11) the indicative confirms the certainty on the part of the speaker. In (12), the speaker can either indicate certainty or doubt depending on the use of the indicative or the subjunctive. Olivella de Castells mentions that the use of the 65 subjunctive or indicative in subordinate adjective clauses also signals the difference between certainty and uncertainty in another way. The indicative is used when the adjectival ‘LW “‘\ clause modifies a known or spec1f1c antecedent The ‘W- W“ V 'subjunct1ve is used when the subordinate clause modifies an .- ~41. ... MH‘_. »,. h-«a— —_—-- unknown or undetermined antecedent. The following examples “I a.” “x are presented to make this clear: (13) Te voy a llevar a un restaurante que queda cerca. (14) Te voy a llevar a un restaurante que quede cerca. 'I'm going to take you to a nearby restaurant.‘ (203) In (13) the verb in the subordinate clause is in the indicative; in (14) it is in the subjunctive. However, the text gives the same gloss for both sentences and points out that in (13), the speaker has a restaurant in mind and is going to take his listener to that restaurant, whereas in (14), the speaker does not have any restaurant in mind.but is going to take his listener to any Close restaurant. For (14) a more accurate English gloss would be: 'I'm.going to take you to some restaurant nearby.‘ This is clearly they, +specific/+definite vs. -specific/-definite case. Spanish indicates this by the differing morphological endings of the indicative and subjunctive moods. l f 66 English has a linguistic means of expressing this difference through the definite and indefinite articles. Spanish has this distinction and uses it in addition to the indicative/subjunctive moods to indicate +specific vs. - specific. In fact, in the following examples, the indefinite or definite article, the use or non-use of the personal a, and the use of mood all combine to indicate whether the speaker has someone in mind or not: (15) Necesitamos al empleado que trabaje en ese departamento. 'We need the employee who works in that department.‘ (16).Necesitamos a un empleado que trabaja en ese departamento. 'We need an employee who works in that department.‘ (17) Necesitamos un empleado que trabaje en ese departamento. . 'We need any employee who works in that department . ' (203) In these sentences, there is a complex relationship between syntax and meaning, but the text fails to point this out and merely gives the English glosses and mentions that the subjunctive is used with unknown antecedents and the indicative with known antecedents. What needs to be explained 67 to the learner is that in (15), the specificity of the employee, which is known both to the speaker and the listener and which is manifested in English only by the definite article the, in Spanish, is indicated redundantly by the use of the personal a, the definite article el, and by the use of the indicative mood in the subordinate clause, trabaje. It also needs to be pointed out to the learner that in (16), the speaker has an employee in mind, but he is one of many who have this characteristic. Such an instance might be when there is a meeting and a worker, who is normally not at the meetings, is present. Another worker asks, 'Why is x .here?' and the response is: Necesitamos a un empleado que trabaja en ese departamento. In (17) , the speaker does not have anybody in mind; he is only mentioning that they need somebody, anybody, that works in that department. Here the difference is manifested by the lack of the individualizing personal a, the lack of which indicates that what follows does not refer to any person, but rather to a class ('un empleado'), and by the use of the subjunctive mood which indicates that the antecedent is undetermined. It is not important who the employee is as long as he or she works in that department. The relationship between syntax and meaning is not obvious to the learner and must be explained, a shortcoming of the text and of most language texts based on traditional descriptive grammars. The author then points out that in adverbial clauses, which modify the verb or the main clause, the speaker must 68 choose the mood which reflects the reality of the event. The following examples are presented to point this out: (18) Hace el pedido cuando llega. 'He places the order when he arrives.‘ (19) Haré el pedido cuando llegue. 'I'll place the order when I arrive.‘ (20) Hagalo como le dicen. 'Do it the way you have been told.‘ (21) Hagalo como le digan. 'Do it however they may tell you.‘ (22) vamos a salir aunque hace frio. 'We are going out even though it's cold.‘ (23) vamos a salir aunque haga frio. 'We are going out although it may be cold.‘ or 'It's cold but we're going out despite the cold.‘ (208-209) The text presents the examples with their English glosses and briefly explains that in (18) the indicative indicates the event occurs habitually and in (19) the subjunctive indicates a possible future event. In (20) and (22), the indicative m t .w u...‘ _ ..H - ‘~Yhdm __ indicates factual events, while the subjunctive in (21) and (23) signals possible events. However, the text fails to point out that the use of the subjunctive in (23) could signal a factual but insignificant or irrelevant event. a.“ . . Olivella presents expressions in Spanish which always . .n-- u “w. .... 69 require the subjunctive in the verbs that follow them. Logically, these expressions include: lpara que, a fin de que, 'so that, in order that'; sin que, 'without'; a menos que, 'unless'; con tal (de) que, 'provided that' and, antes (de) ‘\+_L_._ ...—_‘ ...-.___ ... qu,fi'before'. Often, as in this book, these expressions are introduced to students as expressions which always require the subjunctive. Although this is true, the explanation of these expressions does not point out how it is the semantic or pragmatic significance of these phrases —— that they introduce ..I’H" ‘ unreal or poss1ble future events -- which requires the use of h-1_~.._W the subjunctive. ‘Wfiyflymhfl-‘fi' um p'i‘fiwh ‘I’, r“"‘" In addition to the above expressions, there are other syntactic frames in Spanish that always require the subjunctive such as querer que___ and ojalé (queL___. Perhaps students can learn these structures more easily because they need only refer to syntactic rules for usage. Semantic/functional explanations are not necessary, although most learners do want to know what phrases mean, how they are used and why. This is even more important when the student has to decide whether to use the subjunctive or the indicative based on meaning. A grammatical rule, based on the syntactic frame of the utterance, however, does not always help in these cases, since the speaker can choose either mood. The English native speaker's semantic rules do not always transfer in these cases, and the learner has to understand the subtle semantic and pragmatic differences signaled by the use of mood. It nay be better for learners if this is pointed out 70 from the beginning so that they can be aware that certain structures in Spanish depend on the meaning the speaker wants to give his utterance. In chapter nine, Olivella de Castells presents the use of the subjunctive in independent clauses with ojalé, quizé, quizas, and tal vez. With ojala, the subjunctive is always used because ojala expresses a desire. It can be followed by the subjunctive in different tenses. There is no explanation of the semantically dependent nature of these clauses, which even in Gili y Gaya's structuralist description of Spanish is pointed out. Olivella de Castells gives the following ‘examples: (24) inalé no desconecten esa.méquina! 'I hope they don't unplug that machine.‘ (25) :Ojalé no desconectaran esa maquina! 'I wish they wouldn't unplug that machine.‘ (249) It is pointed out that the imperfect of the subjunctive is used in place of the present when the speaker wants to express an even more negative attitude towards the possible realization of the action or that the action is contrary to the reality of the present. The chapter ends by presenting the phrases quiza, and tal vez, and pointing out that the subjunctive is usually used, but the indicative can also be used if the speaker is quite sure that the action has occurred or will occur. Wh Sa Dr St: 71 (26) Tal vez compren un avicSn supersénico. 'Perhaps they may buy a supersonic plane.‘ (27) Tal vez compran un avion supersénico. 'Perhaps they will buy a supersonic plane.' (250) The use of the subjunctive in (26) shows more doubt on the part of the speaker than the use of the indicative as in (27). As can be seen by this overview, the presentation of the present subjunctive is basically a syntactic one (the other tenses of the subjunctive are presented as well but are not reviewed here because the present study focuses on the ‘acquisition of the present subjunctive); the semantics and the pragmatic functions of the subjunctive are discussed only when the student cannot make the choice between the subjunctive and the indicative by syntactic frame alone. The textbook does ' not lead the student to analyze the syntactic structures and to link them to the semantic and functional properties of the Spanish subjunctive mood. 3.4. Conclusion Neither traditional descriptions nor most language texts, which are based upon the traditional descriptions, satisfactorily describe or explain the complex nature of the Spanish mood system. Again and again, the focus is on presenting the syntactic structures and only linking the structure with semantics and pragmatics when a syntactic description does not suffice to explain the structure. It 72 appears that these authors see it as undesirable to include semantics and/or pragmatics in any syntactic description. This trend has been carried over into generative/transformational linguistic description as well. Syntactic explanations of structures that include semantics are seen as insufficient or as having to apply late rules of semantics in order to satisfactorily explain the structure. By the late 1960's, however, linguists and language teachers alike were beginning to recognize that language in general, and complex structures, such as the Spanish mood system, were not fairly accounted for by traditional, structural or .generative/transformational descriptions. There was a need to look at language, language use and language learning from.new and different perspectives and at different levels. ENDNOTES 1The first edition of Bello's Gramatica came out in 1854, as far as I could discern from.the 1984 edition. His grammar is very much a prescriptive one, not based on actual usage. He believed that a grammar should be unified and largely unchanging. 2The first edition of Rafael Seco's Manual came out in 1930 in two small volumes. The present discussion is based on Manuel Seco's larger and revised 1989 edition which first came out in 1954. CHAPTER FOUR: SYNTACTIC, SEMANTIC AND PRAGMATIC EXPLANATIONS OF THE SPANISH MOOD SYSTEM 4.1. Introduction This chapter reviews theoretical studies of the Spanish subjunctive which have found that it is necessary to discuss the semantic and/or pragmatic nature of this complex structure. Many of the researchers were Spanish teachers attempting to come up with a description or explanation of the subjunctive mood that might help learners understand when and how to use the structure. Others are linguists seeking to understand just why native speakers use the subjunctive. It .will be seen that descriptions of the subjunctive mood have been difficult to put together because the Spanish subjunctive is, in Lunn's words, 'a seemdngly intractable puzzle' (1989b: 691). It was found in the present research that, by the end of the data collection period, the learners were beginning to produce the subjunctive, along 'with a great deal of variability in the use of other verb forms, for only certain syntactic frames, i.e. a specific set of matrix verbs. There appears to be a relationship between the difficulties researchers have had accounting for the nature of the Spanish subjunctive mood and the way both native—speaker children and non-native speaker adults acquire this structure. .Mbstly, it is when semantic and pragmatic subtleties come into play in describing the subjunctive mood that researchers have had a 73 74 difficult time. It is also in these areas that children and adults have difficulties acquiring this structure. What are these semantic and pragmatic subtleties? Different researchers have looked at them in different ways. Most have attempted to build upon one or more observations of earlier studies. This chapter divides these studies into two groups: (1) 'semantic studies' and, (2) 'semantic/pragmatic studies'. 4.2. Semantic Studies These studies attempted to explain the subjunctive by taking into account the interaction between syntactic and _semantic features, such as presupposition and assertion (Bolinger 1968; Rivero 1971; Terrell and Hooper 1974; Terrell 1976; Goldin 1974; Garcia and Terrell 1977; Lantolf 1978; and Bell 1980) or semantic features such as subordination and independency (Takagaki 1984; and Reider 1989, 1990). Bolinger, Terrell and Hooper, and Goldin admit that each of their studies is a search for a rule for the Spanish subjunctive that will aid in teaching. Others merely seek to establish a ‘generalization that will help to explain the different uses of the subjunctive. As will be seen, the complexity of the Spanish subjunctive has made both goals difficult to reach. 4.2.1. Bolinger Bolinger's (1968) work first brought to our attention the need to bring semantics into the picture in descriptions of syntactic phenomena by pointing out that the semantic 75 distinction of the Romance subjunctive mood is manifested in English, where English does not allow main verb phrases to be postposed. For example, if in English the postposed main phrase (PMP) is allowed, then the corresponding sentence in Spanish would require its complement to be in the indicative; if the main phrase cannot be postposed, then the resulting sentence in Spanish would require the subjunctive in its complement: (1) a. I believe they're ready. -> They're ready, I believe. b. Creo que estén listos. (2) a. I don't believe they're ready. -> *They're ready, I don't believe. b. Nb creo que estén listos. Although the resulting rule he presents is as clumsy for teaching purposes as are the syntactic 'textbook' rules he rejects, it is illuminating to point out that indicative complements can be stated independently; subjunctive ones cannot, In addition, this distinction is revealing in terms of the underlying semantic possibilities which distinguish the indicative and subjunctive complements. Bolinger points out that both Spanish and English represent reality in a similar way: by emphasizing that representation. However, Spanish uses the indicative and English allows the representation to be fronted. When Spanish requires the subjunctive, English 76 does not allow the representation to be fronted and therefore to be emphasized. There are some problems with this parallel concept . In Spanish there are complements that can take either the indicative or the subjunctive. When the subjunctive is used, Bolinger' s rule works . But the rule doesn't work when the indicative is used. In fact, Bolinger's rule would predict that the indicative cannot occur. For example, (3) I don't think he's coming. -> *He's coming, I don't think. 'would correctly predict the occurrence of (4) .No creo que venga. (subj) but should exclude the occurrence of (5) .Nb creo que viene. (ind) (22) (5) is, however, a grammatical sentence. Bolinger explains that this is due to the fact that (3) has a mixed status and both verbs, the main and the complement, share the negation. His evidence for this comes from the idea of 'absorbed negation' and the allegation that for native speakers (6) and (7) below are closer in meaning than (8). 77 (6) Nb creo que es asi. 'I don't believe it's like this.‘ (7) Creo que no es asi. 'I believe it isn't like this.’ (8) No creo que sea asf. 'I don't believe it's like this.‘ (23) This 'absorbed negation' or 'not-transportation' is explained more satisfactorily twenty-two years later by Reider (1990) who points out that due to the Neg-trace Condition, 'the semantic interpretation of a verbal complement is affected by .the presence of...(a) negative element' (213). Another problem of Bolinger's analysis is that he deals with noun clauses only and not adjectival or adverbial clauses which also require analysis. Despite the limitations and the lack of a formal theory upon which to base his observations, the work of Bolinger helped to launch the semantic-based analyses that dominated the field in the 1970's and 1980's. 4.2.2. Rivero Rivero's (1971) detailed analysis of the Spanish subjunctive linked. the then-current transformational- generative approach in both syntactic and in semantic analyses. By looking at a number of transformations, Rivero provides evidence that subjunctive and indicative complementizers have different underlying structures and that the subjunctive complementizer behaves differently from the 78 indicative, both semantically and syntactically. The indicative complementizer manifests a positive presupposition about the truth of the complement, whereas the subjunctive complementizer expresses a neutral attitude about the truth of the complement. The nature of the presupposition of these two kinds of complementizers must be reflected in the underlying structure because it cannot be explained by surface structure interpretation rules. Rivero contends that it is the presuppositional nature of the complement itself which is reflected in the underlying structure which affects the syntactic behavior of sentences. .Contrary to what Bolinger suggested just a few years earlier, Rivero maintains that, syntactically, an indicative complement: ...cannot undergo Neg-transportation, subject- raising, Equi-NP Deletion, or Neg-incorporation, although it has the derived structure of a regular complement. It is not subject to tense restrictions either (332). Semantically, the indicative complementizer '...involves a presupposition which is not made by the subject of the matrix verb but by somebody else who is not mentioned in an overt manner...‘ (332). This explains how two sentences with the same matrix verb can take two different complementizers. In the sentences: (9) a. Admite que zinc e1 inspector (Ind.) 79 'He admits that the inspector came.‘ b. Admite que ziniera e1 inspector (subj.) 'He confesses that the inspector came' (although he's not confirming that he came). the matrix verb, admitir 'to admit' does not contribute to the difference in meaning. The difference in meaning can only be attributed to the difference in mood. Rivero explains that (9.a) presupposes that the inspector came and the speaker knows it, but (9.b) does not presuppose that the inspector came. Rivero's definition. of presupposition. makes her ‘explanation a bit confusing. Although she does not precisely define the term, it appears that what she means by 'presuppose' is actually 'confirm' as Terrell (1976) points out (footnote 2, p. 240). If we understand Rivero to mean 'confirm' when she says 'presuppose' then by using the subjunctive in (9 a.), the speaker does not confirm.that the inspector' came (for‘ example, under torture, the speaker confesses something that he does not know whether is true or not). Rivero argues that although surface structure interpretation rules could account for examples like (9) above, '...we would need two separate rules, one for each surface structure, for what is a common and unique phenomenon' (330). In attempting to explain the Spanish subjunctive Rivero endeavors to apply rules from transformational generative grammar and link them to semantic phenomena. 80 4.2.3. Terrell and Hooper/Terrell Although Terrell and Hooper (1974) also take into account the interacting role of syntax and semantics in their analysis of the subjunctive, they caution against basing analyses on T- G grammar, pointing out that transformationalists have attempted to work with this type of analysis but 'encountered difficulties because the choice of mood involves a complex relationship between syntax and semantics, and no comprehensive semantic theory is yet available in the transformational framework' (484). Terrell and Hooper argue for a semantically based analysis of mood where 'the choice of .mood in Spanish is directly correlated with what the sentence as a whole expresses about the truth of the proposition included. in the sentence' (484). Based. on a semantic hypothesis they set up a classificatory system of six sentence types according to attitudes speakers adopt towards the semantic notions of assertion and presupposition. The system they propose is the following: fl_1/ W CLASS M9912 ~ ASSERTION (1) Assertion Ind (2) Report Ind 1 PRESUPPOSITION (3) Mental Act Ind (4) Comment subj NEITHER (5) Doubt subj (6) Imperative subj (488) 81 The following are examples for each class: (1) Assertion: Sé que va a ir con nosotros. 'I know that he is going to go with us.‘ (2) Report: Nos dice que Maria quiere jugar tenis. 'They tell us that Mary wants to play tennis.‘ (3) Mental Act: Me parece que Ud. debe de quedarse aqui. 'It seems to me that you should stay here.‘ (4) Comment: Es.maravilloso que estudie tanto. 'It's marvelous that she studies so much.‘ (5) Doubt: Nb es seguro que vaya con nosotros. 'It's not certain that she's going with us.‘ (6) Imperative: Quiero que nos quedemos un rato mas. 'I want us to stay a little longer.’ (486-487) The authors explain that the semantic difference between classes (1) and (2) and classes (5) and (6) is clear but 'the existence of classes (3) and (4) shows that we have not been able to associate the choice of mood consistently with a certain semantic notion' (488) such as assertion and presupposition. They argue that it is not that their analysis is faulty but that this is an area that may be in a state of instability. Some native Spanish speakers accept indicative 82 complements in all types of presupposed complements so that sentences such as Me sorprendié que vino are just as acceptable as Me sorprendié que viniera ('It surprised me that they came') (488) . They suggest that if the system stabilizes, it is likely to do so by choosing the indicative for class (4) and then the semantic notions of assertion and presupposition would be associated with the indicative and the lack of these notions with the subjunctive. Other studies (Garcia and Terrell 1977 and Lantolf 1978), confirm that native speakers more readily accept the indicative forms in complements following comment matrices. In a subsequent article, Terrell (1976) continues to maintain that the syntactic properties of mood in Spanish 'are directly .related to the semantic notions of assertion and presupposition' (239) . He points out that what he and Hooper attempted to show was that with sentential complements, the relation of both the mood of the verb in the complement and the matrix into which the complement is embedded is dependent upon this factor of assertion, and that the choice of mood is meaningful and not transformationally derived (221) . However, he seems to have found a solution to the problem of the area of instability in presupposed complements which Hooper and Terrell left unresolved. He defines the term 'assertion' as 'a proposition expressed in a declarative sentence' (224) and proposes that the notions of assertion and 83 weak presupposition are compatible, and that is why the indicative is used in these two cases. On the other hand, the subjunctive is used in cases of nonassertion and strong presupposition. The difference between weak and strong presuppositions is that 'in sentences with strong presupposition, the complement is accepted as true under any conditions' whereas in sentences with weak supposition 'the truth of the complement cannot be inferred' (223). In example (10), the complement is true under any condition and is stated in the indicative. In (11), the complement is in the subjunctive and therefore its truth value cannot be inferred. (10) Nb supe que se habia cancelado e1 vuelo. 'I didn't find out that the flight had been canceled.’ (11) .Mo me sorprende que hayan podido hacer e1 viaje. 'It doesn't surprise me that were able to take the trip.’ (223) Terrell believes that by defining these terms in this way and by examining syntactic correlates other than mood he can show that (1) weak presupposition is treated syntactically and semantically as a type of assertion and that (2) the syntactic processes involved may be explained by the semantic properties of the class 84 of matrices to which they are restricted (226). Terrell examines the properties of matrices with six different types of verbs: factives (strong presupposition), semifactives (weak presupposition), opinion (announcing assertion), reporting (indirect assertion), doubt (lack of assertion), and volition (command). In examining the effects of negation on different verbs, he shows that, in general, negating a nonassertive matrix (containing verbs of faction, volition and doubt) does not affect the complement (except in cases where the negated natrix is equivalent to a positive .opinion matrix, such as negating dudar). However, in cases of negating opinion. matrices, semifactives and. matrices of reporting, the situation is more complex: the effect of negating depends on the asserted proposition's relationship with the matrix. Terrell gives the following examples: (12) .Nb era obvio que fuera tan importance. 'It wasn't obvious that it was so important.‘ (13) .Nb era obvio que era tan importante. 'It was important, but it wasn't obvious that it was.‘ (231) In (12), the negated form of the opinion matrix (ser obvio) transforms it into a doubt matrix and the subjunctive is used in the complement. The speaker both denies that the proposition was obvious and has doubts about the importance of 85 the proposition. In (13), the speaker asserts the proposition, asserts that it is true, but negates its relationship to the matrix. In examining the behavior of sentences undergoing complement Preposing (CP), (Bolinger's 'postposed main phrases') Terrell concludes that 'all classes of assertive matrices allow Complement Preposing' (236) in Spanish, whereas nonassertive matrices do not because the effect of CP is to make the complement proposition the main assertion of the sentence and nonassertions cannot be asserted, as evidenced in the following examples: (14) *Lo buscé en el diccionario, dudo. *'He looked for it in the dictionary, I doubt.' (15) *No lo encontraras, quiero. *'You won't find it, I want.'(236) 4.2.4. Garcia and Terrell and Lantolf In a later study, Garcia and Terrell (1977) investigated whether or not the Spanish system of mood is, indeed, in a state of change, and hypothesized that the semantic and syntactic rules governing the use of 'mood must be due to variable constraints. The authors found, as predicted by Terrell and Hooper's classificatory system (1974), that the indicative form.is most acceptable in complements following matrices which express subjective comment (with presupposed complements) and least acceptable in imperative sentences. In 86 between, they found that indicative complements following matrices expressing doubt are accepted to a degree depending on the relative intensity of the assertion-doubt matrix. They also found that indicative complements expressed either in the preterite or future indicative were more acceptable because the subjunctive does not have these tenses. Complements expressed in the present, imperfect, present perfect and past perfect indicative forms were not likely to be accepted because these forms do exist in corresponding subjunctive forme. Their subjects were Mexican and Mexican-American high .school students living in the U. S.—Mexican border cities of El Paso, Texas and Ciudad Juarez, Chihuahua, and Mexican university students living in Mexico City, and they concluded that the mood distinction is being lost more rapidly among Mexican American speakers, because of fewer normative pressures to follow the prescriptive nonm. Lantolf (1978) also argues for a semantic theory of mood providing evidence of his own and others (specifically Garcia and Terrell's (1974) study) that 'mood is not syntactically ‘h—‘I’ determined by the m§§§1x1X§£§1§EE is dependent upon the type -..... --—.. ',~_',—..——‘-~—o-"‘"-._fl._o —. — ofinfonmationfthewspeaker desires to convey about a specific hum—fl ’...a-M." _,_,..—‘--— mun-...»u-z.» u , Tu; W proposit‘igp‘immll) and that 'mood is meaningful even in F“ sentences of the volitional category' (211) when prescriptive pressures are strong for use of the subjunctive. Lantolf gathered his data by distributing questionnaires to Spanish-speaking Puerto Rican Americans living in 87 Rochester, New York. He looked specifically at these subjects' acceptance of indicative complements following matrix verbs of volition, doubt and presupposition. Prescriptively, these kinds of complements would carry the subjunctive mood. He found that subjects tended to select the subjunctive mood for complements following matrix verbs of volition in a greater percentage than for other verb types. He found more variability for mood selection and a greater percentage of acceptability for indicative complements following verbs expressing doubt. He mentions that some of the sentences of the doubt category which prescriptively call for subjunctive are in reality more affirmative than doubtful... [and that speakers]...in the absence of normative pressure,... are freer to react to the partially assertive nature of such sentences and choose the indicative mood (201). As he had predicted, based on Garcia and Terrell's findings, in the presupposition class he found the 'subjunctive is substantially diminished in frequency of occurrence...more than in any other semantic class' (206-7) because, he explains, 'presupposition is the closest to assertion' (202). This is support for Hooper and Terrell's claim.that it is in this category that there is the greatest instability and that if the Spanish subjunctive is in danger of giving way to the indicative, it will begin in the presupposition category (Lantolf: 203). Additional support for this claim can be seen in the non-linguistic constraint of 88 age in the selection of mood: younger speakers were more likely to allow indicative complements in all categories than older speakers. Finally, he found, as Garcia and Terrell did, that complements which displayed the highest degree of indicative were those that had. an embedded 'verb in the jpreterite (following a preterite matrix verb). He explains that presuppositions deal with facts, as assertions do, and those facts are confirmed by being marked in the past tense. 4.2.5. Goldin and Bell The notion of presupposition is a concept upon which Goldin (1974) also bases his analysis of the subjunctive; however, his use of the team is unique to him - he doesn't use it as other linguists do. He suggests there are two ordered principles which native speakers use in their selection of mood; they are: (1) the Reaction Principle, and, (2) the Presupposition Principle. The first depends on the main verb, which, if it expresses a reaction, (alegrar, 'to be or make happY'; sentir, 'to regret'; ser léstima, 'to be a shame'; etc.) requires the verb of the subordinate clause to be in the subjunctive. If the main verb does not express a reaction then the second principle comes into play. The Presupposition principle states: 'When a speaker has positive presupposition about an event or state in a subordinate clause, he uses the indicative mood. When his presupposition is negative or indefinite, he uses subjunctive mood' (297). This definition 89 suggests that Goldin's use of the temm is closer in meaning to the word, 'attitude' than presupposition. Goldin states that these two principles do not require (of the learner, I assume) any knowledge about grammatical categories or structures other than the fact that the subjunctive only occurs in the subordinate clause of complex sentences. He claims that the variation between use of the indicative and subjunctive 'can be described simply and easily using the general principles of reaction and presupposition' (300). Goldin's account of the subjunctive is aimed at giving .classroom.teachers and students of Spanish an understandable formula for being able 'to produce and understand sentences' in the subjunctive. However, as Bell (1980) implies, Goldin has sacrificed 'the complexity of linguistic theory...to the immediate aim.of finding better ways of teaching the Spanish subjunctive' (Bell 1980: 377-78). Although Goldin's account is simplistic, offers little to linguistic theory, and uses confusing terminology, it's main point is that at least there is shown a need to account for and point out to learners that the subjunctive is not explicable only in terms of syntax; there is more to the use and explanation of the Spanish subjunctive. [In his review of the recent examdnations of the Spanish subjunctive, Bell, although recognizing the contribution to current linguistic theory of accounting 'for the phenomena of :mood in terms of as few semantic and/or syntactic principles 90 as possible' (382), rejects unitary accounts that attempt to explain the uses of the subjunctive based either on their common negative characteristics or their common positive attributes. And, he concludes, that any explanation of the use of this structure must incorporate both multiple semantic and syntactic features. He states: I think it is important to recognize that not just one semantic feature is involved, but rather a range of features, with a range of semantic functions (389). He comments that these recent studies have revealed that 'the Spanish subjunctive is 'responsive to quite basic semantic f‘" -5. ,‘r’ --.......-—o —>.-.4-Q-h--'- values expre381ng the 1nformat1on content of an utterance: what is asserted by its utterance, and what is presupposed; how 1s a particular noun phrase referred to; what is the logical truth value of a particular sentential complement; etc? ”(377)W“4Bellflma1nta1ns that a comprehensive account of mood in a language -- in this case the Spanish subjunctive -- far from being relegated to pure semantics, is a necessary part of any linguistic theory. 'A proper description of Mood is obviously an essential part of an adequate linguistic theory, as much as of the restricted description of Spanish grammar' (377). He points out that comment type sentences and non-comment sentences behave differently. The presuppositions stated in the complement of comment sentences are not affected when the 91 main verb is negated, as in (16) and (17) below: (16) Siento que se haya roto e1 plato. 'I'm sorry that the plate has broken.' (17) .Nb siento que se haya roto e1 plato. 'I'm not sorry that the plate has broken.' (379) In non-comment sentences, however, negating the main verb does change the truth value of the complement, as seen in (18) and (19) below: (18) EB cierto que el plato se rompié. 'It's true that the plate has broken.' (19) .Nb es cierto que el plato se haya roto. 'It's not true that the plate has broken.' Bell says that the difference in Comment sentences is that there is semantic separation between the complement and its matrix. In Bolinger's analysis the main clause can only be postposed by making two separate sentences, i.e. 'The plate broke. I'm.sorry.' In comment sentences the matrix can be negated without changing the presupposition of the proposition of the complement. In non-Comment sentences, however, the complement depends directly on the matrix and when the matrix is negated, the subordinate clause is semantically affected (379). However, as Lantolf confirms, native speakers do accept complements in the indicative after matrices of doubt, 92 in which case, non—comment sentences would 'behave' in the same way as comment sentences. Bell's analysis does not take this factor into account. Nevertheless, he does emphasize that syntactic factors alone are not sufficient in accounting for the use of mood in Spanish. He concludes the choice of mood comes not from any purely syntactic factor but from the strength of the speaker's commitment - or rather his non-commitment - to the truth of the complement. ...the strength of the speaker's commitment is expressed by means of a particular matrix verb, with its own peculiar semantic force, and the choice of moodi ... The two factors, matrix verb and.mood, interact and combine to produce a singular semantic effect (383-384). 4.2.6. Takagaki and Reider Takagaki (1984) also claims that what governs the use of the subjunctive mood in the complement is some semantic factor in the main clause. To explain this governing, Takagaki introduces the term 'independency', which is 'the quality of a proposition “affirmatively evaluated” and also 'stated", (251) and contrasts it with 'subordinance', which, he states, has not yet been assigned 'independency'. Propositions which are affirmatively evaluated (whether or not they are stated in the affirmative or in the negative) and have been stated, are expressed in the indicative as in (20) and (21) below: (20) La muchacha es bonita. 'The girl is pretty.‘ (21) La muchacha no es bonita. 93 'The girl is not pretty.‘ (250) Takagaki hypothesizes two levels of meaning assignment: the first one, which represents the logical structure of the proposition, is at the level of logical semantics, and the second is the surface semantic interpretation. He presents the logical semantic structures of (20) and (21) as (22) and (23) below, respectively: (22) A = B.(where A is la muchacha and B is the adjective bonita) (23) A =~B (1a muchacha = no bonita) (251) which, at the level of surface semantic' interpretation structure is formally represented in the indicative. In (24) below: (24) *La muchacha (sea, ser, siendo, sido) bonita. (251) Takagaki explains that the 'stating' force has not yet been achieved. He maintains that 'the subjunctive form.at this level corresponds roughly to the logical structure immediately prior' to the surface, ‘which. has not jyet been assigned I'independency"' (251). The fact that the subjunctive is not found in independent sentences is further evidence for the hypothesis that the 94 subjunctive is the marker of subordinance and not of independency (251). Takagaki regards commands as 'variant of complex sentences with the main clause omitted either completely...or partially' (251); and in sentences commencing with tal vez or quizé, etc. followed by subjunctive complements, he considers the adverbs 'as a kind of "semi- matrix“ and ... [these kinds of sentences] another subtype of the complex sentence' (251). He claims that the sentence, (25) Creo que viene. 'I believe he's (she's, it's) coming.‘ is a coordination of two simple declarative sentences in its underlying logical structure; and the sentence, (26) .Nb creo que venga. 'I don't believe he's (she's, it's) coming.‘ has the underlying logical structure of a complex sentence. If we interpret the sentences at the surface level, the complement of (25) 'is assigned independency to form the juxtaposition of the two independent sentences, ... [and in (26) above]... the complement remains unprovided with this pragmatic force' (252). Takagaki's analysis supports Rivero's claim that indicative and subjunctive complements derive from.different ‘underlying structures and refines Bolinger's analysis that 95 only independent main phrases, in Takagaki's term, can be postposed. It was Bolinger who pointed out that the main phrases of Spanish indicative complements could be stated independently and therefore postposed in their English translations; the main phrases of subjunctive complements could not be stated independently and could not be postposed. Using Takagaki's model, Reider (1989) presents evidence for a semantically based analysis of mood and clitic promotion in Spanish complementizers. He shows how certain restrictions for clitic promotion have a semantic explanation based on the type of proposition expressed in the verbal complement. These semantic features are those which have been used in analyses to characterize the semantic interpretation of mood choice in Spanish. He states that the close correlation between the mood of the complement clause and the ability of clitics to move from the complement to the matrix verb is aphenomenon which has certainly not gone unnoticed...[and that]... the semantically-based hypothesis advocated...provides a reasonable explanation as to why the correlation between Clitic Promotion and mood holds true for some sets of matrix verbs but not for others (284). Based on Takagaki's hypothesis that propositions expressed in the indicative have been assigned 'independency' at the level of surface semantic interpretation in the sense that they have been 'affirmatively evaluated' and are 'stated', Reider points out that propositions that are stated in the subjunctive are lacking in these features, and that it 96 is precisely these semantic features which constrain clitic promotion as well (290). He concludes: ...just as semantic interpretation affects the choice of mood in Spanish verbal complements, it also plays an essential role in restricting the promotability of clitic pronouns. In particular, the correlation between Clitic Promotion and mood in Spanish has been explained in terms of two semantic features assigned to the complement proposition, [evaluated] and. [stated], whose respective values are inherently determined by the meaning of the governing matrix (293). In a later work, Reider (1990) continues to draw on Takagaki's independency hypothesis and 'brings in Chomsky's trace theory to account for the different underlying structures and the derived semantic interpretations of indicative and subjunctive complements. His analysis supports both Bolinger's observations and Bell's claim that negation belongs semantically to both the main clause and the complement. Basically, his claim is that when a negative is moved out of a subordinate complement to the matrix clause through the Neg-transportation rule, a 'neg-trace' is left behind. Then, 'if a verbal complement from.which no has been extracted by NT still retains a trace of the negation in derived structure, then that complement cannot achieve independency...and will therefore be expressed in the SUbjunctive' (216). This explains the difference in structure and meaning of the following two sentences: 97 (27) Nb creo que el profeta vuelve. 'I don't think the Prophet will return.‘ (28) .Nb creo que el profeta vuelva. 'I don't think the Prophet will return.‘ (218) According to Reider, in (27) the proposition stated in the indicative states someone else's belief, but the speaker denies it. In (28), the speaker expresses his belief that the proposition is not true. (28) has undergone NT and therefore contains a neg-trace in the complement, which at the level of surface semantic interpretation is expressed in the subjunctive. (29) has not undergone NT, the negation originating in the matrix. Reider proposes the Neg-trace Condition (NTC) to explain this. The NTC states , in effect , that the speaker...cannot be committed to the truth of P, if P contains a neg-trace —- a trace of raised no which, although phonetically null, is interpreted as a denial of the truth of P. Since P containing a neg-trace does not get affirmed, therefore it remains subordinated and is expressed in the subjunctive (218). By utilizing Chomsky's trace theory and Takagaki's 'independency' hypothesis, the NTC appears to account for Conflicting data and in this way 'serves as a crucial link between the syntactic and the semantic components in the grammar of Spanish' (221). 98 4.3. Semantic/Pragmatic Studies In the previous section only the Garcia and Terrell (1977) and Lantolf (1978) studies analyzed the way native speakers actually use the Spanish subjunctive; and this was at the sentence level. In this section we review studies that have analyzed how native speakers use the subjunctive in discourse contexts, either spoken or written. These studies examine the relationship among syntax, semantics and pragmatics. They include those of Lavandera (1983, 1984) who concludes that the subjunctive mood in Spanish is used by native speakers as a discourse strategy in developing an argumentative style; Lunn (1989a, 1989b) who claims that many phenomena related to mood choice can only be accounted for by examining the discourse context in which speakers mark less- than-optimally relevant information with the subjunctive, Krakusin and Cedeno (1992) who claim: that mood. choice following e1 hecho de que is constrained by the information value of the subordinate proposition, and by Mejias-Bikandi (1994) who argues how a redefinition of the term.'assertion' based on speakers' intentions accounts for mood choice. 4.3.1. Lavandera Lavandera (1983, 1984) examines the use of the subjunctive in discourse and maintains that a great deal of the complexity of the uses of the subjunctive derives from.the exploitation of the forms as a discourse strategy among native speakers. In examining a set of texts where the subjunctive 99 is used, she accepts the semantic analysis that attributes the semantic feature of [+assertion] to the indicative mood and [- assertion] to the subjunctive mood but considers the analysis still 'insufficient for my purposes' (211). She feels an analysis of the subjunctive should account for why a speaker would 'make statements containing nonassertive meanings at all' (211). Specifically, Lavandera attempts to establish: 1. How the [-assertive] modality signaled by the mood morphology matches the lexical expression of modality in the neighboring environment, 2. What might motivate the apparent redundancy resulting from grammatical and lexical expression of modality, 3. At which points in discourse switches to utterances in the subjunctive mood occur, 4. How these utterances relate to the background utterances in the indicative mood (1983: 211-212). In order to answer these questions, she examined the texts of recorded face-to-face interviews of Spanish-speaking subjects from. Buenos Aires, Argentina. The interviews consisted of conversations about the subjects' own feelings and beliefs related to the Argentine way of life, traditional family practices and the current economic situation in the country. She looked at the background of narration in which the subjunctive was used, with special attention to when, where and why the subjunctive was used and how it related to other linguistic features such as lexical signals and in In 100 intonation. What Lavandera found from the text analyses was that the modality of the utterance is often expressed by both grammatical and lexical signals and that in order to get an acceptable stretch of discourse, the different signals must match. She gives the following examples taken from the texts: (29) .mientras que a vos no te falte (subj.) nada, como vos decis 'as long as you don't feel (subj.) deprived and anything, as you say' (30) y a mi no me molesta (1nd.) dértelo en absoluto 'and I don't mind (1nd.) giving it to you at all' In (29), both the lexical signal of 'hedging' in mientras que, 'as long as', and the shifted responsibility message of coma vos decis, 'as you say', fits with the non-assertive mode of the subjunctive. In (30), the assertive meaning of en absoluto, 'at all', fits with the assertive mode of the indicative. Both the linguistic environment of the moods and the lexical signals match. In addition, in the texts examined the utterance with the subjunctive mood is often either a repetition or an anticipation, in a different form, of what has already been, 01? later will be, asserted. Lavandera gives examples from each of the four texts she discusses. The following is an 101 example from one of the texts: (31) .Nb es porque yo vea (subj.) mal el comunismo. 'It's not because I disapprove of communism.‘ (32) no lo veo (1nd.) mal 'I don't disapprove of it' Lavandera explains that the shifting back and forth of moods in discourse and the use of the subjunctive to bring up an issue and dismiss it as irrelevant exemplifies one form of the argumentative style in discourse. This shifting back and forth is a strategy which consists of supporting a claim.with an apparently complete list of all relevant facts. Some facts are asserted, others are simple raised and disposed of quickly with no conclusion being drawn from them. . . the examples in the subjunctive...show how they participate in creating an argumentative style, and how their insertion relates to the stage of the argumentation that has been reached (231). Lavandera further explains that the morphology of the moods...is used to discriminate among utterances that refer to events and conditions in terms of more or less 'relevance'; utterances with the subjunctive place the facts that they describe at the bottom of the scale of 'relevance', and such hierarchization is exploited in the organization of texts (231). By being placed at the 'bottom. of the scale' of 102 'relevance', these facts cannot be relied upon. Lavandera maintains that the subjunctive mood is used to introduce material which the speaker does not want to omit, which is necessarily true of everything we say, but to which he does not want to draw too much attention or have the listener to rely upon too much. Lavandera adds that although linguistic environments such as temo que..., dado que..., es posible que... make explicit the non-assertiveness of the statements which. makes the use of the subjunctive redundant, the 'morphology provides a quick and condensed linguistic means of anticipating information that may be developed elsewhere in the text and the context' (233). Besides, Lavandera points out, employing the indicative in such linguistic and semantic environments would be contradictory. The type of analysis Lavandera has carried out reveals properties of the Spanish mood system that previous analyses at the clause and sentence level could not account for. The subjunctive is not only used for propositions that cannot be asserted, or which have not attained (syntactic clause) independence, but is used to refer to events the speaker feels are not relevant to his argument, but which he cannot leave out . 4.3.2. Lunn Lunn (1989a) also argues for a pragmatic analysis of the Spanish mood system at the discourse level. She points out that the Spanish subjunctive is partially grammaticized in 103 that certain matrix verbs require the use of the subjunctive in the complement. She states that Hooper and Terrell's (1974) semantic analysis of the subjunctive based on the notions of assertion and presupposition 'inspired a boom in the semantic description of Spanish mood' (Lunn 1989b: 689). Their analysis helped to account for data which had previously been treated as exceptions to the syntactic rules, but did not break with the tradition of analysis at the sentence level. Lavandera's (1983) analysis provided another turning point in analyzing the subjunctive. Lunn further explains: Pragmatic analyses such as Lavandera's have paved the way for a prototype analysis of the subjunctive. Seeing the subjunctive as an option which speakers may choose to exercise opens up the possibility -— indeed, the necessity -— of developing a description of all uses of the morphology (690). She also points out that All linguistic analyses of the subjunctive assume that context constitutes the justification for mood choice. What has changed over the years is the definition of context. Syntactic analyses of the subjunctive look at main-clause context, semantic analyses at whole-sentence context, and pragmatic analyses at discourse context. Much of the data in the arguments to follow can only be explained by discourse context, so the explanation to be developed is necessarily pragmatic in nature. But the larger argument can be made that what is necessary to explain the data here also serves to explain the syntactically and semantically conventionalized uses of the subjunctive (250). 104 A pragmatic analysis of the choice of mood can be explained by a single generalization which 'correlates mood choice with the information value of clauses: verbs in clauses of relatively high information value are marked with the indicative; verbs in clauses of relatively low information value are marked with the subjunctive' (249). This explanation of mood choice rejects solely semantic analyses of the Spanish subjunctive in that it 'acknowledges the pragmatic role of the speaker in mood choice...[where] speakers...use subjunctive and indicative as discourse organizers' (249). It rejects grammatical analyses because, as Lunn argues, the indicative/subjunctive mood choice is not completely grammaticized but is left up to the prerogative of the speaker. Based on Sperber and Wilson's 'Principle of Relevance' (1986), which argues that 'relevance is an organizing factor in linguistic communication' (Lunn 1989a: 250), Lunn maintains that speakers mark less-than-optimally relevant information by using the subjunctive. To support her arguments, Lunn provides evidence from.literary works, journalistic writing and advertisements, among other examples. In the evidence she points out that where grammatical and semantic analyses of mood would predict the use of the indicative, the speakers (or writers) employ the subjunctive, which can only be explained by a pragmatic analysis of 'relevance'. In journalistic Spanish, writers often use the subjunctive to mark information that readers already know (or should know); as Lunn points 105 out, this 'potentially assertable but un-newsworthy information' (255) is low in contextual relevance. In advertising, the subjunctive can be used to concede truthful information but minimize its importance. In other examples, it is shown that speakers can choose to mark certain information as low-priority by using the subjunctive. In the following example, the 'relevance' explanation can explain the two readings given for the Spanish utterance: (33) Aunque esté forrado el tio, no me casaré con él. a. 'Although the guy might turn out to be loaded, I won't marry him.‘ b. '80 what if the guy's loaded? I won't marry him.‘ (256) In (33a) the speaker is not asserting that the guy is rich, only that he might be. In (33b) the speaker is asserting that the guy is rich but rejects the fact as important. An analysis based on assertion and presupposition cannot account for the two different readings, but a relevance analysis can. Example (23) from chapter three is repeated below as (34) and examined in light of this new analysis: (34) vamos a salir aunque haga frio. a. 'We're going out although it may be cold' or b. 'We're going out despite the cold' (Olivella de 106 Castells 1982:209). Although the text gives two alternative English glosses for the sentence, it does not explain why there are two. The relevance analysis accounts for the two different readings in the following manner. Sui (34a), the speaker concedes the possibility that it's cold outside; in (34b), the speaker concedes the reality of its being cold outside but denies that this is important; they will go outside anyway. This analysis can also explain why native Spanish speakers sometimes use the indicative when the subjunctive is called for prescriptively. Lunn presents an example from.the novel El beso de la mujer arafia by Argentine author Manuel Puig: (35) .Me da léstima que se terminé. 'It makes me sad that it's over.‘ (257) This is a sentence in which the main clause is a comment on presupposed subordinate clause information. This kind of variable mood usage has been pointed out by others (Terrell and Hooper 1974, Lantolf 1978, and Garcia and Terrell 1977) as being in a state of instability. In a relevance analysis, the speaker's using the indicative signals the hearer to attend to the information. Lunn explains that 'When indicative is analyzed as an instruction to hearers to attend to a piece of information, it is possible to explain why a speaker mdght 107 choose to mark previously-mentioned information with the indicative' (257). Speakers have the option of insisting on the continuing relevance of redundant information. Finally, Lunn points out (endnote, p. 258) that this analysis of the Spanish subjunctive could also explain why it is a late-learned structure by both L1 children and classroom L2 learners. 'Subjunctive endings appear on verbs in clauses that a speaker has chosen not to emphasize... It is not surprising that the morphology in de-emphasized clauses is not noticed - and so not acquired - until late in the acquisition sequence' (Endnote 3, p. 258). In a subsequent article, Lunn (1989b) claims that both untrue and true but concurrently presupposed information share the same quality of being less-than-optimally relevant information and, therefore, unassertable and marked by the subjunctive mood. Here Lunn looks at what characterizes subjunctivizable information and claime that a 'prototype of assertability' analysis of the Spanish subjunctive can explain a lot of data that in.syntactic and semantic analyses have to be treated as exceptions, anomalies or ungrammatical usages. The analysis explains: (1) the use of the -ra subjunctive to background information; (2) the use of the subjunctive in journalistic writing to mark true information that readers can be expected to know (especially the use of the subjunctive in nonrestrictive relative clauses, which. most prescriptive grammars do not allow); (3) the use of the subjunctive in other syntactic environments where it usually does not appear 108 (e.g. main clauses: quisiera (past subj.) .hacerle una pregunta); (4) the contrast between the -ra and -se past subjunctive foams, the latter serving as the less assertive, more formal and polite form, and its diminishing use; (5) the use of the subjunctive in aunque clauses to convey both meanings of 'true-but-uninformative' and 'untrue-and— therefore-uninformative' (e.g. Aunque sea .mi hija, la encuentro muy guapa, which has the two English readings: 'Despite the fact that she's my daughter, I find her very pretty' and 'Although she may be my daughter, I find her very pretty'); and, (6) the use of the past subjunctive in si clauses to express contrary-to-fact information. In these clauses, based on Lunn's analysis, the speaker does not assert the converse of the proposition; the listener has to infer it. Lunn concludes that the Spanish case shows how recourse to prototype descriptions can clarify a seemingly intractable puzzle. Except in their negative relationship to the prototype of assertability, the categories of subjunctivizable information cannot be explained by a single generalization (691). The prototype of assertability analysis clarifies the relationship between presupposed information and untrue information which other analyses of the Spanish subjunctive mood could not. 109 4.3.3. Krakusin and Cedeno Krakusin and Cedeno (1992) make use of Lunn's (1989b) prototype analysis in explaining the variability in the use of the subjunctive mood in factive clauses which follow the syntactic frame e1 hecho de que, 'the fact that' . In their analysis of the writings of the Mexican journalist and colummist, Mariano Grondona, they found that he used the indicative following e1 hecho de que when these clauses contained propositions of high informative value (usually when they came from sources which the writer did not wish to question) as in the following example: (36) Esta tesis ignora el hecho decisivo de que la gran mayoria de norteamericanos xix: arriba y no abajo de la frontera de la pobreza (15/29 de diciembre de 1986) (1291). 'This thesis ignores the decisive fact that the large majority of Northamericans live (1nd.) above and not below the poverty line.' [MLQ's translation] The subjunctive, however, was used to mark factual propositions to which the writer did not want to draw much attention on the part of the readers, as in the following example: (37) Desde la perspectiva de un moderado como Sanguinetti, e1 hecho de que los fundamentalistas organ en el Estado, no los separa tanto como los une (1291). 'From the perspective of a moderate like Sanguinetti, the fact that the fundamentalists 110 believe (subj.) in the State, doesn't separate as much as it unites them.’ Krakusin and Cedeno conclude that the variability in mood choice is constrained by the information value of the subordinated proposition, i.e. the selection of mood is pragmatically determined by the speaker's (or in this case, the writer's) communicative intentions. 4.3.4. Mejias-Bikandi Mejias-Bikandi (1994) also develops an analysis of the subjunctive mood in Spanish based on assertion. However he contends that, rather than appealing 'to different degrees of assertion to accommodate the use of the indicative mood in complement clauses' (900) as Terrell and Hooper do, or to the notion of relevance as Lunn and Lavandera do, what is needed is a redefinition of the term 'assertion'. He explains whether a proposition is asserted or not depends not so much on whether that proposition is true or false, but on what are the intentions of the speaker' when s/he decides to jpresent the information expressed. by the proposition to a particular audience (892). By bringing such notions as intentions of the speaker into the definition of 'assertion', the problematic data of Terrell and Hooper's analysis for presupposed clauses is taken care of. It should be remembered that Terrell and Hooper 111 wanted to correlate the indicative mood with the notion of assertion and the subjunctive mood with the notion of non- assertion. However, sentences that are presupposed are also not asserted, yet often take the indicative mood. Mejias- Bikandi points out that for the two sentences (38) Pedro se ha dado cuenta de que tienes razén. 'Pedro has realized that you are right.' (39) Pedro se alegra de que tengas razon. 'Pedro is glad that you are right.' (896) the complement of both sentences is presupposed and not asserted; however, the complement of (38) is in the indicative and the complement of (39) in the subjunctive. The intention of the speaker, however, is not the same for the two sentences. In (38) the speaker intends to assert the proposition of the complement clause as part of Pedro's view of reality, even though the proposition is logically presupposed. In (39) it is not the intention of the speaker to present the complement clause as part of Pedro's or the speaker's view of reality. This proposition is not asserted and is expressed in the subjunctive mood. IMejias-Bikandi's redefinition of the notion of assertion, based on the pragmatic notion of speaker's intentions, allows Terrell and Hooper's distinction to be maintained: that the indicative is associated with the notion of assertion and the subjunctive with non-assertion. 112 4.4. Conclusions The role of semantics and pragmatics in the description and explanation of the use of mood in Spanish in these studies cannot be diminished. In the earlier studies, although most agree that semantics governs how the complement verb will behave, each investigator views the role of the matrix verb in different ways. Although Bolinger, Rivero and Takagaki all agree that indicative and subjunctive complements derive from different underlying structures, Takagaki maintains that it is the matrix verb which governs mood choice in the complement; Rivero claims that mood choice depends on the nature of the complement verb itself; and Bolinger, along with Bell and Reider, sees mood choice as the result of an interaction between the matrix and complement verbs. Although Takagaki contends that the matrix verb is responsible for determining the mood in the complement, he also believes that the matrix verb accounts for the presuppositional nature of the complement. Terrell and Hooper also maintain that matrix verbs govern mood choice; however, this is because the matrix verb is associated with a certain semantic notion: either assertion or presupposition. Lantolf, although he never states it explicitly, seems to agree with Rivero that mood choice is not syntactically determined by the matrix verb but, rather, is due to what kind of information the speaker wishes to convey in the proposition of the complement. Lantolf finds that 113 prescriptive rules are not always followed by native speakers, who seem to treat the semantic notions of assertion and presupposition as basically the same. As can be seen from these studies, although different authors claim that semantics interacts in different ways to produce either indicative or subjunctive verbal complements, the majority sees the use of the subjunctive as a semantic notion manifested syntactically in a certain morphological ending in subordinate clauses. This semantic notion is related to [-assertion], which is shared by both the matrix verb and the complement verb. Whatever their conclusions, the fact is that none of these studies could leave out the importance of examining semantic phenomena in accounting for the Spanish mood system. Recognizing this has helped to account for the Spanish mood system. However, many questions remained unanswered about the use of the Spanish subjunctive. The most recent studies have included examining pragmatic features of the use of the subjunctive and how these interact with the syntactic and semantic phenomena already studied” Lavandera, in. her analysis of spoken language, finds that the modality of the utterance is expressed in a multiple way by syntactic, lexical and momphological cues and that these signals must match. Most importantly, Lavandera identifies a particular style in discourse and the exploitation of the subjunctive marker as a linguistic means of developing that style. Lunn builds on Lavandera's notion of relevance in looking at am pre but the prc thi to 91 ass tha may any ‘ tha Stm an aha: 114 at discourse units in written language. She suggests an analysis of the subjunctive based on the speaker's (or writer's) judgement of the relevance of the propositions presented. From this analysis she notes that untrue and true but presupposed information share the same notion of less- than-optimally relevant information and from this develops a prototype of assertability of the subjunctive mood to capture this generalization. Krakusin and Cedeno find this analysis to be true in their own study of complement clauses following e1 hecho de que. Finally, Mejias-Bikandi attempts to refine the notion of assertion in explaining mood choice by bringing in the idea that despite the fact that information may be presupposed, it may be the speaker's intention to assert that information anyway, thus again taking up Terrell and Hooper's assumption that assertion is correlated with the indicative and non- assertion with the subjunctive. What can be noticed from the foregoing review of these studies of the Spanish subjunctive is that no account can give an adequate description if they use only one level of analysis. The syntactic accounts, based mostly on the notions of the dependency/independency of the two clauses, could not avoid discussing how the semantic nature of the matrix verb determined mood in the complement. The semantic accounts were based principally on the notions of assertion or presupposition; however, it was shown that by integrating pragmatic notions, such as a speaker exercising his/her pre or the ind one to see sub rel: lea: 115 prerogative to mark less-than—optimally relevant information or to assert a presupposed proposition, generalizations about the subjunctive mood can be captured. It is seen that the nature of the subjunctive mood is indeed, a complex one -- one that must be studied at more than one level of analysis, to be sure, in order for the researcher to fully understand the structure being studied. It will be seen that many of these phenomena regarding the Spanish subjunctive mood, including assertion, presupposition, relevancy and speaker prerogative, cause difficulties for the learners in the present research study. CHAPTER FIVE: THE ACQUISITION OF THE SPANISH SUBJUNCTIVE 5.1. Introduction This chapter reviews studies, both child L1 and adult L2, which have examined the acquisition process of the Spanish subjunctive mood. Unfortunately, this is an area that has not been researched sufficiently; however, there are studies which have revealed and confirmed that the subjunctive is a difficult and late-learned. structure (Gili. y' Gaya 1972; Gonzalez 1975; Floyd 1990; and Stokes 1988, 1990), or that adult learners use the strategy of non-attention to verb morphology for producing and comprehending the subjunctive (Terrell, Baycroft and Perrone 1987; Lee 1987). One recent study claimed that learners are not able to generate the complex syntax, such as subordinate clauses, in which the subjunctive is required (Collentine 1995), contrary to what the present research found. The majority of the studies do not go much beyond a descriptive account of the acquisition process. Collentine does however, attempt to account for the data in his study by analyzing them in terma of Givon's model of language development (1990). .In the Blake (1991) study we begin to see a more detailed account of the acquisition process in which the multiple linguistic factors of syntax, semantics and pragmatics are taken into account and hypotheses are made concerning how these interact and influence the 116 117 acquisition process of this complex structure. Finally, Al— Kasey (1993) found that by distinguishing between factive and non-factive predicates taking the subjunctive, she could account for the similarities and differences between children L1 and L2 and adult L2 learners of Spanish and the interaction of syntactic and semantic features. In these studies, however, some parallels may be drawn with the present research. Some uses of the subjunctive which have been found to be late-learned for children, ie. not until eight or nine years of age, are not used by the adult subjects in the present study. However, there are also some differences; the children in Blake's study did use the subjunctive in adjectival and adverbial clauses, whereas the adult subjects in the present study did not, except for the adverbial cuando. .Al-Kasey's study also found some parallels and some differences between adult L2 and child L1 and L2 learners. 5.2. Review and Discussion of the Studies Gili y Gaya (1972), in a study of four to seven-year-old Puerto Rican children, found that younger children used the subjunctive only after optative verbs and after certain adverbial conjunctions such as a que 'so that', para que 'in order to', and cuando 'when'. He concluded that children do not acquire the more subtle uses of the subjunctive (after dubitative and factive verbs), where semantic and pragmatic contexts dictate usage, until they are much older, about eight 118 or nine years of age. Unfortunately, Gili y Gaya's study is a purely descriptive one and he does not provide any explanation as to what these 'semantic and pragmatic contexts' are, nor how they intervene in the acquisition process. It is unfortunate that one of the leading authorities in Spanish grammar does not discuss this. As was seen in chapter four, it is precisely these kinds of uses (the subjunctive following dubitative and factive verbs) that cause so many difficulties for researchers attempting to account for the nature of the subjunctive mood. The dubitative verbs include dudar que 'to doubt that', no creer que 'to not believe', etc. and the factive verbs include constructions such as e1 hecho de que 'the fact that'. It was seen in the previous chapter that adult native speakers vary in their use of the subjunctive in the complement verbs following these verbs according to the relevance of the proposition or according to their intention to assert that proposition. These are subtle semantic meanings and discourse markers in adult speech, and this may explain the lateness with which children acquire the subjunctive in these constructions. Surely an understanding of the acquisition of these uses of the subjunctive would complement the studies that have attempted to describe the semantic and pragmatic notions which influence subjunctive marking. Gonzalez (1975), in a study of bilingual children in Texas, found that his subjects were using the present subjunctive by the age of three and the past subjunctive by 119 the age of 4;6. His findings confirm those of Gili y Gaya. His subjects used the optative subjunctive (mostly after the matrix verb of querer que 'to want'), and after the adverbial conjunctions para que and cuando. Floyd (1990), in her review of the findings of other studies of syntactic development in Spanish as they relate to children's use of subordinate clauses within complex sentences reports that according to Merino (1976) '[subjunctives], which involved complex sentences, were among the most difficult structures, with significant differences observed by grade for children's production of subjunctives and conditionals' (Floyd 1990: 489). In adult second language acquisition of Spanish, Stokes (1988, 1990) found that former Mormon missionaries, after having spent between sixteen months and two years in the target language culture, became 'quite fluent in Spanish, but fail[ed] to gain control over the subjunctive' (705). However, Stokes also found that advanced students who had previously lived in the target language country and were studying Spanish at the university level profited more from instruction on the subjunctive than students who had never lived in the target language culture. It appears that formal instruction combined with immersion within the target language culture help learners acquire the subjunctive. In studies that examine adult classroom learners' strategies used in the acquisition of the Spanish subjunctive, Terrell, et al (1987) say that adult learners do not 120 productively use the subjunctive marker because for them it is redundant and does not add significantly to the content of the message in most cases. In addition, native speakers have no difficulty understanding the non-use of the subjunctive by non—native speakers. 'The choice of the verb form appears to be quite secondary to clause marking insofar as comprehension by native speakers is concerned' (27). The authors suggest that learners could be making the same assumption about the lesser importance of subjunctive verb forms in the comprehension of sentences. Lee's (1987) findings appear to support this view. In testing reading comprehension of the Spanish subjunctive by adult second language classroomt learners, he found. that 'learners who have never been instructed in the subjunctive mood are able to comprehend the meaning of sentences and/or noun clauses which contain this verb form' (54). Collentine (1995) found that for two groups of university students, at the end of their second year Spanish course, did not produce the complex syntax, such as subordinate clauses, in which the subjunctive appears. The two groups completed two different types of tasks. The first group participated in ten-minute conversations with the researcher who attempted to prompt the students to reply in the subjunctive. These learners produced the subjunctive only 13% in the contexts where it was required -— a finding very similar to those found in the present study for the subjects in the first interview. The second group completed highly controlled oral-production 121 tasks in which the subjects were required to respond to questions related to drawings shown to them. These subjects produced the subjunctive 34% of the time for required contexts. Collentine concludes that these learners were not ready to produce complex syntax and make the subtle morphological distinctions required to correctly use the subjunctive. However, in the present study it was found that subjects had no difficulties producing complex syntax where the present indicative was required. Collentine's study is important, however, in that it attempts to link syntactic and morphological features of the subjunctive in trying to understand how these learners use the structure. In his 1991 study, Blake attempts to link the complex interaction between syntax, semantics and pragmatics and takes 'them.into account in describing the complex process of the acquisition of the Spanish mood system. In his study, Blake looked at the acquisition of the Spanish subjunctive by 134 Mexican, middle-class children between the ages of four and twelve years and the use of the Spanish subjunctive by 39 adult Mexican university students. He found that these children do not acquire the subjunctive for different matrix verbs all at the same time, a finding similar to Gili y Gaya's. The children in Blake's study used the optative subjunctive with few errors by the age of four, and by the age of five or six years had begun to use the subjunctive with certain adverbial conjunctions (para que, cuando, hasta que, 122 etc.). By six years of age these children still had not acquired the subjunctive following dubitative verbs. Blake states: ...starting at five years of age (a statistically significant division) this dominion of adjectival clauses also includes the pragmatic distinction between an existing referent, marked by [+indicative], and an 'irreality', marked by [+subjunctive] (235). [MLQ's translation] Blake says that the use of a language influences the development of the competence in that language and should, therefore, be included in any description of its properties (231). He maintains that the Spanish subjunctive mood, in all of its uses, cannot be acquired by Hispanic children without taking into account semantic and pragmatic properties of the context of usage. ...the learning of the subjunctive implies understanding the context in which it is used as much as managing certain basic principles about the syntax (233). [MLQ's translation] He contends that if we cannot describe and/or explain the contexts which require the use of the subjunctive then we do not have an adequate account of the acquisition of the subjunctive by Hispanic children. Blake initially mentions that an ideal analysis of the subjunctive might state that subjunctive-requiring matrix verbs are subcategorized in the lexicon by a dependent clause 123 requiring an operator of [+ subjunctive]. This would work for the optative subjunctive, which is required following matrix verbs which express desire, commands or volition. This use of the subjunctive is obligatory, without exceptions, and is based on purely syntactic criteria. However, like other researchers before him, Blake found that this type of analysis does not work in explaining other uses of the subjunctive which are not as syntactically regulated. Such is the case in adverbial and adjectival clauses where the selection of mood relies on semantic and pragmatic factors. In addition, dubitative verbs (e.g., no creer, dudar, no dudar, etc.) and factive verbs (e.g., ser lastima 'to be a shame', gustar 'to be pleasing', alegrarse 'to be happy', etc.) can be followed by either the subjunctive or the indicative where the speaker relies on mood choice to produce subtle semantic and pragmatic differences in meaning (232~233). For these reasons Blake feels that it is imperative that any description or explanation of linguistic competence include discussions of semantic and pragmatic features upon whidh the child relies in developing his knowledge of the language. He concludes: ...the lexical and pragmatic restrictions, contrary to being especially secondary to the acquisition process of the mother tongue, appear to form.part of the child's knowledge of Spanish, what we have called linguistic competence (240). [MLQ's translation] 124 Al-Kasey (1993) also attempts to link syntactic, semantic and pragmatic properties to account for mood choice by adult native speakers and to account for the acquisition of mood by children L1 and adult L2 learners. She points out that there are two different semantic categories that take subjunctive clauses: factives and non-factives, and that their syntactic qualities are better predicted by semantic type than by mood. She states: ...there is not a clear division of mood and truth values. It is clear that the discourse use of the predicate and embedded proposition, the function of the speech act, must be taken into account for predicting mood selection (11—12). In her study, Al-Kasey asks subjects to judge indicative and subjunctive complements as indicating presupposition or not. She had. hypothesized. that the L2 learners would correlate subjunctive mood and non-presupposition (or non- factivity). These learners only did so for predicates of belief (or doubt) such as dudo que 'to doubt that' and no creer que 'to not believe'. The child L1 learners' responses did not pattern in the same way. Although.Al-Kasey argues that her L1 and L2 data support the principles and parameter theory of language acquisition and that adult L2 learners do have access to universal grammar, her most important contribution is in pointing out and maintaining that both indicative and subjunctive complements differ according to factivity. Not all 125 subjunctive complements are alike; and learners respond to this difference in responding to data that reflects this difference. This distinction may help to explain why both children L1 and the adult L2 learners of Spanish in the present research study use and acquire non-factive uses of the subjunctive following syntactic frames expressing desire, futurity and necessity before the factive uses of syntactic frames expressing doubt. 5.3. Conclusions It is evident from the studies reviewed here that any viable account of the Spanish mood system must include an analysis of the mmltiple syntactic, semantic and pragmatic phenomena which intervene to constrain its meaning and use. In addition, as Blake says, any description or explanation of the developing linguistic competence of the learner must also take into account these multiple factors. This same conclusion. was reached. in reviewing and discussing the theoretical studies devoted to analyzing the subjunctive mood in Spanish. This brings us back to Ellis' (1990) stance that in second language acquisition research both linguistic form.and how' it interrelates ‘with.:meaning and function. must be addressed in order to completely understand and explain the acquisition process. Analyses must go beyond mere descriptions and look for explanations as to why the 126 subjunctive is a late—learned structure in child language acquisition and a difficult an/or unused structure for adult second language learners. Analyses of the acquisition of the Spanish mood system must taken into account the multiple factors related to its meaning and use which appear to constrain these phenomena for adult native speakers. CHAPTER SIX: DATA COLLECTION AND METHODS OF ANALYSIS 6.1. Introduction The present research project consists of eliciting and analyzing, for its evidence regarding the acquisition of the present subjunctive, natural oral and written data from a group of young, American university students studying Spanish at the Universidad.Aut6noma de Querétaro in Querétaro, Mexico during the months of April - June 1992. 6.2. Data Collection The data elicited for analysis consisted of informal writing samples taken from the journals of the subjects under study and two sets of taped interviews. The first interview was carried out soon after the students' arrival in Mexico, and the second after three months of living there, at the end of a trimester course in intensive Spanish language and Mexican literature, history and culture. The interviews were designed to elicit as many subjunctive verbs as possible. Questions were posed in the subjunctive and it was hoped subjects would respond in kind. The subjects were told that the interviews were informal conversations to be used by their teachers to assess their level of competence in speaking and understanding Spanish, the first interview to be used in determining their incoming level and the final interview to measure their level of achievement. 127 128 It was stressed that the interviews would not be used in any formal evaluation, such as assigning a grade or determining the students' placement in the program. The journals were a weekly writing assignment required by the students' Spanish teachers. The students were told they could write about any subject they wished but were encouraged to write about their experiences living, studying and traveling in Mexico. It was hoped that in producing the oral and written data the subjects would be attending to meaning and not form. The students were encouraged to express their ideas without feeling pressured to be grammatically accurate. Many researchers agree (e.g. Labov 1970, 1972; Tarone 1982, 1983; Ellis 1986, 1990) that the most systematic language use is evidenced when the speaker is not concerned with ' grammatical accuracy. In formal styles, where speakers attend more to form, 'the normal pattern of language is disturbed' (Ellis 1986: 88). 6.3. Subjects The subjects of the study were sixteen American students, all native speakers of English, from Dartmouth College, Hanover, New Hampshire. The students were interviewed as to their previous exposure to Spanish, All had recently completed an intensive first—year Spanish. program: which concentrated Dartmouth's standard three-term in two terms. In .Mexico, the students continued an intensive program which was 129 to fulfill and complete their own college's requirement of two years of foreign language study. Some of the subjects had also had some Spanish previously in high school but these subjects were not eliminated from the study due to the fact that, despite this, they lacked the skills to qualify them to test out of their college's language requirement. The subjects were all freshman or sophomore students, ranging from eighteen to twenty years of age. The first-year Spanish Language Program at Dartmouth College utilizes the Natural Approach Method developed by Krashen and Terrell (1983), in which conscious, grammatical instruction is secondary to developing communicative speaking and listening skills, although the program does follow a structural syllabus. The text used for this course is Dos Mundos: A Communicative Approach (Terrell, Andrade, Egasse & Munoz 1986) . The second-year Spanish program utilizes varying texts. The text used for the Spanish in Queretaro Program, at the time of the present research, was Olivella de Castell's Mundo Hispano: Lengua & Cultura (1981) . 6.4. Setting All students enrolled in the program lived with Mexican families, one student to a family, in Queretaro, Mexico, for the three-month duration of the program. During this time, the students were expected to spend a great deal of time with the host families. These families were considered to be middle or upper-middle class socially and economically, with 130 at least one member, usually the father, being a professional: doctor, lawyer, professor, teacher, businessman. etc. The families resided in middle class or upper-middle class neighborhoods with all of the modern comforts that most American university students are accustomed to. The families chosen, in general, had children approximately the age of the American students. In past programs, this arrangement worked very well, and the students were encouraged to be involved in the family life of their host families as much as possible. From the content of the final interviews, it appears that most students did so. Many of the American students from past programs had continued to maintain close ties with their Mexican families, which indicates the level of involvement both culturally and linguistically that these students had with their Mexican families. 6.5. Language Program in Queretaro The language program in which the subjects were enrolled is an intensive fifteen-hours per week, twelve-week program. The students have three hours daily of Spanish language instruction: one hour of grammar, one hour of conversation, and one hour of a tutorial where the student meets one-on-one with a Mexican graduate student in Modern Languages (for further work on any specific area the student or the tutor deem necessary). In addition, the students have two three- hours-per-week courses in the History and Culture of Mexico and Modern Mexican Literature and Writing, both given 131 primarily in Spanish, although English, it was reported by many students, was used occasionally. All classes were given in the Modern Language Department of the Universidad Autonoma de Querétaro. The language professors are members of the faculty of the Modern Language Department. The history professor is on the faculty of another national university located in Querétaro and was especially invited to give the history class. The literature professor was a visiting professor from Dartmouth College who normally teaches Peninsular and Latin American literature courses at that college. In addition to the language and subject-matter courses, the students were required to participate in two planned three- to four-day cultural excursions in which they visited important historical, archaeological or cultural sites in Mexico. These excursions were conducted by the professor of the Mexican Culture and History course. The students also visited numerous touristic sites during their free afternoons and weekends. In sum, the program's requirements and the students' own personal interests in traveling and getting to know the country provided them with, in fact, an extremely intensive immersion program in which they actively participated and interacted with the country, culture, people and language. 132 6.6. Interviews As previously mentioned, the interviews were designed to elicit the subjunctive (mostly the present subjunctive but also the past subjunctive) but had the appearance of an informal, casual conversation. To add to this appearance of informality, the interviews were conducted by Mexican Modern Language Department students majoring in Spanish. The actual interview format was designed by this researcher (Quesada) and the students were trained by the researcher to carry out the interviews. The use of native Spanish-speaking interviewers, who speak little English, hopefully prevented the subjects from.resorting to English when communication began to break down (if indeed it did). The first set of interviews was carried out on April 10— 12, 1992, five days after the students' arrival in Mexico. The second set of interviews was carried out on June 7-9 of the same year, near the end of the course. Each interview lasted approximately fifteen minutes. The interviews were recorded and later transcribed into conventional Spanish orthography by this researcher. The tapes were of very high quality and the sound was quite good. Occasionally the interviewee began speaking as he/she entered the room.and the interviewer had already begun the recording. These were the only instances when the sound was not good. In these cases, when speech was incomprehensible, this was noted in the transcripts and this data was not used for analysis. Out of the thirty-two interviews there were only two occasions when 133 the morphology of the verb form.could not be detected, ie. it could not be understood if the indicative or the subjunctive form. had. been used, and these two occasions were not considered for analysis. The style of the interviews is quite informal (see Appendix A). The questions at the beginning are easy and very predictable; these were designed to put the subjects at ease. In section B of both interviews, questions are posed in the subjunctive. If the subjects did not understand or did not respond appropriately to the question, the interviewer was to ask the simpler questions suggested in parentheses. If the subjects were capable of understanding and responding to the difficult questions, the interviewer continued with all the questions in section B. In section C of the interviews, questions were again simple, so that the subjects would leave the interview feeling somewhat confident of their language ability. This was especially important at the end of the initial interviews, so that the subjects would not remember the first interview and feel tense and nervous during the second. 6.7. Journals The students were required by their writing teacher to keep a weekly journal in which they wrote entries three times a week in Spanish. The tOpics on which they could write were of the students' on choice, but as previously mentioned, students were encouraged to write about their experiences in 134 Mexico. At the end of the course, with the students' permission, the journals were made available to me for interlanguage analysis. Again, it was presumed that the data from.these journals was quite natural, where the writers were attending to meaning and not form. 6.8.. Methods of Analysis In order to address the hypotheses set forth in chapter one, the following methods of analysis were carried out. Once the tapes of the thirty-two interviews had been transcribed, an error analysis was done in which these data together with the sixteen journals were analyzed for the use and non-use of the subjunctive. In this type of analysis the number of occasions are counted in which the structure under study would normally occur in native speaker speech or writing, and this is compared with the number of times the structure is actually used. In order to determine which occasions would be used for the analysis of this study, the researcher considered any occasion which the subjects' textbook pointed out the subjunctive should or could be used. These occasions were mostly of four types: (1) verbs in complement clauses of matrix (main) verbs of desire or hope, querer que, esperar que, etc.; of doubt, no creer, no pensar, etc.; of comment, alegrarse que, etc.; and. of impersonal expressions, es ‘posible/necesario que, etc.; (2) verbs in adverbial clauses such as cuando, antes/después de que, hasta que, (3) verbs in 135 noun clauses such as lo que, and (4) main verbs following ojala. In the discussion of the results these occasions will be referred to as 'syntactic frames' or 'frames.' In order to ascertain if the occasions selected for analysis did indeed require or allow the subjunctive, the occasions selected from the transcripts by the researcher were presented to two groups of native Spanish speakers. In the first group there were nine Spanish majors from the Universidad Autonoma de Querétaro's Department of Modern Languages. The second group was made up of four native speakers of Spanish who were graduate students but not language majors and who claimed to know very little about language, grammar or linguistics. The occasions were presented to the native speakers in the written format in which they are presented in Appendix ' Two. When the utterance or sentence could not be understood in isolation, the complete discourse passage in which it occurred was provided. These speakers were asked to provide judgements of whether the occasions selected required or allowed the present subjunctive in Spanish. Their responses were in written form although many judgements were commented on orally with the researcher. Of the 239 occasions presented to them, they judged 223 as requiring or allowing the present subjunctive in Spanish. For some of the sentences, the native speakers said that they would use a sentence structure that did not require the present subjunctive. In fact they confirmed that, not 136 surprisingly, there were numerous other ways of saying the same thing. However, all reached a consensus that for the final occasions selected for analysis, the present subjunctive was required or allowed by the way the sentence was set up. Next, the utterances (either clauses, sentences or discourse units) in which the occasions appeared were listed according to syntactic frame. Then, for each syntactic frame, the verb form.which was used, either the present subjunctive or another verb form in place of the present subjunctive, was tabulated (see Appendix B for the list of occasions according to syntactic frame and the verb form used for both sets of interviews). A great deal of variability of verb forms was found in both interview situations, and. of syntactic frames in interview two. In the first interview there was very little subjunctive marking and this was spread out over only three of the sixteen subjects. But the variability of verb forms used was spread throughout the sixteen subjects. In the second interview, both subjunctive marking and other verb—form marking was spread throughout the sixteen subjects. Following this analysis and in order to eliminate the possibility that this variability occurs according to chance, a chi-square test was performed on the data, specifically on the syntactic frames and the forms produced for each frame. Finally, in order to attempt to understand why these subjects selected the subjunctive mood only for certain syntactic frames, a cross-products chart (Adamson 1990) was set up. 137 Analysis of the data suggests that second language learners construct prototype schema for rule application of complex syntactic and semantic structures. The results of these analyses are presented and discussed in the following chapter. CHAPTER SEVEN: RESULTS AND DISCUSSION 7.1. Introduction The results of the analysis support some part of each hypothesis set forth in chapter one, presented here again. H1: Young English-speaking adults do use the subjunctive in the upper-beginning/lower- intermediate stages of language learning. This use, although limited to few linguistic environments, is systematically variable. Hypothesis One is confirmed by the results; the young English—speaking adults in this study do use the present subjunctive in the stage of language acquisition that they are in, i.e. the upper—beginning/lower-intermediate stage. This use is limited to specific linguistic environments but is systematically variable . H2: The subjunctive will only be used in linguistic environments where (1) the syntactic frame is overtly produced, (2) the morphology is highly salient and (3) the meaning/function of the utterance is to signal either a possible future action or event or a desire on the part of the subject of the main clause. 138 139 Hypothesis Two is partially confirmed in that the results of the analysis show that these subjects tend to use the present subjunctive more often when the syntactic frame conveys both a possible future action or event or a desire on the part of the subject of the main clause, and when the subordinate clause verb is an irregular verb for which the verb morphology is presumably more salient than for regular verbs. However, it could not be confirmed that the overt appearance or production of the syntactic frame influenced, either favorably or not, subjunctive use. H3: The subjunctive will not be used in linguistic environments which are lacking either one or more of these features (listed in H2). Hypothesis Three was confirmed in the same fashion. When the syntactic frame does not convey a possible future action or event or does not convey desire on the part of the main clause subject and the subordinate clause verb is regular, the probability that the verb will be marked for subjunctive greatly decreases. Again, the results of the data could not confirm whether the lack of an overtly produced syntactic frame influences subjunctive marking for these subjects. H4: There will be variability in the use Of alternative forms used when the subjunctive is not used. 140 The results confirmed Hypothesis Four; there is a great amount of variability in the use of alternative forms when the subjunctive is not marked. This variability is not seen in the indicative subordinate clauses analyzed from the same data. In addition, the results of the chi—square test carried out confirm. that the variability seen is statistically significant and systematic and not random. H5: Young English-speaking adults in the upper- beginning/lower-intermediate stages of language learning do produce complex sentences and there is no correlation between this production and subjunctive use. The results of the analyses of both the present subjunctive and the present indicative subordinate clauses confirmlHypothesis Five. For just the present indicative and subjunctive there was a total of 522 subordinate clauses. The number' of subordinate clauses for' other tenses was not tabulated; however, if they had been, this number would have increased substantially. There was not a lack of complex sentences. In addition, the production of complex sentences neither contributed to nor inhibited the use of the subjunctive. Finally, the variability in the alternative forms and the sentence patterns in which they appear suggest a developmental sequence of the Spanish subjunctive in which learners move 141 from a dependency on L1 syntax and morphology to greater reliance on the L2 syntax, morphology and semantics. The results of the analysis of the data will be discussed in more detail in the following sections. 7.2. Error Analysis and Grouping According to Syntactic -Frame Analysis revealed that there was a total of 223 occasions in which the present subjunctive should or could be used across the two interview sets. For Interview One there were 108 such occasions and for Interview Two there were 115. Unfortunately, for the purposes of this study, the data from the journals could not be used. Journal entries by most of the subjects were extremely short -- less than half a page -- and were not systematically recorded. In the journals there were only sixteen occasions identified in consultation with the native speakers -- not a large enough sample to reveal any pattern of use. The journals, it is to be recalled, were a weekly writing assignment; topics were not assigned and neither the present subjunctive nor any other structure was tapped, unlike in the oral interviews. Undoubtedly, where all possible, the subjects chose to avoid using a structure they were still in the process of acquiring. Tables 1 and 2 present the results of the error analysis and subsequent grouping for the data from.Interviews One and Two respectively (see Appendix B for the 225 occasions of clauses requiring or allowing present subjunctive verbs). 142 Table l. Subjunctive (S) and Other (0) Marking According to Syntactic Frame - Interview One. Syntactic Number of 0-Marking S-Marking % % Frame Occasions . o-Marking S-Marking l 9 9 0 100% 0% 2 15 13 2 87% 13% 3 30 24 6 80% 20% 4 2 2 0 100% 0% 5 5 5 0 100% 0% 6 33 33 0 100% 0% 7 4 4 0 100% 0% 8 2 2 0 100% 0% 9 8 8 0 100% 0% Total 108 100 8 92.5% 7.5% Syntactic Frames Type 1: cs posiblclnecesario/logico/etc.) que_ Type 2: cuando_ Type 3: querer que___ Type 4: no pensar que— Type 5: recomendar que— Type 6: esperar que— Type 7: antes/despues del para que— ‘I‘ype 8: gustarse que— Type 9: lo que/a dande— 143 Table 2. Subjunctive (S) and Other (0) Marking According to Syntactic Frame - Interview Two. Syntactic Number of 0-Marking S-Marking ‘5 % Frame Occasions (ll-Mills) (S-Mnrklng) 1 13 10 3 77% 23% 2 24 12 12 50% 50% 3 20 6 14 30% 70% 4 9 9 0 100% 0% 5 21 13 8 62% 38% 6 8 8 0 100% 0% 7 7 4 3 57% 43% 8 7 0 7 0% 100% 9 6 3 3 50% 50% Total 1 15 65 50 57% 43% Syntactic Frames Type 1: es posible/necesario que_ Type 2: cuando— Type 3: querer que_ Type 4: no creer que_ Type 5: esperar que_ Type 6: hasta/(sin/para/etc. que_ Type 7: alegrarse que_ Type 8: ojala (que)— Type 9: Others 144 These tables show the syntactic frame types, the number of occasions and the amount of subjunctive marking or lack of marking, both raw numbers and percentages. Table 1 shows that in Interview One present subjunctive marking is very low, only 7%, and only occurs in the complement verbs following the syntactic frames cuando_ 'when_', and querer que_ 'to want that_', although other syntactic frames requiring or allowing present subjunctive in their complements occur in these subjects' production. These include es posible (necesario, importante) que_ 'it is possible (necessary, important) that', esperar que_ 'to expect (hope) that_', antes/después de que, para que_ 'before/after, in order to__', gustarse que_ 'to be pleased that_' and lo que; a donde_ 'that which; to wherever_' . From Table 2, it can be seen that subjunctive marking increases to 43% in Interview Two and now is produced in complement verbs following a greater range of syntactic frames, including es posible (necesario, importante) que_, cuando_, querer que_, esperar que_, alegrarse que_ 'to be happy that' , ojala_ 'oh, how I wish that_' and six assorted syntactic frames of which there is only one occasion each. Complement verbs following the syntactic frames no creer que_ 'to not believe that' and hasta (sin, para) que_ 'until, without, in order to___' are never marked for the subjunctive. The tables reveal several interesting facts. First, present subjunctive use increases from 7% to 43% overall from interview situation one to situation two. Syntactic frames 145 followed by complement verbs with present subjunctive verbs increase from two in Interview One to seven in Interview Two. In the frames in which there are present subjunctive verbs in the complements in. Interview One, the number of present subjunctive verbs increases in Interview Two, from.13% to 50% in the complements following the frame cuando___, and from.20% to 70% for the frame querer que___. That use of the present subjunctive would increase after twelve weeks of intensive Spanish study in the target language country is not surprising. It does however, confirm both Schachter's and Huebner's findings that forms are initially used in one or a few environments and for one function or meaning, and then gradually spread to more environments and for more functions or meanings. In addition, this finding leads to more compelling questions. First, when the subjects do not mark for present subjunctive, what verb forms do they use? Is there one verb form they choose above the others? Why? Is there variability in the verb forms they choose? Are the choices of these verb forms systematic and/or semantically or pragmatically related? Secondly, what is the nature of the environments for which the subjunctive is being used -- or better, when the present subjunctive is marked, what might explain this? 146 7.3. Subjunctive-Marking Vs. Other Verb FormrMarking The 223 occasions for obligatory or possible use of the subjunctive were analyzed for what verb form was used by subjects on these occasions (again, see Appendix B). The two interview situations will be discussed separately and then compared. 7.3.1. Interview One In Interview One there was a total of eight different types of verb forms used in the nine syntactic frames of the 108 occasions. Not surprisingly, the preferred form was the present indicative. Other forms included: the infinitive, the future indicative, the past preterit indicative, the present participle (gerund), the present indicative with a wrong person or number used, and invented forms (new forms created by the subjects which don't exist in the language). Table 3 shows the distribution of the total use and percentage of the use of alternative forms for the nine syntactic frames. Instead of a positive correlation between correct usage of the subjunctive and of a particular alternative verb form, precisely the opposite was found. In the complement verbs following the syntactic frames for which the subjects tend to use the subjunctive more, a higher percentage of other forms is also seen. In cases where the subjunctive does not appear in complement verbs, or scarcely appears, there is also less variability; a smaller number of other forms are used. 147 Table 3. Forms Used for Present Subjunctive by Syntactic Frame - Interview One. Synucuc #of Pres. Pres. Inf. Fut. Pret. Pres. Pres. Invent. Frame Occas. Subj. Ind. Ind. Ind. Part. IndJWF Form es posible! necesario 9 - 7 - 2 - - - - ‘1“6 78% 22% cuando 15 2 4 3 4 1 - 1 - 13% 27% 20"" 27% 7% 7% querer 30 6 6 10 - 1 1 4 2 que 20% 20% 33% 3% 3% 13% 7% no pensar 2 - 2 - - - - - - que 100% momendar 5 - 2 2 - 1 - - - 9m 40% 40% 20% esperar 33 - 1 1 18 - 1 - 2 1 que 33% 55% 3% 6% 3% antes] dupe): de 4 _ 3 1 _ _ _ _ _ ‘1“ an 25% que 75% tarse 2 - - 2 - - - - - gas 100 % lo que/a 8 - 8 - - - - - - donde 100% Total 108 8 43 36 6 4 l 7 3 Percentage 100% 7% 40% 33 6% 4% 1% 7% 3% 148 The syntactic frames in which there is a higher percentage in the use of the present subjunctive in their complements are cuando___ and querer que;__. For cuando___, there is 13% use of the subjunctive and 87% non-use of the subjunctive. The other verb forms which are used following the frame cuando___ are the present indicative (27%), the infinitive (20%) the future indicative (27%), the preterit indicative (7%) and the present indicative/wrong form.(WF). The use of the present subjunctive following this frame is very rare (only two occasions), but the fact that these learners use a variety of forms instead may be important. It may be significant, too, that for complement verbs following cuando___, the preferred other form is not only the present indicative, but also the future indicative, both chosen 27% of the time. The syntactic frame querer' que___ shows a similar pattern. There is some subjunctive marking in its complement, 20%, and a variety of other forms: present indicative 20%, infinitive 33%, preterit indicative 3%, present participle 3%, present indicative/WF 13%, and for invented forms 7%. In fact subjunctive marking is even greater for verbs following querer que_ than for cuando_, and the number of other forms employed is also greater, seven following querer que___ and six after cuando___. The preferred other form.used following the frame querer que___ is the infinitive (33%) and not the present indicative, which is chosen only as often as the subjunctive (20%). 149 For the frames which are all adverbial clauses, antes/después de que/para que_, there was no use of the subjunctive; 75% use of the present indicative and 25% use of the infinitive. The frame, esperar que___, oddly enough, is like cuandb___ and querer que;__ in the variety of other verb forms used in its complement clause, but lacks uses of the present subjunctive. Here, as for querer que___, the infinitive is the preferred choice (55%), followed by present indicative (33%), the present indicative/WP (6%), the preterit indicative and invented forms (each 3%). The syntactic frame recomendar que_, also displays some variability in the choice of other forms; the present indicative and the infinitive forms were again the favored alternative, seen 40% of the time, with the preterit indicative being used 20% of the time. The remaining syntactic frames studied all showed either little or no variability in the choice of other verb forms in the complement clause. Following the frames es posible/necesario/légico que___, the present indicative was used 78% of the time and the future 22% of the time. The frames no pensar que___ and lo que/a donde___ both showed 0% marking of the subjunctive and subjects employed the present indicative 100% of the time in their complement verbs. For the frame gustarse que___ the infinitive was used 100% of the time in substitution for the present subjunctive. In total, the present subjunctive was used in only 7.5% of all possible or obligatory occasions and the present 150 indicative in 40% of these. The infinitive appeared 33% of the time, followed by the present indicative/WP for 7%, the future indicative for 6%, the preterit indicative 4%, invented forms 3% and the present participle form.just 1% of the time. The variability seen in these subjects' choice of verb forms in the complement clause following certain syntactic frames which require or allow the present subjunctive mood appears to be considerable. But is it systematic and significant, or is the variability due to chance? The analyses carried out to attempt to answer this question are discussed in the following section. 7.3.2. Indicative Clause Analysis and Chi-Square Test First, occasions in the data for present indicative marking in complement subordinate clauses from the data of the interviews were counted. The number of present indicative' verbs and the number of other verb forms used according to syntactic frame were tabulated in the same way as has been done for occasions of present subjunctive .marking (see Appendix C for a list of subordinate clause syntactic frames requiring or allowing indicative verbs). Table 4 shows the results of this tabulation. First of all, out of 125 obligatory occasions, subjects employed the indicative mood 115 times or 92% of the time. They used the present indicative/WP 3% of the time, the infinitive 2%, and the preterit indicative, the imperfect indicative and the subjunctive only once each for 1% of the time for each form, 151 Table 4. Forms Used for Present Indicative by Syntactic Frame - Interview One Syntactic # of Pres. Infin. Pret. Pres. Imperf. Subj. Frame Occas. Ind. Ind. IndJWF cuando 14 13 1 - - - - creer/pensar/ 22 20 l 1 - - - oir/ decir que (n0) sabcrqud 15 15 - - - - - donde/coma] coal/xi (adjectival) 29 27 - - 2 - - que si 8 8 - - - - - donde 6 6 - - - - - porque 31 26 1 - 2 1 1 Total 125 1 15 3 1 ' 4 l 1 Percentage 100% 92% 2% 1% 3% 1% 1% 152 Two of the present indicative/WF choices were errors of concordance or number, as in the following examples: (1) .Mi clases son muy dificil porque las .profesores habla espafiol muy rapido'My classes are very difficult because the professors (3 per., pl.) speak (3 per.,sing.) Spanish very rapidly.‘ (2) No puedo entender espanol porque la gente hablan muy rapidamente. 'I can't understand Spanish because people (3 per., sing.) speak (3 per.,pl.) very quickly.‘ In fact, in example (1), there are three concordance errors (mi clases), (son dificil) and (las profesores). In example (2), the subject has understood the subject, la gente 'people', to be plural as it is in English. Therefore, if these occasions are analyzed as errors of concordance and not of present indicative mood choice, then the amount of correct present indicative marking increases to 93%. Nevertheless, either analysis concludes that there is considerably less variability in verb form choices for the present indicative than for the present subjunctive. According to one definition of acquisition, (93% accuracy in obligatory occasions, Schacter 1986), these subjects have acquired (or have almost acquired) the present indicative. The high accuracy rate of present indicative verbs in these occasions and the very small number of other verb forms may 153 suggest that the amount of variability seen in the choice of other verb forms in occasions for subjunctive verbs is probably not random or due to chance. To support this conclusion, a chi-square test was performed on the data, using the software program Minitab (formatted for Windows 3.1). A chi-square test calculates the expected frequency and measures it against the observed frequency of some numerical value. In this case, the number of times a verb form occurred according to syntactic frame was compared with the number of times it would be expected to occur if it were not due to chance. Because the number of occasions for certain syntactic frames was too low for statistical purposes, it was necessary to combine some of the frames. E's posible/necesario/etc. and recomendar que were combined because recomendar que_ is essentially the same as the syntactic frame, es recomendable que_ which also requires the subjunctive mood in the complement clause. They are all impersonal clauses which semantically involve the idea of futurity. Gustarse que_ was combined with querer que_ because they are similar in meaning. Lo que/a donde_ was combined with antes/después de que/para que_ because the way in which these subjects use lo que/a donde_ is similar syntactically to the other frames they are combined with, although, lo que/a donde_ on other occasions may not require the subjunctive. All, except for lo que, are adverbial clauses. To test the randomness or systematicity of the verb forms 154 used by these subjects according to syntactic frame, only present subjunctive 'verbs, jpresent indicative 'verbs, infinitives and other were entered into the program. The other verb forms combined future indicative, preterit indicative, present participle, present indicative/WF and the invented forms because the number of occasions of use individually were too low for statistical purposes. The results of the chi-square test were significant: 2?: 41.527, with a degree of freedom of twelve (df12), and a probability of error of less than .001 gp<.001). The raw and statistical figures are presented in Table 5. The results of this test support what the comparison of the subjunctive data with the indicative data suggest, which is that the choice of verb forms is not random or due to chance, but that there is a statistically significant amount of variability in the way these subjects choose verb forms according to syntactic frame. In addition, the analysis of the subordinate clauses in which the present indicative is used, support Hypothesis Five that these adult L2 learners do produce complex sentences. 7.3.3. Discussion Why did subjects choose certain verb forms above others? Why was there a predominance of present indicative and infinitive forms? A look at the clauses or sentences and the discourse units in which these alternative verb forms appear may reveal something about how these subjects are processing the L2 and what strategies they may be using to communicate in 155 Table 5. Tabulated Statistics of Chi-Square Test - Interview One Rows: Form; Columns: Syntactic Frame; Cell Contents: Count and Expected Frequency 1 2 3 4 5 All 1 0 2 6 0 0 8 1.06 1.13 2.42 2.49 0.91 8.00 2 9 4 6 1 1 1 1 41 5.42 5.80 12.38 12.76 4.64 41.00 3 2 3 12 18 1 36 4.75 5.09 10.87 11.21 4.08 36.00 4 3 6 8 4 0 21 2.77 2.97 6.34 6.54 2.38 21.00 All 14 15 32 33 12 106 14.00 15.00 32.00 33.00 12.00 106.00 Chi-Square = 41.527 With D.F. = 12 p<.001 Row 1: Subjunctive Column 1: es posible/necesario/recomendableletc. que Row 2: Indicative Column 2: cuando Row 3: Infinitive Column 3: querer/gustarse que Row 4: Other Column 4: esperar que Column 5: antes/despues de quelparallo que/a donde 156 it. In some cases the subjects have transferred the syntactic structure from.English almost word for word, as the following examples demonstrate: (3) *Afi parientes no demandan nada de mi; quieren .mi estar alegre. 'My parents don't demand anything of me; they want me to be happy.’ (Well-formed Spanish: Mis papas no demandan nada de mi; quieren que yo esté alegre.) (4) *Mi novia me quiere ir a Tequisquiapan, pero no sé qué quiero hacer. 'My girlfriend wants me to go to Tequisquiapan, but I don't know what I want to do.‘ (Well-formed Spanish: Mi novia quiere que vaya a Tequisquiapan, pero no sé qué quiero hacer.) (5) fMis padres quieren me estudiar. 'My parents want me to study.' (Well-formed Spanish: .Mis padres quieren que yo estudie.) In cases such as the preceding, interference from.the first language in syntactic surface structures is obvious. However, syntactic transfer doesn't always occur. In fact, there were examples in the data where if positive syntactic transfer from the first language had occurred, it might have 157 resulted in the correct usage of the Spanish subjunctive, as in the following examples: (6) *Es necesario que el va a otro escuela por aprender ser un abogado. (lit.)*'It's necessary that he goes to another school for to learn to be a lawyer.‘ (Well-formed Spanish: Es necesario que el vaya a otra escuela para estudiar leyes.) 'It's necessary that he go (subj.) to another school to study law.‘ (7) *Recomendamos que ella v16 un doctor. (lit.)*'We recommend(ed) that she saw a doctor.‘ (Well-formed Spanish: Recomendamos que ella viera (vea) a un doctor.) 'We recommend that she see (subj.) a doctor.’ As in (6) and (7), in some dialects of English the subjunctive is still used -- one of the very limited uses of the subjunctive in English —- although in cases such as (7) above, it is also disappearing in some dialects, giving way to the present indicative as in 'We recommend that she sees a doctor.‘ In other dialects of English, the infinitive is now standard use for sentences such as (6) as in 'It's necessary for him to go to another school.’ Apart from the cases of first language transfer, the presence of certain other verb forms may indicate that 158 learners understand, although in a very limited sense, some of the meanings or functions of the subjunctive mood in Spanish. For the syntactic frame, cuando___, subjects used the future indicative just as often as the present indicative and more often than the infinitive. The fbllowing are examples of this: (8) *Cuando regresaré a los Estados Uhidos, voy a comenzar estudiar una vez.mas. (lit.)'*When I will return to the United States, I am going to study one time more.‘ (Well-formed Spanish: Cbando regrese a los E.U., voy a comenzar a estudiar de nuevo.) 'When I return (subj.) to the 0.5., I am.going to begin to study again.‘ (9) *Cuando regresaré a México, quiero que nos visitan. *'When I will return to Mexico, I want that they visit us.‘ . (Well-formed Spanish: Cuando regrese a México, quiero que nos visiten.) 'When I return (subj.) to Mexico, I want them to visit us.’ (10) *Cuandb me graduaré de Dartmouth, voy a seguir estudiando. *'When I will graduate from Dartmouth, I am.going to keep on studying.‘ 159 (Well-formed Spanish: Cuando me gradue de Dartmouth, voy a seguir estudiando.) 'When I graduate (subj.) from Dartmouth, I am.going to keep on studying.’ These examples reflect neither the structure of English nor of Spanish. What they do reflect is that the learner in some way understands that the verb which follows the syntactic frame cuando___, expresses some kind of possible future action or event. They may not realize that this possible future action in Spanish is expressed in the present subjunctive, but neither do they transfer the English structure which expresses the meaning of possible future action or event in the present indicative. Perhaps the learners understand in some way that this meaning is expressed in Spanish with a form other than the present indicative. Another possibility is that learners confuse the first person present subjunctive, regrese with the first person future indicative regresaré, since the former is more familiar to them. The only other syntactic frame in which subjects substituted the future for the present subjunctive in the complement verb was es logico/necesario/posible que , which also can indicate a possible or necessary future action. That subjects do not transfer the syntax of the present indicative from.English with expressions of possible future action, may be due to the fact that there are times when the subjects, unsure of the correct morphological inflection, opt for 160 expressing the meanings of the language in as direct a manner as possible. It seems at this point, in a very limited way, semantics is interacting with syntax and morphology and providing a way for learners to communicate in the second language. The results of the error analysis lead one to conclude that these learners have not yet acquired the Spanish subjunctive mood. However, looking at the other verb forms they use in complements following different syntactic frames provides an indicator that they at least have captured some of the semantic features of the Spanish subjunctive mood. It also provides evidence of where first language transfer plays a systematic role in second language production. The data reveal that the second language acquisition process is not a simple, linear one, but rather a complex process where learners take morphological, syntactic and semantic and, perhaps at later stages, pragmatic factors, both from the first language and from the second, into account in the difficult task of communicating in the second language. A look at some of the discourse units from the data reveal that, although learners do not always produce the correct form, this does not necessarily imply that they do not have some understanding of the structure, meaning, and use of the form. In fact, for some structures which are complex both syntactically and semantically, morphological form may be acquired simultaneously with syntactic and semantic competence. It may even be that syntactic and semantic 161 competence is developed before the morphological form is under productive control. The following example of discourse from Interview One, shows an emerging dominance of the use of the subjunctive by the learner. An analysis at the clause or the sentence level does not show this development in process. Only an analysis at the discourse level can do this. (11) 1 NS: aTus papas? (:Qué opinan? 0 sea, gquieren que signs estudiando, o que, o que trabajes, o qué quieren que tu hagas? 4 NNS: Quieren... que... que yo tar-mine, termina, mis, mis estudios, en la universidad, y tambien quieren que me divertirse (acdmo se dice?)... que.me adivertirsa? 8 NS: Quieren que me divierta. NNS: Si. Quieren que.me divierta, uh,.mmm, aque se alegre? Cbsas como asi... uh, después de, de la universidad, de.mis estudios, quieren que yo busca an, un trabajo. 13 NS: Que busques un trabajo. NNS: Si. E50. 1 NS: 'Your parents? What do they think? That is, do they want you to keep on studying, or that 162 you work, or what do they want you to do?‘ 4 NNS: 'They want...that...that I finish (ind.), finish (subj.) my studies, in the university, and also they want that to enjoy myself (infin.)? (how do you say it?)...that I enjoy myself?‘ (infin. with reflexive particle inflected) 10 ' NS: 'They want me to enjoy myself.‘ NNS: 'Yes. They want me to enjoy myself (subj.), uh, mmm, that I be happy (pres.subj./WF)? Things like that..., uh, after, the university, my studies, they want that I look (ind./WF) for a, a job.‘ 16 NS: 'They want you to look for (subj.) a job.‘ NNS: 'Yes. That's it.‘ In line 4, the learner first replies in the present indicative, third person singular, but then self-corrects and provides the present subjunctive. Here the use of the present indicative reflects neither interference from English nor the structure of Spanish. The learner is searching for the correct form.for the occasion and this time finds it. In line (6), the learner employs first the infinitive, attempts to self-correct, and then in line (7) invents a new form. It seems that she has constructed a type of subjunctive utilizing the infinitive plus the reflexive -se as the root form, building the inflection off of the -se ending rather than from the third person singular present indicative form. The native 163 speaker, the interviewer, gives her the correct form and the learner repeats it. In lines (9) and (10) it is not clear whether the learner has used the subjunctive correctly and has not used the reflexive pronoun correctly (que.me alegre 'that I be happY') or has used the adjective alegre with the wrong form of the verb ser (que sea alegre). Finally, in line (12), the learner uses the present indicative again but in the third person singular. It may be significant that this learner doesn't use first person singular, which could be attributed to interference from the first language. It seems she understands that the occasion requires another form.and is trying to construct and produce it, but doesn't have success. When the interviewer gives her the correct form, she recognizes it and confirms that is what she is trying to say, responding: Si. 630. ('Yes. That's it.') An analysis at the level of discourse shows that the learner controls the syntactic structure and is conscious of some of the semantic features of the subjunctive. What she still doesn't control is the morphological form, Nevertheless, it is evident that the lack of control over the :morphology does not necessarily imply that the learner is not in the process of acquiring the present subjunctive in Spanish. In other examples of discourse, the learner first appears to translate word—for—word from English, then hesitatingly 'self—corrects' the utterance to approximate Spanish, often 164 times asking for confirmation from the native speaker, as seen in the following example: (12) NS: Tbs papas. gouieren que tu trabajes, o que estudies? 'Your parents. Do they want you to work, or to study?‘ NNA: Oh. .Mis padres quieren.me estudiar. Quieren que yo aestudie? 'Oh. My parents want me to study. They want me to study [lit. that I study (pres.subj.)]?' NS: Hmm,.mm. NNA: Pero yo tengo mi trabajo. 'But I have my job.‘ Other' examples showr the learners searching for the correct fomm: (13) NS: gQué me recomendarias ver? 'What would you recommend that I see?‘ NNS: Ah, yo recomiendo que tu, ah, vea, eves? avistes? 'Oh, I recommend that you, ah, see, see? saw? NS: veaa. 'See.' (pres.subj.) 165 NNS: ...Vbas, um, 1a capital, washington, D.C.; es muy interesante... 'See (pres.subj.), um, the capitol, Washington, D.C.; it's very interesting...‘ (14) NNS: Ellos esperan que yo estqy'aprendi-, aprend-, apron... 'They expect that I am learn-, learn—, lear—...' NS: Aprenda espafiol. Learn (pres.subj.) Spanish. NNS: Si, aprenda espaflol. 'Yes, learn (pres.subj.) Spanish.‘ (15) NS: ngrqué estas estudiando espafiol? 'Why are you studying Spanish?‘ NNS: No sé exactamente. .Porque me gustaria trabajar en Espafia cuando..1mmm...graduo... 'I don't know exactly. Because I would like to work in Spain when...mmm...I graduate (pres.ind.)...' NS: gCuando te gradfies? 'When you graduate (pres. subj.)?' NNS: Si. 'YGS.’ -' 166 NS: aSi? 'Yes?‘ NNS: Gradu...gradu... 'Gradu...gradu...' NS: Gradues. 'You graduate (pres.subj.)' NNS: Graduo, agradfie? 'I graduate (pres.ind.), I graduate? (pres.subj.)' NS: Uh, huh. The variable use of forms seen in these lines of discourse, the insecurity in their use, the hesitation and the asking for confirmation, and the recognition and acceptance of the correct forms when given, all reveal an emerging awareness on the part of these speakers of the use of the present subjunctive in Spanish. In addition to the fact that the non-marking of the present subjunctive is systematically variable, it can be seen that the use of other forms is due to either universal language learning strategies such as transfer from the L1, overgeneralization of the target language rules, or a misunderstanding of how the L2 morphological form. is constructed. It is also seen that meaning, even at the beginning stages of the acquisition of structures which are complex both syntactically and semantically, is attended to and interacts with the other linguistic features of syntax and morphology. 167 Clearly, the results of the analysis of the data from Interview One support both Hypotheses One and Four as set forth above. These subjects, as has been seen, use the subjunctive only in complement subordinate clauses following the syntactic frames, cuando___ and querer que___. However, although this use is very limited, both the comparison between indicative and subjunctive use in obligatory occasions, and the results of the chi-square test confirm that there is systematic variability in the use of alternative forms used when the subjunctive is zero-marked. In addition, it is seen that this systematic variation is due to the combination of universal acquisition strategies and an interaction of multiple linguistic factors. After twelve weeks of intensive study and living in the target language country, do these subjects continue to show such systematic variability? (In order to answer this question, the data from Interview Two and the analysis performed are now presented and discussed. 7.3.4. Interview Two A summary of the results of the analysis of the data from the second interview are presented in Table 6. The most immediate observation is that subjunctive marking increases dramatically -- from 7% in Interview One to 43% of the 115 possible or obligatory occasions produced in Interview Two. In addition, these subjects produce more variety of syntactic frames in which the present subjunctive is required or allowed 168 Table 6. Forms Used for Present Subjunctive by Syntactic Frame - Interview Two. Syntactic # of Pres. Pres. Inf. . Fut. Para. Pret. Pm. Frame Occas. Subj. Ind. Ind. Fut. Ind. mm as posible! necesario l3 3 4 1 2 - - 3 que 23% 31% 8% 15% 23% cuando 24 12 4 - 5 - - 3 50% 17% 21% 12% querer que 20 14 4 - - - l l 70% 20% 5% 5% no creer 9 - 9 - - - - - que 100% esperar que 21 8 5 2 3 3 - - 38% 24% 10% 14% 14% hasta/sin! 8 - 8 - - - - - para que 100% alegrarse 7 3 3 1 - - - - que 43% 43% 14% ojala 7 7 - - - - - - 100% other 6 3 3 - - - - - 50% 50% Total 115 50 40 4 10 3 1 7 Percentage 100% 43% 35% 3% 9% 3% 1% 6% 169 (see Appendix B, Part II), yet the number of different types of verb form decreases (from eight in Interview One to seven in Interview' Two). The jpresent indicative remains the preferred alternative form used and other verb forms include the infinitive, the future indicative, the paraphrastic future (ir + infinitive), the preterit indicative and the present indicative/WP. There are no longer invented forms and there are no occasions of the present participle form. Clearly, the fact that there is an increase in the number of syntactic frames is due to the subjects' increased knowledge of vocabulary. In addition to the syntactic frames es .posible/necesario/importantei que;__, cuando___, querer que_, and esperar que_ produced in Interview One, the following frames are also produced: no creer que___, hasta/sin/para que___, alegrarse que___, ojala (que)___, (adjectival) que_, no porque_, e1 hecho de que_, lo que___, no sé si___ and quien sabe si___. The last six frames are grouped together as one syntactic frame for purposes of analysis because there is only one occasion for each environment. However, this frame could not be included in the chi-square test because the environments do not all share either syntactic or semantic properties. A pattern that is observed in Interview One is not quite the same in Interview Two, i.e. it is not always the case that the greater the percentage of present subjunctive marking, the greater the amount of variability in the use of other forms. In fact, for some frames, it is seen that present subjunctive 170 marking in the complement clause greatly increases, but the amount of variability in the choice of other verb forms decreases. For complement verbs following the frame cuandb___, present subjunctive marking increases (from.13% in Interview One) to 50%, while the variability of other forms used decreases (from six in Interview One) to four. In complement verbs of the frame querer que___ present subjunctive marking increases from 20% to 70% in Interview Two and the number of other verb forms decreases from seven in Interview One to only four in Interview Two. The other frames, however, do present a similar pattern for use as seen in Interview One. For the syntactic frame es .posible/necesario que;__, whereas in Interview One there is no subjunctive marking and there is little variety of forms (only two), in Interview Two there is now some present subjunctive 'marking (three out of thirteen or 23%) and more variability in the choice of other forms -- five. The other forms include 31% use of the present indicative, 8% use of the infinitive, 15% use of the future indicative and 23% use of the present indicative/WP. The present indicative continues to be the preferred other form.used in complement verbs following this syntactic frame. The frame, esperar que___, patterns in a similar way. Whereas in Interview One there is no present subjunctive :marking in its complement verbs, in Interview Two the present subjunctive is marked 38% of the time. The number of forms remains the same at five. However, the preferred other form 171 used is no longer the infinitive as in Interview One, but is now the present indicative, which is used 24% of the time, followed by the future and paraphrastic indicative futures, both used 14% each, and the infinitive which is used only 10% of the time. The use and percentage of forms used with the remaining syntactic frames cannot be compared with those from.lnterview One because they are not comparable. However, a similar pattern of use is seen overall. Where there are no subjunctive verbs in the complement clause, as in following the syntactic frames, no creer que_ and hasta/sin/para que___, there is also no variability in the use of other verb forms. For the syntactic frames alegrarse que___ and 'others', there is some present subjunctive marking and some variability. However, although the percentages of subjunctive marking are quite high, 43% and 50% following alegrarse que___ and 'others' respectively, the number of occasions is so low, seven and six, respectively, that it is difficult to draw any conclusions about patterns of use. . Following the syntactic frame, ojala, present subjunctive marking is 100% and there is no variability. Ojala, however, is a different case from the other linguistic environments examined -- what follows ojala is a main clause, not a subordinate one. This is discussed more thoroughly later on. In sum, the amount of subjunctive marking dramatically increases in Interview Two, to 43%, while the preferred other form.chosen continues to be the present indicative, although 172 it drops slightly from 40% to 35%. The use of the infinitive drops dramatically, from 33% in Interview One to only 3% of the time in Interview Two. The use of the future doubles to 12% of the time if the percentages of the two future indicatives are combined. The use of the preterit indicative remains the same at 1%, which indicates this type of error may be due to incorrect stress placement (i.e. pronouncing termine, present subjunctive 'finish,' as terminé, preterit indicative 'finished'). The use of the present indicative/WP doubles to 6%. Although any one of these increases could be accidental and therefore insignificant, the pattern of increase is unquestionable. Finally, while the number of syntactic frames requiring or allowing present subjunctive verbs in the complement clause increases, perhaps due to an increase in lexical knowledge on the part of the subjects, the number of variable forms decreases, probably due to an increase in the control of the target language morphology. Although the variability in the choice of verb form.is less in Interview Two, is it still significant? Is it systematic? The discussion in the following section will attempt to present the answers to these questions. 7.3.5. Indicative Clause Analysis and Chi-Square Test For the data from Interview Two, again the occasions for present indicative marking in subordinate clauses were examined (see Appendix C, Part II). Table 7 presents the results. Table 7. Forms Used for Present Indicative by Syntactic Frame - Interview Two. 173 Syntactic Number of Present Present Present Frame Occasions Indicative Indicative/WF Subjunctive cuando 22 19 1 2 crest/pensar 31 31 - - que si 9 9 - - donde 6 6 - - aunque 2 2 - - quien 6 6 - - parecer 2 2 - - corno 3 3 - - lo que 6 6 - - porque 26 25 l - (adjectival) 56 56 - - que Total 174 170 2 2 Percentage 100% 98% 1% 1% 174 What is remarkable is the almost complete lack of variability and the constancy of indicative marking. For the 174 obligatory occasions, the present indicative was marked 98% of the time. These subjects have acquired the present indicative in subordinate clauses and there is virtually no variability. Again a chi-square test was performed on the data from interview two, and again the results were significant. As for A the data from.lnterview One, two of the syntactic frames were combined and two others were eliminated because of such small numbers of occasions. The verb forms tested were the present subjunctive, the present indicative, the infinitive and the other forms were combined. The results of the test appear in Table 8: x“ = 50.160, with a degree of freedom of fifteen (dfls), and a probability of error of less than .001 gp<.001). Again, as with the data from Interview One, the results of this test support the conclusion that the choice of verb foams is not due to chance, but rather that there is a statistically significant difference in the way these subjects choose the different forms based on syntactic frame and that this is done systematically. 7.3.6. Discussion An interesting pattern emerges from the analysis on the variable use of the subjunctive and other verb forms in these occasions. In the first set of interviews it was seen that the use of the subjunctive, however limited, was also 175 Table 8. Tabulated Statistics of Chi-Square Test - Interview Two Rows: Form; Columns: Syntactic Frame; Cell Contents: Count and Expected Frequency 1 2 3 4 5 6 All 1 2 12 14 0 8 3 39 4.68 9.36 7.80 3.51 8.19 5.46 39.00 2 4 4 4 9 5 11 37 4.44 8.88 7.40 3.33 7.77 5.18 37.00 3 1 o o 0 2 0 3 0.36 0.72 0.60 0.27 0.63 0.42 3.00 4 5 8 2 0 6 0 21 2.52 5.04 4.20 1.89 4.41 2.94 21.00 A 12 24 20 9 21 14 100 11 12.00 24.00 20.00 9.00 21.00 14.00 100.00 Chi-Square = 50.160 With D.F. = 15 p<.001 Row 1: Subjunctive Colunnn 1: es posible/necesarioletc. que Row 2: Indicative Column 2: cuando Row 3: Infinitive Colunnn 3: querer que Row 4: Other Colunnn 4: no creer que Column 5: esperar que Column 6: hasta/sinIparallo que/porque 176 accompanied by a considerable amount of variability in the choices of other verb forms. However, in Interview One, the syntactic frame esperar que_ also exhibited a considerable amount of variability in the choice of other verb forms in its complement verbs but there was not one instance of use of the present subjunctive. Then, in the data from the second interview, for the syntactic frames in which subjunctive marking was the greatest, 50% for cuando_ and 70% for querer que_, variability decreased. It appears that when learners are developing competence in a structure and in its use and form, this developing competence is accompanied by a great deal of variability. The learner is 'trying on', so to speak, a number of forms to see which one 'fits'. As the learner becomes more competent in the structure, this variability decreases. The learner refines and limits the choices available. In Interview One, subjects are developing competence in subjunctive marking in the complement verbs for the syntactic frames cuando_ and querer que_. In Interview Two, as competence in subjunctive marking for these two environments improves, variability decreases. In fact, some of the variability seen in verb forms chosen is due to self— correction, as exemplified in the following examples: (16) Me gusto mucho las playas de México y cuando tannins. .., gtermine? 1a escuela, voy a vivir en la p1 aya . 177 'I like the beaches of Mexico very much and when school finishes...(pres.ind.), finishes? (pres.subj.), I am going to live on the beach.’ (17) .Es que, cuando regrese, ah, no...regrese a los Estados Uhidos, teneré mucho prisa. 'It's that, when I return (pres.ind.), ah, no...return (pres.subj.) to the U.S., I'll be in a hurry.‘ (18) Espero que podré tener, ah...,tendré, no... espero que pueda, asi?...ah, tener todos los dias con ellos. 'I hope that I will be able to have, ah...will have, no...I hope that I can (pres.subj.), yes?...ah, have all the days with them.‘ In Interview One, the syntactic frame esperar que;__, exhibits considerable variability in the verb forms in its complement verbs, although there are no present subjunctive verbs. It appears that subjects are developing competence in subjunctive marking for this frame. Indeed, in Interview Two, they employ the present subjunctive 38% of the time in complement verbs of this frame. For other frames where there are no present subjunctives in no creer que___ and hasta/sin/para que;__ neither is there variability. Another frame where there is no variability is ojala . 178 As mentioned before, this is one of the few cases in which the subjunctive mood in Spanish is found in an independent clause. In addition, unlike most of the other syntactic frames except for querer que___ and hasta/sin/para que___, the subjunctive mood is always required here, no matter what the context or the speaker's attitude. Learners do not need to struggle with a syntactically or semantically complex linguistic environment and they have no choice but to choose the subjunctive. Although the number of occasions in Interview Two for this frame is small, only six occasions, the fact that these subjects mark the verbs 100% for the subjunctive may indicate that they now have the morphological form of the present subjunctive under control, a finding reinforced by the fact that there are no longer any invented, 'subjunctive-like' forms as in Interview One. In addition, ojala___, shares certain features in common with the syntactic frame querer que_. Both of these invariably require the subjunctive in their complements -- there is no choice according to speaker attitude or context, and both share the semantic features of futurity and desire. It is probably no coincidence that these two frames both cause a high percentage of subjunctive-marking in Interview Two. The fact that there is still some variability in the choice of verb forms in complements following querer que___ could be due to the fact that querer que___ occurs in complex sentences. In addition, querer may be followed by an infinitive (when there is no que and the subject of querer and the complement 179 is the same). It appears, therefore, that except for the syntactic frame ojala_, the acquisition of the Spanish subjunctive mood is accompanied by a considerable amount variability in the choice of verb forms in the complement clauses. In fact, it may be that this variability or this 'trying on' of forms is a necessary stage in the learning process in order for competence in this complex structure to develop. Another interesting phenomenon that is exhibited in the data is the change in sentence patterns from the first interview situation to the second. Whereas in the first interview there was a great deal of evidence of negative transfer from the L1 in the sentence patterns and a high percentage of infinitives in complement clauses, in the second interview, the sentence patterns more closely resemble Spanish. Also, subjects produced more sentences containing the syntactic frames. This may help explain why in the first set of interviews there was such a low incidence of present subjunctive marking. It is difficult for learners to focus on the morphology of the verb when they are still acquiring the sentence patterns of the L2. The results of the analyses carried out on the data of Interview Two also confirm Hypotheses One and Four discussed above. Subjects are still using the subjunctive only 43% of the time, not enough to be able to state that they have 'acquired' the subjunctive. However, the very limited use seen in Interview One has spread to more environments. 180 Although subjects continue to evidence a great deal of variability in their choice of verb forms in the complement clauses, the variety of verb forms apparently available to them.has diminished slightly. The preferred alternative verb form.continues to be the present indicative; however, the use of the infinitive (with no marking for mood, tense, person or number) has greatly diminished. _Use of the future indicative has increased to 12% of the time. This may be due to the fact that when. these future indicatives appear, they' are in complement clauses following syntactic frames whose semantic meaning conveys possible futurity. Finally, the invented forms, although limited in the data from.lnterview One (only 3% of the time), do not appear at all in the indicative data and have disappeared as alternatives in the subjunctive data in Interview Two. From this, in addition to increased subjunctive marking, it can be seen that the subjects are stabilizing the verb morphology for the present subjunctive. The differences that are observed from.the data and the analyses carried out on the data from the two interview situations clearly show a shift in language learning strategies from transfer of L1 syntax and morphology in Interview One to a construction process that uses multiple linguistic features or levels to construct and use the L2 in Interview Two. 181 7.4. Prototype Features of Syntactic Frames and Complement Verbs Why is it that these subjects mark the present subjunctive in complement clauses following only certain syntactic frames -- two in Interview One and spreading to more than seven in Interview Two? What are the features of these syntactic frames that cause the subjects to :mark their complement verbs for subjunctive some of the time? Do the complement verbs themselves have certain features in common which promote or inhibit use of the present subjunctive? And what is the nature of the sentences or the discourse units in which the subjunctive tends to be marked? Is there also some significant feature of these sentences or units that promote or inhibit subjunctive marking? In order to identify what these features might be and to attempt to ascertain their possible influence on the marking or non—marking of the present subjunctiwe, a cross-products chart (Adamson 1989, 1990) was set up so that it would be possible to see in what way certain features may combine and interact to promote or inhibit subjunctive marking. 7.4.1. Interview One Following the analysis of data from.lnterview One, it was noticed that the linguistic environments where the subjunctive was marked had several features in common. First, the syntactic frame was always overtly marked by the learners, i.e. the learner produced a complete sentence in his/her 182 utterance, as in the following examples: (19) Cuando.me vuelva a mi casa, voy a almorzar. 'When I return home, I am going to eat lunch.‘ (20) Quieren que.me divierta. 'They want me to have fun.‘ In addition, as in the examples above, it was observed that when the verb in the complement clause was an irregular one (volver ~ vuelva; divertirse ~ me divierta), there was a greater tendency for that verb to be marked for present subjunctive. Finally, it was observed that both the cuando_ and querer que___ frames convey a meaning of futurity and querer que___, the syntactic frame most often marked for subjunctive, also conveys the meaning of desire on the part of the subject of the main clause. It may be that these learners have constructed a conceptual category or prototype schema for the marking of the present subjunctive that includes the features of +overtly marked syntactic frame [+OF], +futurity [+fut], +desire [+des], and +irregular verb [+irreg]. Adamson (1988, 1989, 1990) proposed that learners construct prototype schema for certain complex syntactic structures. These schema are conceptual categories which are not clear-cut but rather are networks of essential properties or features. JMembers of the category may not have any properties in common but instead ‘will share properties in common with a prototype which is most 183 typical of the category. The more features a structure has in common with the prototype, the more likely that the structure will promote marking of the target language form. Specifically for this case, it may be that these learners have a variable rule that states, 'when the sentence contains the overtly marked syntactic frame which conveys either futurity or desire (or both) and when the complement verb is irregular, then mark for present subjunctive'. Of course, as mentioned earlier, a variable rule states the linguistic environment for which the rule is more likely to apply, and the more features a sentence has in common with the prototype, the more likely it will be marked with the subjunctive. Therefore, the 108 occasions for subjunctive marking from Interview One and 109 occasions from Interview Two were analyzed and these [1] prototype features were calculated for each occasion (see Appendix D) and a cross-products chart was set up. Although there were 115 occasions in Interview Two, in order to regularize the data, six of these were not counted in the application of the prototype features. It was found that the learners tended to monitor or self-correct a good deal more in the second interview and so the first production of a form was not used if the learner immediately self- corrected and used another form, Figure l is the cross-products chart showing the way the four different features combine to form.aixteen environments for possible subjunctive marking. Since there were only eight occasions in which the complement verb was marked for present 184 [ti-'eaturesl Subimctlve- Marking Mme] 3114 (217.) [+DOS] I m: (1257.) (+Futl l-irregl [drug] 25 (407.) l-Desl hon l-irreel Mme] (+Desl I l-irrogl l-Futl [Ame] LIFT-I: l-irtegl [drag] [+Desl I [+Fut] [4.1708] [drug] l-trregl l-OFI Mme] [+093] I l-imgl l-I-‘utl Mme] H1 l-Des) l-irtegl Tom 81108 (71.) Other Verb tom-Marking 11/14 (797.) 21/24 (8757.) 3/5 (607.) 13 (1007.) 3 (1009.) 2 (1W1) 2 (1137.) 3 (1007.) 8 (“WM 25 (1007.) 3 (1m!) 1 (100%) 10002) 3 (1007.) 1 (1M!) twill” (9370) figure 1. Cross Products Chart: Four features yielding sixteen environments for subjuntive marking - Interview One 185 subjunctive, these results must be considered very cautiously and only in a preliminary fashion. Of the 108 occasions, 66 are [+OF]; and of these 66, 12% are marked for subjunctive and. 88% are not marked for subjunctive. However, of the 42 that are [-OF], none are marked for subjunctive and 42, or 100% are marked for another verb form. It may be that for these learners to mark a complement verb with the present subjunctive, a sentence must contain an overtly marked syntactic frame that would require its complement verb to be in the subjunctive. When an occasion is marked for subjunctive, it is always [+fut]; 75% of the time it is marked for [+des] and 25% marked for [-des]. When the subjunctive is marked, 62.5% of the time the verb is [+irreg] and 23.5% it is [~irreg]. Again, out of only eight total occasions, these figures may not be reliable. Next, the combination of features is observed to see how features may interact to promote or inhibit subjunctive marking. When the complement verbs are marked for subjunctive, 100% of the time the occasion has the features of [+OF], [+fut], and 75% of the time they are [+OF], [+fut], [+des]. Only 25% of the markings are [+OF], [+fut], [-des]. Although the data are very limited, it appears that in order for a complement verb to be marked with the present subjunctive for these learners, the occasion must have the features [+OF], [+fut]. The likelihood that a verb will be marked with the subjunctive increases if the features are also [+des], 186 [+irreg]. For the first set of interviews the subjunctive is never marked when the occasion is [-OF] or [—fut], and is never marked when the occasion is [—des] and [-irreg]. Again, the results of this analysis support Hypothesis One in that the subjunctive is used only in very limited linguistic environments. However, as can be seen by the cross-products chart, the marking of the present subjunctive in these environments is in. considerable 'variation. with marking for other verb forms. However, for these linguistic environments, other verb formemarking is diminished slightly. In the linguistic environments in which there is no subjunctive marking, other verb form-marking is 100% or categorical. In addition, the results appear to support, but only very preliminarily, Hypotheses Two and Three. The subjunctive is only marked for environments where (1) the syntactic frame is overtly marked, (2) the verb morphology is highly salient, as it is with irregular verbs, and (3) when the meaning of the syntactic frame is either possible futurity or possible futurity and desire on the part of the subject of the main clause. The data from Interview Two are examined in a like manner . 7.4.2. Interview Two After assigning the same iprototype features to the 109 occasions from.the data of Interview Two, it was seen that 92, or 84%, of the occasions were [+OF], and only 17, or 16%, were 187 [-OF]. Of the 92 [+OF] occasions, 46% were marked for the present subjunctive and 54% were marked for another verb form, Of the 17 [-OF], a similar distribution was seen: 47% were subjunctive-marked and 53% were other verb formrmarked. In these data then, [¢OF] appears to have no influence on the promotion or inhibition of present subjunctive marking and was, therefore, not included in the cross-products chart combining prototype features for the data from.lnterview Two. The features [ifut], [ides] and [iirreg] are included and the results of their combination are represented in Figure 2. Here it can be seen that when a syntactic frame has the features [+fut] and [+des], subjunctive marking is favored 64% of the time for irregular verbs and 63% of the time for regular. The feature [iirreg] neither promotes nor inhibits subjunctive marking. However, when the syntactic frame has the feature [-des], subjunctive marking decreases, although when the complement verb is [+irreg], it is the same as :marking for other verb forms. When the complement verb is [- irreg], subjunctive marking is. even more inhibited. Therefore, [iirreg] does not influence subjunctive marking ‘when the syntactic frame is [+fut], [+des], but when coupled ‘with [-des], it does influence subjunctive marking. It appears at first glance that syntactic frames that Ihave the features [-fut], [+des] and are followed by complement verbs that are [+irreg] promote subjunctive marking (they are subjunctive-marked 67% of the time) and that when these are followed by complement verbs that are [-irreg]. they 1188 [thahuesl hhxggl fines] l-irrogl («I-Full] l+imgl [435] l-irregl [+irrogl (4X53 l-lmgl FFufl 003:5] Idle) l-imgl Tom 50/109 (463) :amuuncwwr Inning 16/26(643) 12/19(638) 7/14(503) 6/24(258) 4/6(67X) l/3(333) 2/7(298) 0/6(08) ounmlnmb Ibnmdnmkhu: 10/26(368) 7/19(378) 7/14(503) 16/24(753) 2/6(338) 2/3(67x) 5/7(718) 6/6(1008) 59/109(548) Ifigurell CkusslhxmhnnstCharc 1Hununknuuresyifldingnightemndnmnnunustbr subdumunimennarking-flhmuuwiew71kvo 189 inhibit subjunctive marking (they occur only 33% of the time). However, it must be observed that the number of obligatory occasions that has these features is very small; there are six occasions with the features of [—fut], [+des], [+irreg] and only three for [~fut], [+des], [-irreg]. However, this pattern is also observed for syntactic frames that are [+fut], [-des] and those which are [-fut], [-des]. When syntactic frames have the features [-fut], [-des], subjunctive marking is greatly inhibited (it occurs only 29% of the time), and when combined with [-irreg], it is completely blocked (0% of the time). The feature [iirreg] does not promote or inhibit subjunctive marking when the syntactic frame has the features [+fut] and [+des]; however, it does when the syntactic frame has either [+fut] and [-des] or [-fut] and [+des], and especially when it has the features [-fut] and [+des]. The feature [—fut] does not appear to influence subjunctive marking unless coupled with [-irreg] and/or [~des]. These results suggest that subjects have constructed a prototype schema for present subjunctive marking that is [+fut, +des, +irreg]. The more features a sentence has in common with this prototype, the more likely that sentence will be marked for the present subjunctive. When the syntactic frame is [+fut], [+des] there is a greater likelihood that these subjects will mark the complement verb for subjunctive. If the syntactic frame is missing one of these features, subjunctive marking can be promoted when the complement verb 190 is irregular. That irregularity of the verb promotes subjunctive marking comes as no surprise. This confirms Wblframvs 'principle of perceptual saliency'. Wblfram (1989) found, it is to be recalled, that the closer phonetically the past tense form.is to the present tense form, the less likely it would be that his subject would mark for past tense. Indeed, the regular subjunctive marking is very close phonetically to the indicative, so close, in fact that NNS's may not often perceive it in NS's speech. This imperceptibility is strengthened by the fact that the syllable in which the morpheme appears is never stressed. For native English-speakers an unstressed /afl.is not very different from an unstressed /e/. For many irregular forms there is a greater difference between the present indicative forms and the present 'subjunctive forms; for the subjunctive there are root and suffix changes which help to make the form.more salient, more noticeable to the learner. Some irregular verbs, such as hacer 'to make, do' and decir 'to say, tell', are used often in everyday speech and are used often in command forms which are a form of the subjunctive. This may explain why the feature [+irreg] helps to promote subjunctive marking in some cases. If the syntactic frame is missing both of the features, [ifut] and.[ides], it greatly reduces the likelihood that the subjunctive will be marked. The meaning of the syntactic frame coupled with the morphological shape or saliency of the 191 complement verb are syntactic, semantic and morphological features which combine to influence how and when, and even with what likelihood, a certain structure will be produced by learners. In Chapter Four, I discussed Lunn's (1989b) proposition that for native Spanish speakers there is a prototype of the subjunctive mood that contains the feature [-assertability]. It was discussed that complements that were not asserted by the speaker are marked in the subjunctive, but that also complements which contain untrue information and true but less-than-optimally relevant information were unassertable. How does this native-speaker prototype compare with that of learners in this study? To the extent that any information that is [+fut], [+des] is also [-assertable], the prototype of the subjunctive that these learners have set up is the same. However, it was seen that for Interview One, the feature [+OF] influenced subjunctive marking in conjunction with [ifut] and [ides]. In the second interview, the feature [+irreg] for the complement verb influenced subjunctive marking. For second language learners overt syntactic and/or morphological features combined with semantic features are important in signaling subjunctive marking. However, although information that is [+fut] and [+des] is also [-assertable], it is not necessarily so the other way around. Information that is [-assertable] is not always [+fut], [+des]. According to Lunn's (and Lavandera's and.Mejias-Bikandi's) analysis, the prototype that 192 native speakers have constructed is a pragmatic one where speakers may choose to assert information or not, and whether this information is [+fut] and/or [+des] is not part of the prototype. In addition, native speakers do not rely on the overt syntactic feature of an overtly marked syntactic frame [+OF] or on the salient morphological marking of irregular verbs [+irreg] as second language learners do. For these learners the prototype schema is a syntactic, morphological, semantic one, not a pragmatic one. This is made even. more evident by looking at the environments that are marked for subjunctive and those that are not marked for the subjunctive. Subjunctive marking and variability in complement verb marking in both sets of interviews is highest for syntactic frames that are [+fut], [+des]. These frames include querer que___, esperar que___, and oja1a___. Subjunctive marking and variability in complement verb marking is second highest for syntactic frames that are [+fut], [-des]. These include cuando_ and es posible/necesario/etc. que___. Subjunctive marking is lowest or not marked at all in complement clauses following syntactic frames that are [-fut], [+des] or [—fut], [-des]. These include alegrarse que , estar alegre/contento que , no creer/pensar que___, sin/lo que___ and e1 hecho de que___. Syntactic frames that are [+fut], [+des] almost always obligatorily require the subjunctive in the complement verb. Those that are [+fut], [-des] require the subjunctive less often obligatorily; and those that are [-fut], [+des] or [- 193 fut], [-des] are the syntactic frames that leave the choice for subjunctive marking up to the speaker. It is here that pragmatics and the speaker's intention come into play. Information that the speaker wants to assert will be marked in the indicative, and information that the speaker does not want to assert will be marked in the subjunctive. Apparently, these learners have not yet reached the level of acquisition where their prototype for the present subjunctive matches that of native speakers. 7.5. Conclusions That learners have such difficulty producing the subjunctive in sentences that are [-fut], [+des] or [~fut], [— des] is certainly understandable. It will be recalled that it is precisely this area of the subjunctive mood that was ignored by both traditional descriptions of Spanish grammar and most textbook presentations, probably because it is not understood. Also, this is the area where linguists have had the most difficulty describing and explaining the nature of the Spanish mood system. Bolinger (1968) first pointed out that for these kinds of sentences, only propositions that could be fronted in English.would take the indicative mood in their corresponding Spanish sentences. Rivero (1971) pointed out that subjunctive and indicative complements had different underlying structures, both semantically and syntactically. Later, Terrell and Hooper (1974) wanted to correlate the indicative mood with assertion and the subjunctive mood with 194 non-assertion but had problems accounting for matrix verbs which were presupposed but not asserted but for which native speakers followed with indicative complement verbs. Takagaki (1984) and Reider (1989, 1990) attempted to apply the notions of independency and subordinance to the distinction between the indicative and subjunctive moods. Finally, Lavandera (1983, 1984), Lunn (1989a, 1989b) and Mejias-Bikandi (1994) point out that mood selection is the prerogative of the speaker in choosing to assert or down-play propositions. Finally, it was seen that researchers in child language acquisition found that children do not acquire the subjunctive in these kinds of syntactic frames until later. Gili y Gaya (1972) found that children do not acquire the more subtle uses of the subjunctive after dubitative (dudar, no creer, etc.) and factive verbs (alegrarse, estar alegre) and e1 hecho de que, etc.) where semantics and pragmatics dictate usage. Because adult native speakers vary precisely in these uses of the subjunctive, sensitively judging the relevance of the proposition and deciding whether to assert the proposition or not, it becomes more understandable that these are the areas of difficulty in language acquisition, both for children and adults acquiring Spanish as a second language. It is confirmed. that it is the semantic/pragmatic complexity of the Spanish mood system that causes such difficulties for adult second language learners. Indeed, it was confirmed that complex sentences containing subordinate clauses themselves were not difficult for the learners in this 195 study. Although Floyd (1990) and Collentine (1995) had found that their subjects, both children and adults, had difficulty producing complex sentences that contained subordinate clauses, the results of the present research study found that these learners produced abundant complex sentences with present indicative verbs. In fact, by the time of the second interview, these learners were committing very few syntactic errors in which they transferred the sentence patterns of English. In fact, like Blake's (1991) child subjects, these learners had greatly improved in accuracy for obligatory subjunctive marking following optative matrix verbs (querer que, esperar que,) and ojala. That both the child's and the adult second language learner's prototype of the Spanish subjunctive mood does not match. precisely the adult native speaker's is evident. However, it is seen that this does contain some of the features that are part of the native speaker's schema. The prototype schema of the Spanish subjunctive that these learners have constructed is evidently [+fut], [+des] which, when combined with a complement verb which is [+irreg], is more likely to be marked morphologically with the subjunctive. Adamaon.maintains that language learners refine the prototype schema as language develops. In order for adult language learners to adjust their prototype of the Spanish subjunctive, the kinds of language experiences they encounter, both in and outside the classroom, may be crucial in order for them to develop a closer match of their prototype to that of adult 196 native speakers. CHAPTER EIGHT: IMPLICATIONS FOR TEACHING AND FUTURE RESEARCH 8.1. Summary Results of the analyses discussed in Chapter Seven support Hypotheses One, Four and Five and part of Hypotheses Two and Three. The young English—speaking adults in this study used the present subjunctive in very limited linguistic environments in Interview One (two environments) and in more environments in Interview' Two (seven). However, these learners do not use the subjunctive at all or in very few instances in syntactic environments for which subjunctive marking involves a great deal of semantic and pragmatic complexity. It is in these areas that traditional grammars, textbook descriptions and synchronic/theoretical studies of the Spanish subjunctive all had difficulty in accounting for .its use. In addition, it was seen that children, as well as adults, have difficulty acquiring these more subtle and complex uses of the present subjunctive. For both interviews, it was seen that the variable use of the present subjunctive was systematically variable according to syntactic frame and that the variability was statistically significant. It was. seen that ‘when the obligatory' occasion for subjunctive marking contained the features [+fut],[+des] there was a greater likelihood that these subjects would mark for the present subjunctive. When the obligatory occasion was lacking in one or both of these features, there was less 197 198 likelihood that subjunctive marking would occur. However, if the complement verb was [+irreg], this feature increased the possibilities that the subjects would. mark for present subjunctive. It was seen that for the first interview, if the syntactic frame was [+OF], there was a greater likelihood that the subjects would mark for present subjunctive; however, this feature did not appear to either promote or inhibit subjunctive marking in the second interview.) It may be that in the very earliest of stages of acquisition of the Spanish subjunctive an overtly marked syntactic frame may help to promote subjunctive marking, but that in later stages this is not a factor. The identification of these syntactic, morphological and semantic features and the analysis of their influence on the promotion or inhibition of subjunctive marking for these learners supports the claimtthat variability is constrained and caused by an interaction of multiple linguistic factors, including syntax, morphology and semantics. The analysis of the variability of the use of alternative forms used when the subjunctive was not marked not only supports Hypotheses Four, it also suggests that perhaps in order for acquisition to take place, especially for syntactically and semantically complex structures such as the subjunctive, variability in the use of other verb forms is a necessary precondition. In the first interview, subjunctive marking was accompanied by a great deal of variety in the use of alternate forms. Where there was variety of alternate 199 forms in the first interviews but no subjunctive marking, in the second interview subjunctive marking was seen. Also, as acquisition increased, ‘variability ‘was seen. to decrease. Onset of acquisition of a form is accompanied by variability as the learner tries out new forms. As acquisition develops and stabilizes, variability decreases. This variability reveals that learners use different strategies in order to communicate in and process different levels of the L2. In the data it was seen that there are occasions when these learners transfer the L1 sentence structure, almost word-for—word, .At other times, they construct the L2 sentence structure but insert Ll morphology. At others, they use the L2 sentence structure and use some other verb morphology which reflects neither the L1 nor the L2. On some occasions, this morphology reflects a semantic notion that the sentence expresses, as in the choice of the future indicative for a possible future action. In other instances, the morphology appears 'subjunctive-like' as in the choice of the third person singular in place of the first person singular. Finally, when acquisition is taking place, these learners construct sentences with both L2 syntax and morphology. The final stage of acquisition would be the interaction of L2 syntax, morphology, semantics and pragmatics. At all times and at all stages there is evidence of interaction of the learner's knowledge of the syntactic, morphological and semantic features of the target language and of his/her 200 accessing of universal learning strategies such as L1 transfer, overgeneralization, hypothesis formation and prototype schema construction. The findings support Hypothesis Five; these adult L2 learners of Spanish produced abundant complex sentences in both interviews. In subordinate clauses where the present indicative is allowed or required, these subjects made very few errors. In both interviews the lack of the use of the present subjunctive was not due to the lack of use of subordinate clauses. It appears that subordination is neither a prohibiting nor a promoting factor for subjunctive use for these learners. The acquisition process is not a simple one, nor one that is the result of a few factors. The results of this study and others reveal that the learner must take into account and does take into account, syntactic, morphological, semantic and pragmatic kinds of linguistic knowledge in order to construct the grammar of his second language. If these kinds of knowledge make up his linguistic ‘ competence then it is inevitable that this knowledge would interact in significant ways to produce the kinds of structures evident in interlanguage. 8.2. Insights into the nature of the Subjunctive Mood and Acquisition In Interview One there was very little use of the present subjunctive. It was used in the complement verbs following 201 only two syntactic frames. Accompanying this very limited subjunctive use was a great variety in the use of other verb forms. In Interview Two subjunctive use increased dramatically for complement verbs of these two syntactic frames; variability in the use of other verb forms decreased slightly. Subjunctive use also increased for complement verbs following other syntactic frames but not for others. Most notably there was no subjunctive use following the syntactic frames expressing doubt no creer que 'to not believe that', the adverbial syntactic frames hasta/sin/para que 'until/without/so that' and never with any adjectival clauses. Recall that dubitative frames such as no creer que were problematic for researchers to account for. The complement verbs following this syntactic frame can take either the indicative or the subjunctive mood based on the point of view or the intentions of the speaker. Faced with making this subtle pragmatic distinction in their speech, it is possible that these learners, incapable of doing so, opt for using the indicative mood, their default verbal form when unable to make decisions -- or unaware of their prerogative to do so. Although these learners use the subjunctive with the adverbial syntactic frame cuando 'when' (50% of the time in Interview Two), they never use it with the adverbial frames hasta/sin/para que. From.this evidence we cannot make claims about adult learners acquiring certain grammatical functions, such as nominal clauses before adverbial clauses. However, these learners are clearly distinguishing between two 202 different types of adverbial clauses requiring subjunctive mood. Learners are taught that the indicative mood is used following cuando when it indicates a habitual action or event, and the subjunctive when it indicates a possible future action or event. These are fairly comprehensible time concepts for learners. Hasta/sin/para que adverbials, which obligatorily require the subjunctive, are more complex time concepts and also indicate more of a conditional meaning than cuandb. This complexity may be confusing for learners and may explain why these learners do not use the present subjunctive mood with these adverbials. It can be seen from.the difficulties researchers have had in describing and accounting for the subjunctive mood, and the difference in the use and acquisition of the present subjunctive according to syntactic frame and verbal complement by these adult learners, that the subjunctive is indeed a complex structure. In addition, we can see how syntactic structure, grammatical function, meaning and pragmatic features of the subjunctive mood all combine to make some aspects easier to use and acquire for learners of Spanish as a second language. 8.3. Implications for Teaching From the results of this study of the second language acquisition of the present subjunctive in Spanish, a few implications for teaching may be drawn. First, if learners exhibit a great deal of variability when a structure is just 203 about ready to be acquired or is in the process of being acquired, perhaps we need to take another look at the business of correcting errors. Since 1967, Corder has been urging both language teachers and researchers to take a look at the kinds of errors learners make, claiming that we can learn a lot about the language learning process from error analysis. For over a decade now researchers such as Dulay, Burt and Krashen (1982), among others, have maintained that error correction does little good since the majority of language development occurs as acquisition, which is the unconscious development of language. Although one may wish to argue with their theoretical basis for supporting this issue, they may well have been on the right track by suggesting that we take another look at error correction. When is error correction helpful to learners? The results of this study seem to suggest that thereis a natural development or sequence of development for the Spanish present subjunctive. This development is accompanied by a great deal of variability where the learner is hypothesizing about the correct form to use in different linguistic environments. This variability shows errors, but these errors are, most of the time, quite logical. Sometimes the errors are the result of language transfer from L1, sometimes from misunderstandings about how to form the subjunctive, sometimes based on meaning and at other times due to not attending to the almost imperceptible L2 verb morphology . n) 204 At times, these learners were searching for the foam and when the form was provided by the native speaker, they recognized it and even used the correct form later in the conversation. At these times, it appears that error correction was helpful and useful and contributed to their language learning process. At other times, the native speaker provided the correct form.and the learner appeared not even to notice and did not incorporate the correct form into his/her speech. When the teacher notices that the learner's speech (or writing) is accompanied by a great deal of variability, this may be the sign that the learner is on the verge of acquiring a certain structure and that error correction is necessary in order to assure complete acquisition. Another implication for teaching that can be drawn from this study is the fact that semantics and pragmatics must be ' incorporated into pedagogical descriptions and explanations, especially for syntactic structures like the Spanish subjunctive mood, whose complexity and use depends so much on these levels of the linguistic system. Semantic and pragmatic descriptions must not be left for the more advanced learner. It was seen that these are not aspects of 'late level' rules to be applied after the learner masters the syntactic and Inorphological structure, but rather form.part of the structure as a whole. Understanding more about the nature of the Spanish ssubjunctive mood, how native speakers understand and use it and how children and adults acquire it surely will benefit 205 learners if the results of these studies are incorporated into materials development and language teaching methods. 8.4. Directions for Future Research As always, many questions remain unanswered and new questions arise as one attempts to conclude a research study. This research dealt with the linguistic factors which influence and constrain the production of a certain linguistic structure. It dealt with only one social context: the informal interview situation, although it compared and contrasted the development of the structure from an earlier to a later stage. Although it has been claimed by many that beginning learners are not capable of manipulating a variety of styles and are more influenced by linguistic factors than by external social factors, other research, especially with more advanced learners, may need to include multiple social/psychological factors as independent variables influencing variation in the use of forms. In addition, writing samples were looked at but, most likely because the subjunctive was not encouraged or cued in any way in these samples, learners chose to avoid using it. If the subjunctive had been encouraged, perhaps through the assignment of specific writing topics, or through discrete grammar tests or exercises, would the same amount of ‘mariability be seen as in the oral interviews? As learners aadvance is there more or less variability in the use of other syntactic frames requiring the present subjunctive? Itvaa see: acc a1t rel int the 1101 UN IE' is C8 116 206 seen that acquisition or the onset of acquisition is accompanied by a great deal of variability in the use of alternate forms, but that this variability is systematic and related either to processing or learning strategies interacting with multiple linguistic factors. Is this always the case for all structures? Or do other structures which are more syntactically and semantically/pragmatically less complex not show this pattern? There was virtually no variation in the use of the present indicative mood in subordinate clauses requiring the indicative. It can be suggested, then, that it is not the syntactic or morphological complexity of forms that causes variability in the use of alternate forms but rather the semantic and pragmatic complexity. This claim, however, needs to be researched in more depth. Other studies analyzing the ‘variability of complex syntactic, semantic and pragmatic forms need to be carried out in order to substantiate this claim, For example, does the acquisition of the past subjunctive pattern in the same way as the present subjunctive? Does the acquisition of aspect, the distinction between preterit and. imperfect past tenses, exhibit similar variability? Is the acquisition of passives, or reflexives, or prepositions also as complex as the (acquisition of the present subjunctive? There are, indeed, lmany interesting and important questions to address in future research endeavors . 207 8.5. Conclusion The study of adult second language acquisition is the study of interlanguage. The study of interlanguage is a study of variability. Therefore, analyzing variability in the acquisition and use of specific structures, especially those which are syntactically, semantically and pragmatically complex, is a necessary and valid endeavor and one which can contribute to a theory of second language acquisition. In addition, the results of such studies can lend insight not only into the acquisition process but into the nature of the linguistic system.itself. Finally, by understanding how the learner's knowledge of this linguistic system interacts with both his/her L1 system and with processing and learning strategies, we can come to a better understanding of how to instruct future learners who need and want to learn another language. APPENDICES APPENDIX A APPENDIX A THE INTERVIEW FORMATS Interview One (Conducted five days after arrival): A. Opening: 1. 2 3. 4 Hola, gqué tal? (gComo estas?) gComo te llamas? gDe donde eres? gEn donde estudias? gQué carrera estudias? (gQué estudias?) gTe gusta México? gQué te parece Querétaro? (gTe gusta Querétaro?) gComo fue el viaje? gComo llegaste? gen avién? gen tren? go Como? B. Development: 7. 8. 10. 11. 12. 13. 14. 15. gSi no estuvieras aqui en Querétaro, gen donde estarias ahora? (Qué haces normalmente en estos meses de abril, mayo, y junio?) Cuando regreses a los Estados Unidos, gqué haras? gQué vas a hacer cuando regreses a los Estados Unidos?) gHabias estudiado espafiol antes? (gHas estudiado espanol antes?) gComo estan las clases? gTe gustan hasta ahora? En este programa, gqué quieres que tus maestros hagan por ti? gQué quieres que te ensefien? gQué esperas que te pidan de ti? gQué piensas hacer cuando termines tu carrera? (tus estudios) gY tus padres? gQuieren que sigas estudiando, o quieren que trabajes? gQué quieren que tu hagas? gY la familia con quien vives? gComo te parece? gQué esperas de ellos? 208 Int of 16. 209 Si yo fuera a ir a los E.U. para las vacaciones, gqué me recomendarias ver? gA donde me recomendarias ir? C. Closing: 17. ;Ah! Debes visitar , es muy bonito. gTe gusta la comida mexicana? En sirven muy ricas. etc. Interview Two (Conducted three months after arrival at the end of the language program): A. Opening: 1. 2. Hola, gQué tal? gComo te ha ido? gCémo te llamas? (Si la entrevistadora no conoce al entrevistado) Ya, estan terminando el programa. gComo te sientes? gComo te fue? Explicame algo sobre tu experiencia aqui. gQué has hecho? gQué has visto? gQué has aprendido? etc. gComo ha sido la experiencia de vivir con una familia mexicana? B. Development: 6. 10. 11. 12. gQué deseas de ellos? gDeseas que te sigan comunicando, o deseas que vayan a visitarte? gQué deseas de ellos? A ver, cuéntame, gha pasado todo que habias esperado? gRealizaste todo que querias realizar aqui? gCumplieron tus'maestros? gHicieron lo que tu querias? gQué mas esperabas de ellos? Cuando regreses a los E.U., gqué haras? Si tuvieras la oportunidad de hacer otra cosa, gqué harias? Si no hubieras venido a México en esta temporada, gqué habrias hecho? gQué esperas que pase en el futuro con tu Vida? 210 gQué quieren tus papas que pase con tu futuro? C. Closing: 13. gExtrafias a tu familia? ga los E.U.? 14. gYa estas listo(a) para regresar? Que tengas un buen viaje, etc. APPENDIX B p—- “In IUI) APPENDIX B OCCASIONS REQUIRING OR ALLOWING PRESENT SUBJUNCTIVE BY SYNTACTIC FRAME Part I. Interview One Es logico/necesario/posible/etc.) que___ l.& 2. En los E.U., cuando se estudia las ciencias de politico, es logico que yo estudio derecho, pero, y, y consiguiré abogado, pero, no sé. (fut. ind.) ‘ 3. Es necesario que yo voy a otro escuela por aprender ser un abogada. (pres. ind.) 4. Para trabajar en un buen trabajo. gsi?, es necesario voy a otra preparacion, gsi? (pres. ind.) 5. No es necesario que esperamos por todos. (pres. ind.) 6. Es posible que mis companeros y yo, ah, vamos a San Miguel Allende este fin de semana. (pres. Ind.) 7. gNo es obvioso que yo necesito estudiar mucho...? (pres. ind.) 8. Es posible que yo estudio la medicina, no sé. (pres. ind.) 9. Es posible que (la situacién) cambiaré en el futuro. Yo espero. (fut. ind.) OA = 9 Variation of Forms = 2 SM = 0 (0%) Pres. Ind. = 7 (78%) Fut. Ind. = 2 (22%) Cuando___ 10.& 11. gCuando yo graduar..graduar...gradué? (inf.) (pret. ind.) 12. Cuando me graduarse de Dartmouth, quiero, voy a gescuela de abogado? (inf.) 211 212 13. Cuando yo, yo estoy, er, gcuando yo tengo, ah, sesenta anos? (pres. ind.) 14. Porque me gustaria trabajar en Espafia cuando graduo. (pres. ind.) 15. No sé que quiero hacer cuando todo es término. (pres.ind.) 16. A: gQué piensas hacer? B: No sé. gCuando yo llegaré? (fut.ind.) 17. Um, cuando, yo, um, gyo me graduo? ;Oh! ;Si!, Um, um, es, voy a estudiar, mas. (pres.ind.) 18. Oh, cuando, mmm, me vuelva a mi casa yo, yo voy a elmorzar con mi familia. (pres. subj.) 19. Ah, cuando ah, re- regresa-ré a los E.U., voy a comenzar estudi—, estudiar una vez mas, una... (fut. ind.) 20. A: Cuando te vayas, gte gustaria seguir? gescribirles una carta? ghablarles por teléfono? gvenirles a visitar? B: ...cuando yo salga. gSi? (pres. subj.) ‘21. Cuando hablar espanol, uh, may, muy bien, uh, me gustaria leer libros de espanol, um, mucho. (inf.) 22.& 23. A: gVas a seguir en contacto con ellos? B: Si, si, si, y también, um, si, um, yo regrese, gregresaré? a Mexico, quiero que, uh visi...gnos visitan? (pres.ind./WF) (fut. ind.) 24. A: Entonces, gcrees que la clase cambia? B: En el futuro, cuando haceré, haré tres o cuatro semanas. (fut. ind.) OA = 15 Variation of Forms = 7 SM = 2 (13%) Pres.Ind. = 4 (27%) Put. Ind. = 4 (27%) Pret.Ind. = 1 ( 7%) 213 Inf. = 3 (20%) Pres.Ind/WP = 1 ( 7%) Querer que 25. No, mi parentes no demand nada, de mi; quieren estar, quieren mi estar alegre...alegre. (inf.) 26.& 27. Quiero que ellos saben, que creemos, si nos intentamos. (pres. ind.) (pres.ind./WF) 28. A: Y tus padres, gquieren que sigas estudiando? o, gquieren que trabajes? B: Quieren que yo estudiando. (pres. part.) 29. Yo quiero estudiar y, porque eso, mi padres quieren que yo estudiar, gsi? (inf.) 30. Mis abuelos quieren que practica, practica como una abogado, pero no quiero ser un abogado. (pres.ind./WF) 31. Mis padres quieren que haga que quiero hacer. (subj.) 32. Mi novia me quiere ir a Tequisquiapan, pero no sé que quiero hacer. (inf.) 33. A: gQue quieres que tus maestros te ensefien? B: Um, solamente, ah, (que) me ayudan con mi espanol y especialmente con, ah, con mi gaprehension? y, porque no puedo entender, uh, espanol, porque, la gente hablan muy rapidamente, uh, grapido? (pres. ind.) 34. Um, mis padres, um, me quieren que estudiar ahora, pero, es no problema Si no quiero estudiar ahora pero trabajar. (inf.) 35. Ahora, ellos, ellos, es que quieren que yo soy en la universidad, después, pero si yo voy a estudiar medicina, hay muchas afios de la escuela. (pres. ind.) 36.-39. .A: gTus papas? gtus padres? gQué opinan? 0 sea, gquieren que s1gas estudiando, o que, o que trabajes, o gqué qu1eren que tu hagas? 214 B: Quieren...que...que yo, terminé, gtermine? (si) mis, mis estudios, en la universidad, y tambien quieren que me gdivertirse? (gcomo se dice?) que me gdivertirsa? (pret.ind.) (pres. subj.) (inf.) (Invented Form) A: Quieren que me divierta. B: Si. 40.-42.B: (Later) Quieren que me divierta, uh, mmm, gque se alegre? Cosas como asi... Uh, después de, de la universidad, de mis estudios, quieren que yo busca un, un trabajo. (pres.subj.) (pres. subj.) (pres.ind./WF) A: Que busques un trabajo. B: Si. 43. Nada mas, quiero que ellos sepan que yo estoy muy feliz en Querétaro, en Mexico, y, uh, es que...(si). (pres. subj.) 44.& 45. A: Tus padres en los E.U. gQuieren que sigas estudiando? go que trabajes? .B: Creo que ellos me quieren continuar mis estudios. (inf.) A: Ah, 0 sea que... B: Si, porque ellos, um, quieren um, me quieren, um, tener una vida mejor que ellos, y... (inf.) A: Ah, gque te superes? B: Si. 46. Quiero que, uh, visi... gnos visitan? or no, no, si, gles visitan? (pres.ind./WF) 47.& 48. Oh, Mis padres quieren me estudiar Quieren que yo gestudie? Pero yo tengo mi trabajo. (inf.) (pres. subj.) 49.& 50. Quiero que los maestros, um, nos dicen como es el Mexico de hoy y nos ayudana la, uh, el espanol correcto. La gramatica. Mi gramatica es horrible. (pres. ind.) (Inv.Form) 51.& 52. Mis padres quieren que yo hago lo que yo quiero. Pero no quieren que estudia para toda la vida. (pres.ind.) (pres.ind.) 215 53. No sé, es duro, pero ellos quieren que trabajar por el “Peace Corps“ por poco tiempo, por un afio. (inf.) 54. Y quiero que (nosotros) ver mucho de Querétaro. (inf.) OA = 30 Variation of Forms = 7 SM = 6 (20%) Pres.Ind. = 6 (20%) Inf. = 10 (33%) Gerund = 1 ( 3%) Pret. Ind. = 1 ( 3%) Pres.Ind/WP = 4 (13%) New Form. = 2 ( 7%) No pensar que___ 55. Yo no pienso ellos son divorciados. (pres.ind.) 56. No pienso que entiendo, no sé, no sé si entiendo o no entiendo, gsi? la pregunta. (pres.ind.) OA = 2 No Variation = 1 Form SM = 0 (0%) ' Pres. Ind. = 2 (100%) Recomendar que 57. Recomendo tu, tu vas a..., I don't know. Es dependiente. (pres.ind.) 58. Yo recomiendo que visitar' Nueva York 3y Boston. y Washington, D.C. (inf.) 59. & 60. B: Ah, yo recomiendo que tu, ah, ves, gves? gvistes? ah...(pres.ind.) (pret.ind.) A: Veas. B: ...veas, um, la capital, washington, D.C.; es muy interesante... 216 61. Yo recomiendo que tu viajar a New Hampshire y Chicago porque soy de Chicago. (inf.) OA = 5 Variation of Forms = 3 SM = 0 (0%) Pres. Ind. = 2 (40%) Inf. = 2 (40%) Pret. Ind. = 1 (20%) Esperar que 62. En el futuro yo espero que (la clase) cambiar. (inf.) 63.-67. A: gQué esperas de ellos? (de la familia) B: No mucho, solamente, que, que ellos me gustan a mi y que comprenden que yo voy con mis amigos mucho tiempo y (que) tengo clases todo el tiempo, y (que) ellos permiten que yo voy a la cocina y preparar mi own comida y todos, no es necesario que esperamos por todos. (pres.ind.) (pres.ind.) (pres.ind.) (pres.ind.) (inf.) 68.—71. A: gQué piensas que ellos esperan de ti? B: Ah, de que hablar mejor y de escribir mejor y de, de que tratar y de géchole ganas? (inf.) (inf.) (inf.) (Inv.Fomm) 72. A: gQué crees que los maestros esperan de ti, en este curso? B: Oh, umm, yo espero, um, que aprender mas espanol. (inf.) 73.-77. A: gQué esperas que tus maestros pidan de ti? B: gQué, qué quieren de mi? A: 0 sea, glos maestros? gQué esperan de ti? B: De que ameliorar, eh, (gcémo se dice?) gameliorar? gmi espafiol? ges? gno? Gracias, uh, (gComo se dice?) de hacer mejor mi espanol, de aprender dichos... 217 A: Te gustan los dichos, gverdad? B: Si, gqué mas? De que escribir mejor en espanol, mm, y de hablar mejor. (inf.)(inf.)(inf.)(inf.)(inf.) 78.-81. A: gQué esperas tu de la familia con quien vives? B: De que placticar conmigo, de, que me ayuda, ah, de, de hacer mi experiencia aqui grata, umm, de hacer memorias lindas de Querétaro, de Mexico. (inf.) (pres. ind.) (inf.) (inf.) 82.& 83. A: gY qué esperas que te pidan de ti? 0 sea, gqué esperas que los maestros te pidan a cambio? B: Oh, no sé. Ummm... A: Que aprendan. B: Si. Que aprendo, y... ah, también, ah, ah, espero que, uh, conocer’ con otros estudiantes... de esta universidad. y de Querétaro también. (pres. ind.) (inf.) 84. También, si ellos, ah...van a los, va a los E.U., uhm, espero que ellos me visitan tambien. (pres. ind.) 85. A: En este programa, gqué quieres que tus maestros te ensefien? B: Espero que ellas ayudarme, ayudarme, um, hablar espanol mejor. (inf.) 86. Mmm, espero que, que ellos reciben mis cartas, pero, um, pero... (pres. ind.) 87.-89. A: gQué esperas que te pidan... de ti? B: Um, que yo esta aqui todas las dias, y que escucha que, todos que los maestros dicen, y que yo hago mi mejor trabajo. (pres.ind,/WF) (pres.ind./WF) (pres.ind.) 90. A: gY qué esperas de la familia con quién vives? B: La familia es perfecto, me gusta mucho. Que ellos hacen todo lo que hacen ahora. (pres.ind.) 218 91.& 92. Ah, espero...que, uh, espero que aprender (espanol), uh, me ayudé espanol. (inf.) (pret.ind.) 93. B: Ellos esperan que yo estoy aprendi-, aprend—, apren- (pres.ind.) A: Aprenda espanol. B: Si, aprenda espafiol. 94. Espero que placticar mucho con ellos, para que luego puedo aprender espanol y conocerlos mas. (inf.) 0A = 33 Variation of Forms = 5 SM = 0 (0%) Pres. Ind. = 11 (33%) Inf. = 18 (55%) Pres.Ind/WP = 2 ( 6%) Pret. Ind. = 1 ( 3%) New Form. = 1 ( 3%) Antes/después de/para que___ 95. Regresamos a los E.U. el 14 de junio y pues, habra un, una semana libre antes de que nosotros debemos a regresar a la Universidad de Dartmouth. (pres.ind.) 96. Yo voy a ir a Kentucky, mi estado, y, um, manejaré a Dartmouth después de yo visito con mi familia un poco tiempo. (pres.ind.) 97. Espero que platicar mucho con ellos, para que luego puedo aprender espanol. (pres.ind.) 98. Es una joyeria - es un estudio. Es para que los estudiantes trabajar con metales. (inf.) 0A = 4 Variation of Forms = 2 SM = 0 (0%) Pres. Ind. = 3 (75%) Inf. = 1 (25%) 219 Gustarse que___ 99. A: gQué quieres que te ensefien? B: A mi, me gusta que me ensefiar hablar espanol. (inf.) 100. A: gQué quieres que te ensefien los maestros? 13: .A mi, me gusta para las maestras, ah, ensefiar ah, algo...dificil y nuevo. (inf.) OA = 2 No Variation = 1 Form SM = 0 (0%) Inf.= 2 (100%) Lo que/a donde 101. Mis padres quieren que, (lo) que es la mejor para mi, pero, no sé. (pres.ind.) 102. Quieren (lo) que yo quiero. (pres.ind.) 103. It dependiente on donde quieres ir. (pres.ind.) 104. Mis padres quieren que haga (lo) que quiero hacer, gsi? (pres.ind.) 105. ...pero ellos (sus papas), ellos 1e gustan cual que yo me gusta, si, si yo soy alegre, ellos son alegre tambien. (pres.ind.) 106. Mi padre me dijo, 'Lo que tu quieres.“ (pres.ind.) 107. ...y que escucha que, todos (lo) que los maestros dicen, y que... (pres.ind.) 108. Mis padres quieren que yo hago lo que yo quiero. (pres.ind.) 8 No Variation = 1 Form 0 (0%) Pres.Ind. = 8 100%) 8 u N 220 Part II. Interview Two. Es posible/importante/necesario/etc.) que 1. Es posible que yo vaya otra vez antes de salir México. (pres. subj.) 2. Posiblemente, es posible que él tiene demasiado transparencias. (pres. ind.) 3. Hace cuatro dias que estamos esperando una contesta y es posible que tenemos cuatro mas. (pres. ind.) 4. No es importante que no sea la verdad o la realidad. (pres. subj.) £5. Cada dia la cosa mas importante es que él termina los transparencias. (pres. ind.) 6. Si la clase es por 2 1/2 horas...no es un problema...es necesario que, que terminamos la lectura. (pres. ind.) 7. & 8. Es necesario que manejar...ah, de, que maneja de Kentucky hasta, hasta Dartmouth en dos dias, porque...para que llego para la escuela en tiempo. (inf.) (pres.ind./WF) 9. Pero, desafortunadamente, es necesario que regresarémos. (fut. ind.) 10. Es posible que me queda una semana mas, pero no sé. (pres.ind./WF) 11. E8 posible que regresa la proximo afio; pero todavia no estoy seguro. (pres.ind./WF) 12. Es necesario que nosotros continuarémos, pero alla en los E.U. (fut. ind.) 13. Y es, siempre es muy util que uno sepa una lengua como espanol. (pres. subj.) OA = 13 SM = 3 (23%) Variation of Forms = 5 Pres.Ind. = 4 (31%) Inf. . = 1 (8%) Put. Ind. = 2 (15%) Pres.Ind./WP = 3 (23%) 221 Cuando 14. & 15. B: Me gusto mucho las playas de México y cuando termina..., gtermine? A: Si, termine. B: ...termine la escuela, la escuela de Dartmouth, voy a vivir en la playa. (pres.ind./WF)(pres. subj.) 16. Cuando llegarémos al aeropuerto, vamos a sentir tristes. (fut. ind.) 17. Voy a comprar una fabrica de sillas y papeles cuando ganaré mucho dinero. (fut. ind.) 18. Cuando regreso a los Estados Unidos, puedo agradecer a Dios. (pres. ind.) 19. Las extrafiaré muchas cosas cuando salga de Querétaro. (pres. subj.) 20. Necesito estar en mi casa y, por eso, cuando vuelva a los Estados Uhidos, voy a dormir y comer:mucho, finalmente. (pres. subj.) 21. Necesito primero comprar regalos para mi familia, y luego, cuando los tenga, voy a comprar una chamarra de piel. (pres. subj.) 22. Los extrafio todavia pero mas cuando vaya de Querétaro. (pres. subj.) 23. Voy a sentir alegre una vez mas cuando regresaré a mi casa. (fut. ind.) 24. Cuando salgo Mexico, llevaré con mi muchas cosas pero mis, lindos, ah, las lindas memorias. (pres. ind.) 25. Y luego me dijo, 'ven aqui cuando quieres." (pres. subj.) 26.8: 27. Es que, cuando regresa, ah, no...regrese a los Estados Unidos, teneré macho prisa porque necesito ir a clases de Dartmouth en el verano. (pres.ind./WF) (pres. subj.) 28. & 29. Tengo clases de ciencias de la tierra, pero cuando es, cuando sea otofio, voy a regresar a México para estudiar los montanas y la geologia. (pres. ind.) (pres. subj.) 30. Cuando yo vuelva atmi casa, los voy a sorprender. (pres. subj.) 222 31” Ellos no entienden que yo... cuando yo regresaré a mi casa, que no voy a descansar bastante. (fut. ind.) 32. Cuando regrese a los Estados Unidos, voy a mi casa y, pero solamente para cinco dias. (pres. subj.) 33. Cuando terminan clases aqui, voy a ser muy feliz y, pero muy triste a la mismo tiempo. (pres. ind.) 34. A: gY qué esperas que pase en el futuro? gYa cuando termines tus estudios? B: gCuando termine aqui? (pres. subj.) A: Si. 35. cuando regresaré a Dartmouth, quiero mostrar, mostrar bien. (fut. ind.) 36. Nb sé que haria, que haré cuando termine la universidad. (pres. subj.) 37. No pienso acerca de mis experiencias en México ahora, solamente sobre mis experiencias cuando regresa a casa. (pres.ind./WF) ' OA = 24 SM = 12 (50%) Variation of Forms = 4 Pres. Ind. = 4 (17%) Put. Ind. = 5 (21%) Pres.Ind./WF = 3 (12%) Querer/desear que___ 38. Mi mama quiere que 'Los Pistones'I ganen. (pres. subj.) 39. No quiero que la comida me hace enfermo. (pres. ind.) 40. Si van todos, quiero que se hagan amigos otra vez. (pres. subj.) 41. Quiero que terminen los clases pero no quiero regresar a Dartmouth ahora. (pres. subj.) 223 42. Quiero que terminen mis clases para regresar pronto a mi casa. (pres. subj.) 43. Solamente no quiero regresar a los Estados Unidos pero si, quiero que mis clases terminen pronto. (pres. subj.) 44. Mis hermanos siempre quieren que yo los ayudé con su inglés. (pret. ind.) 45. Solo deseo que le guste comer los moscos. (pres. subj.) 46. Ellos no quieren que yo vaya tampoco. (pres. subj.) 47. & 48. Mis papés aqui quieren que yo aprenda todo pero mas que disfruta las experiencias aqui. (pres.subj.) (pres.ind./WF) 49. Quiero que mi familia, mis amigos aqui me visiten porque yo espero visitarlos de nuevo. (pres. subj.) 50. & 51. Quiero nos hablan y nos escriben cartas y tarjetas de posta. (pres. subj.) (pres. ind.) 52. Ella no quiere que yo viva aqui para siempre, nada mas un rato. (pres.subj.) 53. Ellos no quieren que yo vaya. (pres. subj.) 54. No es seguro pero quieren que, que yo se doctora en medicina, de medicina. (pres. Ind.) 55. Yo quiero que mi familia me visite en el futuro. (pres. subj.) 56. Quiero que ellos vayan alla...ojala puedan. (pres. subj.) 57. El siempre quiere que nosotros no hacemos caso a el tiempo. (pres.ind.) OA = 20 SM = 14 (70%) Variation of Forms = 4 Pres.Ind. = 4 (20%) Pret.Ind. = 1 ( 5%) Pres.Ind./WP = 1 ( 5%) 224 No creer que 58. Pero esto era antes y ahora no creo que piensan asi. (pres. ind.) 59. No creo que él sabe todo que el piensa. (pres. ind.) 60. Ellos siempre hablaron en inglés y yo... para mi, no me gusta esto... no creo que es la mejor cosa para aprender espanol. (pres. ind.) 61. No creo que vuelvo a Acapulco otra vez. (pres. ind.) 62. Y no creo que se encuentra hospitalidad asi en los Estados Unidos. (pres. ind.) 63. Y'yo, eso es una cosa que, que no me gusta; no...no creo que me gusta, pero para mi seria mejor hablar solamente espafiol. (pres. ind.) 64. Pero ellos no creen que es la verdad, nunca. (pres. ind.) 65. No creo que los mexicanos entienden esto. (pres. ind.) 66. El siempre dice esto y cree que yo hablo mejor que él. (pres. ind.) OA=9 SM = 0 (0%) No Variation of Form Pres.Ind.: 9(100%) Esperar que___ 67. Espero que tengo la gusta a verle otra vez. (pres. ind.) 68. Espero que no voy a estar enfermo otra vez. (paraphrastic fut.) 69. Que tenga un buen dia. (pres. subj.) 70. Esta noche es el préximo partido y espero que sea el filtimo tambien. (pres. subj.) 71.- 73. Espero que podré tener, ah... tendré, no... espero que pueda, gsi?...ah, tener todos los dias y todos los horas con ellos pero eso no es posible. (fut. ind.) (fut. ind.) (pres. subj.) 74. Espero que la semana proxima todo va a ser bien. (para. 225 fut.) 75. A: gQué esperas que pase en el futuro? B: ... que poder ser amigas otra vez. (inf.) 76. Mafiana es el examen de historia y espero que no seré tan dificil que el otro. (fut. ind.) 77. En la fiesta voy a comer mucho pero espero que no me voy a enfermar. (para. fut.) 78. & 79. Espero que puedo, que pueda usar mi espanol otra vez. (pres. ind.) (pres. subj.) 80. Estoy cansado de, cansado de las clases, pero me gusta México mucho y espero que de aprender mucho aqui. (inf.) 81. El espera que nosotros escuchar a él todo el tiempo. (inf.) 82.-85. A: gQué esperas que pase en el futuro con tu vida? B: gCon mi vida en los Estados Unidos? A: Si. gQué esperas? B: Que siga aprendiendo de la, del mundo, de las culturas diferentes y todo...que no olvido mi espanol, que no olvide...ah, mi espanol... y (que) pueda regresar a Mexico pronto. (pres.subj.) (pres.ind.) (pres.subj.) (pres.subj.) 86. A: gQué esperas que pase en el futuro? gcon tu vida? B: Que todo salga bien...ah...eso es todo. (pres. subj.) 87. B: Mis padres esperan esto tambien. A: gQué? gqué esperan ellos? B: Que yo...que salgb adelante. (pres. ind.) OA = 21 226 SM. = 8 (38%) Variation of Forms 2 5 Pres. Ind. = S (24%) Para. Fut. = 3 (14%) Inf. = 2 (10%) Put. Ind. = 3 (14%) Hasta que (cuando)/sin que/para que 88. & 89. No puedo esperar hasta que veo y digo este a mi novio. (pres. ind.) (pres. ind.) 90. Lo que sé ahora es que solo tengo una semana mas hasta que puedo regresar a Los Angeles. (pres. ind.) 91. No sé hasta cuando tenemos clases pero creo que hasta en agosto. (pres. ind.) 92. & 93. No puedo andar en la playa sin que muchas personas me piden dinero o que compro cosas. (pres. ind.) (pres. ind.) 94. Es necesario que manejar...ah, de, que maneja de Kentucky hasta, hasta Dartmouth en dos dias, porque...para que llego para la escuela en tiempo. (pres. ind.) 95. Todos vamos a ir para que celebramos e1 despedido. (pres. ind.) OA 8 No Variation of Form SM 0 (0%) Pres. Ind. = 8 (100%) Alegrarse (de)/estar alegre/contento (de) que___ 96. & 97. Me alegro que clases estén casi terminadas y que no haya mas trabajo.l(pres. subj.) (pres. subj.) 98. Acapulco es un lugar interesante pero estoy alegre de que estoy en Querétaro ahora. (pres. ind.) 99. Estoy alegre que nuestro ensayo de historia esta terminado; ahora tenemos mas tiempo libre para otras cosas. (pres. ind.) 100. Estoy muy contenta de que hablo espanol. (pres. ind.) 101 pue 102 cor 0A SM PIE Inf Oja 103 sen 104 105 101 11 AC: 9)“ 112 227 101. iMe gusta mucho México y Querétaro pero me alegro que ‘pueda regresar a mi pais pronto. (pres. subj.) 102. No fue tan dificil como pensé, pero estoy contento, contenta de que terminar todo bién. (inf.) OA = 7 Variation of Forms = 3 SM = 3 (43%) Pres.Ind. = 3 (43%) Inf. = l (14%) Ojalé (que)___ 103. Ojala que él y yo podamos ir a la playa algun fin de semana. (pres. subj.) 104. Ojala que ellos les gustan México. (pres. subj.) 105. Ojala que ganen los Toros de Chicago. (pres. subj.) 106. Ojala que no tengamos mas tarea ahora. (pres. subj.) 107. Ojala que Querétaro y México estén mas cerca de Dartmouth y Pennsylvania. (pres. subj.) 108. Ojala pueda regresar a México en el futuro cercano. (pres. subj.) 109. Quiero que ellos vayan alla...ojala puedan. (pres. subj.) 7 No Variation of Form 7 (100%) CA SM Other Syntactic Frames 110. Pienso que no hay nadie en nuestro grupo que guste la clase o su estilo de enseflar. (pres. subj.) 111. Regresé un poco desilusionada; no porque no me gusta Acapulco, sino porque he creado Acapulco como paraiso excepcional. (pres. ind.) 112. Yo creo que le gusta e1 hecho de que vivo en un rancho tambien. (pres. ind.) 228 113. Pero como dicemos en inglés, “Haga todo lo que flota su barco.‘ (pres. ind.) 114. Como me falta solamente cinco trimestres en Dartmouth, no sé si sea posible. (pres. subj.) 115. Quien sabe si, 31 lo realice. (pres. subj.) OA II 05 Variation of Forms = 2 SM = 3 (50%) Pres.Ind. = 3 (50%) APPENDIX C APPENDIX C OCCASIONS REQUIRING OR ALLOWING PRESENT INDICATIVE BY SYNTACTIC FRAME . Part I. Interview One Cuando 1. Casi todo estudiantes de los Estados Unidos trabajan en el verano, cuando estan en la universidad. (pres.ind.) 2. Cuando se estudia las ciencias de politico, es logico, que yo estudio derecho. (pres. ind.) 3. Cuando ‘platicamos, necesitamos hablar hasta nuestros pensamientos y forma las palabras en espanol. (pres. ind.) 4. Yo tengo un clase se llama 'drill', cuando yo practico espanol, tambien, durante los veinte semanas. (pres. ind.) 5. Estoy nervioso, nerviosa cuando hablar espanol. (inf.) 6. Mi familia es perfecto, porque estan muy paciente con mi cuando hablo espanol. (pres. ind.) 7. Mi familia ayudan mi cuando hablo espafiol. (pres. ind.) 8. Cuando todos los estudiantes va a escuela durante del verano, después del segundo aflo, casi los estudiantes aqui tienen cuatro dias y then, regresa a Dartmouth. (pres. ind.) 9. Esto es en inglés, cuando hablamos. (pres. ind.) 10. Cuando pica mucho, no me gusta. (pres. ind.) 11. Es muy dificil entender las maestras cuando hablan. (pres. ind.) 12. Les pregunto cuestiones cuando no entiendo, a veces. (pres. ind.) 13. Cuando yo no estoy en Querétaro, estudio en mi escuela, Dartmouth. (pres. ind.) 14. Estas cosas no son las que usaré cuando estoy hablando. (pres.ind.) 14 13 (93%) Variation of Forms = 2 229 CA IM 230 THE. = 1 (7%) Creer/pensar/oir/decir que 15. Creo que yo voy a una otra escuela para matema’lticas. (pres. ind.) 16. Creo que ellos me quieren continuar mis estudios. (pres. ind.) 17. Creo que vamos a San Miguel de Allende.(pres. ind.) 18. Yo pienso que la clase cambiar. (inf.) 19. Pienso que (con) menos tarea y mas conversacion entendimos mas. (pret. ind.) 20. Pienso que son mexicanos. (pres. ind.) 21. Yo pienso (que) el costo or las playas de Nuevo Jersey es el mejor. (pres. ind.) 22. Pienso (que) es en mi casa. (pres. ind.) 23. Pienso que las clases estan bien ahora. (pres. ind.) 24. Pienso que mafiana voy. (pres. ind.) 25. Pienso que los otros y yo vamos a San Miguel de Allende. (pres. ind.) 26. Pienso que me gusta mas la comida aqui que en los Estados Unidos. (pres. ind.) 27. Pienso que hay mucha gente alla. (pres. ind.) 28. Pienso (que) el combinacion e1 sol, la comida es muy rico, muchas cosas. (pres. ind.) 29. Es muy dificil porque nosotros no entendemos espanol tan bien como los profesores piensan que entendemos. (pres. ind.) 30. Pienso que Ud. puede ir a playa o a cine, pero no sé. (pres. ind.) 31. Yo pienso que yo quiero ser un geologa. (pres. ind.) 32. He oido que San Francisco es muy bien, pero no sé. (pres. ind.) 231 33. Pero he oido que las mujeras entran gratis pero, los 'hombres, ;no! (pres. ind.) 34. Oi que hay muchas discoteques. (pres. ind.) 35. Debo decir que me gusta la clase de conversacion porque me gusta aprender los dichos coloquiales. (pres. ind.) 36. Mi profesor me dijo que los estudiantes necesitamos oirnos mas rapido. (pres. ind.) OA = 22 Variation of Forms = 3 IN = 20 (91%) Inf. = 1 (4.5%) Pret Ind.= 1 (4.5%) (No) saber que/donde/como/cual/si :17. Yo no sé si puedo porque es muy lejos y caro. (pres. ind.) 38. Si yo voy alla, yo no sé si voy en avion 0 en camion. (pres. ind.) 39. Nb sé si entiendo o no entiendb la pregunta. (pres. ind.) 40. Yo sé cual es Cancun, pero no fué alla. (pres. ind.) 41. No sé que es 'estarias'. (pres. ind.) 42. No sé qué quiero hacer. (pres. ind.) 43. Pero no sé qué quiero hacer. (pres. ind.) 44. No sé qué vamos a hacer pero probablamente a ir a unas discotecas. (pres. ind.) 45. No sé donde es San Miguel Allende. (pres. ind.) 46. Yo no sé donde exactamente ella vive. (pres. ind.) 47. En los E.U. no sé donde me gusta vacacionar. (pres. ind.) 48. No sé a qué debes ir. (pres. ind.) 49. Nb sé Como yo voy a regresar. (pres. ind.) 50. No sé Como se dice. (pres. ind.) 232 51. the sé Como se dice, pero es muy linda. (pres. ind.) OA IN 15 No Variation 15 (100%) Nominal and Adjectival Que 52. Yo tengo un clase (que) se llama "drill“. (pres. ind.) 53. Es el razon que estoy aqui. (pres. ind.) S4. Solamente hace solo cinco dias (que) estoy aqui. (pres. ind.) 55. Hay cuatro mujeres (que) viven en mi casa tambien. (pres. ind.) ’ 56. Tiene un hijo que tiene, mas o menos, veinticinco afios. (pres. ind.) 57. Hay una sobrina que tienen diecinueve afios también. (pres. ind.) 58. En mis clases, aprendemos sobre las cosas que no valen. (pres. ind.) 59. Normalmente yo soy en mi universidad en Nueva Hampshire (que) se llama "Dartmouth.“ (pres. ind.) 60. and 61. Una cosa aqui es que no pueden comprender que yo me gusta el camién y me gusta caminar y correr. (pres. ind.; pres. ind.) 62. Yo tengo muchos discoteques buenos en los E.U. (que) se llaman, gLa Ciudad Segunda? (pres. ind.) 63. Quiero que ver macho de Querétaro que yo vi solamente de mi casa. (pres. ind.) 64. En los E.U. hay “undergraduate“ que es cuatro afios. (pres. ind.) 65. Después es médico o derecho (que) es la specializacion. (pres. ind.) 66. Las personas son muy simpaticas, los que conocen. (pres. ind./wp) 233 67- iEllos son muy divertidos porque ella le gusta de decir las cosas que yo no conocen, como la ”slang". (pres. ind./wp) 68. and 69. Una cosa que no me gusta es que no es en el centro. (pres. ind.; pres. ind.) 70. Me gusta mucho que hay mucha gente. (pres. ind.) 71. and 72. Vivo en el Estado de Connecticut que es en al noreste de los E.U. en una region que se llama "Nueva Inglaterra' que es 103 seis primeros estados. (pres. ind.; pres. ind.) 73. and 74. A: gEn donde estudias? B: En la, la Universidad que se llama "Dartmouth College“ que es al estado de Nueva Hampshire. (pres. ind.; pres. ind.) 75. Hay muchas cosas que me gusta mucho. (pres. ind.) 76. En los E.U., hay un nombre, A-M—Y, que es “Amy“. (pres. ind.) 77. gY Como se dice la palabra, estos que son en el mar? (pres. ind.) 78. El hijo de catorce ahos, que se llama Isac, es muy, muy amable. (pres. ind.) 79. También a la ciudad de Boston que es una ciudad de Nueva Inglaterra. (pres. ind.) 80. Es una razon que yo voy al Dartmouth. (pres. ind.) 0A = 29 Variation of Forms = 2 IM = 27 (93%) IM/WF = 2 ( 7%) Si 81. Si tu vas a los E.U., puedes venir con mi. (pres. ind.) 82. Es no problema si no quiero estudiar ahora pero trabajar. (pres. ind.) 83. Si yo voy a estudiar medicina, hay muchas afios de la escuela. (pres. ind.) 234 844 Si yo soy alegre, ellos son alegre tambien. (pres. 1nd,) 85.'} También, si ellos van a los E.U., espero que ellos me V1S1tan tambien. (pres. ind.) 86. No me gusta el chile si es muy picante. (pres. ind.) 87. Si quieres, puedes visitar mi ciudad también. (pres. ind.) 88. Si yo voy alla, yo no sé si voy en avién 0 en camién. (pres. ind.) 8 No Variation 8 (100%) CA IM Donde___ 89. Donde yo vivo, no estoy cerca de Filadelfia. (pres. ind.) 90. A mi me gusta donde yo vivo, en Colorado, porque me gusta las montanas. (pres. ind.) 91. New Hampshire es donde Darmouth es. (pres. ind.) 92. Son amables y me ayudan todo el tiempo y dicen donde puedo ir en los fines de semanas. (pres. ind.) 93. Me gusta también e1 nortest de donde es mi escuela. (pres. ind.) 94. Pero mi lugar favorito es donde yo vivo. (pres. ind.) OA = 6 No Variation IN = 6 (100%) Porque___ 95. Es muy dificil porque nosotros no entendemos espanol. (pres. ind.) 96. Estaba muy nervioso cuando llegué a México porque no hablo espanol muy bien. (pres. ind.) 235 97- Me gusta la playa porque no puedo ir mucho. (pres. ind.) 98. Mi clases son muy dificil porque las profesores habla espafiol muy rapido. (pres. ind./wp) 99. Me gusta mas e1 norte porque yo vivo en el norte. (pres. ind.) 100. Querétaro es muy agradable a mi porque no es tan grande. (pres. ind.) 101. Me gusta la clase de conversacion porque me gusta aprender los dichos. (pres. ind.) 102. Asisto a las clases de Dartmouth este verano porque el verano antes de el segundo afio at Dartmouth, los estudiantes tienen que asistir a la escuela. (pres. ind.) 103. Yo no vivo con mis padres porque Dartmouth es cinco horas de Nueva York. (pres. ind.) 104. Para mi es facil hablar espanol porque la gramatica parece a francés. (pres. ind.) 105. Ellos son muy divertidos porque ella le gusta de decir las cosas que yo no conocen, como la I'slang". (pres. ind.) 106. Las clases de gramatica y conversacion son un poco fastidioso porque ya sé todo de gramatica. (pres. ind.) '107. Si, porque casi todo estudiantes de los E.U. trabajan en el verano. (pres. ind.) 108. Porque la maestras son muy formales, mas formales que las maestras en los E.U. (pres. ind.) 109. A: gTe gusta mucho la ciudad de Los Angeles? B: Si. Porque vivo ahi. (pres. ind.) 110. Yo estoy 'biased' porque yo vivo en Nuevo Jersey. (pres. ind.) 111. El este es mejor porque yo vivo en el este. (pres. ind.) 112. Si. Porque mucho personas comprender. (inf.) 113. voy a escuela de abogado porque quiero ser abogado en el futuro. (pres. ind.) 114. Mi familia es perfecto porque estaban muy paciente con 236 mi cuando hablo espanol. (imp.) 115. No, porque tenemos vacaciones en marzo y en abril y mayo y junio hay clases. (pres. ind.) 116. Quieren que yo estudiando porque ahora, todos las personas van al universidad para trabajar en un buen trabajo. (pres. ind.) 117. A: gQué me recomendarias ver? B: Ah, realmente, New England, Vermont, New Hampshire y Maine, porque todo es muy hermoso. (pres. ind.) 118. Me gustaria vivir con mi familia y aprender con ellos porque en mis clases aprendemos sobre las cosas que no valen. (pres. ind.) 119. No puedo contestar estas preguntas porque no tenga los "answers”. (subj.) 120. and 121. ...porque no puedo entender espanol porque la gente hablan muy rapidamente, uh, grapido? (pres. ind.; pres. ind./WP) 122. A mi me gusta donde yo vivo, en Colorado, porque me gustan las montafias. (pres. ind.) 123. Me gusta mucho mi clase de literatura porque es interesante. (pres. ind.) 124. Hoy voy a nadar porque yo nado para la, el equipo de Dartmouth. (pres. ind.) 125. Es dificil porque en los E.U. hay "undergraduate“. (pres. ind.) OA = 31 Variation of Forms = 5 IM = 26 (84%) inf. = 1 ( 3%) mmj.=1 (3%) imp. = 1 ( 3%) IM/WP = 2 ( 6%) Part II. Interview Two Cuando___ 237 l. Aun cuando Zipolite es muy pequefia, tiene comida fantastica. (pres. ind.) 2. Cuando todos de nosotros tenemos los mismos errores en nuestras pruebas, el cree que estamos copiando notas. (pres. ind.) 3. Un rio tiene aqua caliente y el otro el agua fria y cuando se juntan, su mezcla es interesante. (pres. ind.) 4. Bailé mucho con muchachos mexicanos porque es tan dificil decir 'no' cuando ellos preguntan si quieres bailar. (pres. ind.) 5. Ahora, me rio cuando pienso a ese comentario. (pres. ind.) 6. Me parece que siempre es tiempo a ir cuando yo siento bien. (pres. ind.) 7. Pienso que cuando Octavio Paz dice que quiere solamente momentos de alegria, significa la alegria de un vida sin remordimientos. (pres. ind.) 8. Cuando la gente oyen mi acento, saben que soy extranjero. (pres.ind./WF) 9. Cuando llegue a‘mi casa, siempre entro con alegria. (pres. subj.) 10. Cuando yo regreso de un viaje tengo feliz y un sentido de seguridad estar en hogar. (pres. ind.) 11. Cuando un mexicano dice, 'Este es tu casa', es la verdad. (pres. ind.) 12. Pero cuando miro alrededor, es facil ver que casi todos estan solitarios, profundamente, de vez en cuando. (pres. ind.) 13. Cuando él habla de una cosa un momento y otro momento, el esta hablando sobre una cosa diferente. (pres. ind.) 14. Cada dia cuando salga de su clase, no estoy segura de qué pasa en esa clase. (pres. subj.) 15. Cuando ellos vuelven a la casa, tenemos tiempo de familia. (pres. ind.) 16. Cuando vamos a1 mercado, es una experiencia un poquito aburrido. (pres. ind.) 17. Cuando pienso sobre mi casa en Nueva York por supuesto 238 'hay mmchas cosas que no puedo esperar ver. (pres. ind.) 18. Se tiene cuidado cuando se esta bebiendo el agua. (pres. cont. ind.) 19. Es muy ironico que Mexico no tiene nada papel o sillas cuando se realiza que haber enfermos de estomago tantos. (pres. ind.) 20. Me hago loco cuando pienso lo. (pres. ind.) 21. Mis hermanos siempre me rien a mi cuando no sé como decir una palabra o una oracion en espanol. (pres. ind.) 22. Cuando no se comprende el espanol, sonrie y sefia afirmativa hecha con la cabeza. (pres. ind.) OA = 22 Pres. Ind. = 19 (86%) Variation of Forms = 3 Pres..Ind/WF = 1 ( 5%) Pres. Subj. = 2 ( 9%) Creer/pensar/oir/decir/sentir/suponer que___, 23. Yo creo que Mexico es un pais muy interesante por muchas razones. (pres. ind.) 24. En los Estados Uhidos,:mucha gente creen especialmente en los universidades, que se necesita dinero ser feliz. (pres. ind.) 25. Yo creo que éste no es verdad aqui. (pres. ind.) 26. Creo que la gente esta acostombrado viendo basura en los calles y el campo. (pres. ind.) 27. Las olas estan enormosas pero no hay surfeadores, cual yo creo es mejor. (pres. ind.) 28. Yo no podria creer qué simpatico Serafin es. (pres. ind.) 29. & 30. Yo creo que Albert Jones es un hombre muy inteligente pero yo creo que él es un maestro horrible. (pres. ind.) (pres. ind.) 31. El cree que estamos copiando notas. (pres. cont. ind.) 32. Pero Albert cree que debo estudiar mas. (pres. ind.) 239 33. Yo creo, aunque, mucho de mi frustracion es por estoy listo regresar a mi casa y estoy listo por unas vacaciones. (pres. ind.) 34. Ella dijo que los muchachos :mexicanos hacen esto solamente con extranjeras. (pres. ind.) 35. Pienso que los misiones son muy interesante. (pres. ind.) 36. Creemos que es huérfano. (pres. ind.) 37. Creo...pero no estoy seguro... que...que los mexicanos ,pueden mas facilmente entender nuestros accentos, que nosotros podemos entender los suyos. (pres. ind.) 38. Creo que es mejor escribir sobre esto que fabricar un cuento alegria. (pres. ind.) 39. a 40. En la ultimo semestre dije que...que nada dura, y creo que eso es la verdad todavia. (pres. ind.) (pres. ind.) 41. Otra persona me dijo hoy que no tiene nadie de amigo. (pres. ind.) 42. Esta manana, Sra. Heny me dijo que tengo dos madres, Sra. Heny y Sra. Pilar. (pres. ind.) 43. Pienso que muchas personas tienen este problema. (pres. ind.) 44. Mucha de la gente de ahi, creen que el PRD es el mejor. (pres. ind.) 45. Todo piensa que tengo mala suerte. (pres. ind.) 46. Pienso que es un poco patologico que el gobierno ha decido a sacar, o gcomo se dice? gexhumar? g31?... a los cuerpos muertos para mostrarlos en un museo de mémias. (pres. ind.) 47. Pienso que los maestros deben dejar la mitad de la tarea para los estudiantes. (pres. ind.) 48. Pienso que cuando Octavio Paz dice que quiere solamente momentos de alegria, significa la alegria de un vida sin remordimientos. (pres. ind.) 49. Los mmchacos mexicanos creen que las muchachas americanas son mas fac1l que las muchachas mexicanas. (pres. 1nd.) 50. Piensan que ellos son mas mejor que todos los otros estudiantes. (pres. ind.) 240 51. La gente nos dijeron que no debemos ir por alla porque es muy peligroso. (pres. ind.) 52. Dijeron que hay banditos que nos robarian. (pres. ind.) 53. Ahora, supongo que yo solo estoy esperando regresar a mis padres y mis amigos en los Estados Unidos. (pres. ind.) OA = 31 No Variation of Form Pres. Ind. = 31 (No) saber que/si___ 54. Yo sé que la gramatica falta mucho, pero para decir eso, no esta mucho mas facil en inglés. (pres. ind.) 55. Yo sé que hay felicidad en el futuro. (pres. ind.) 56. Sé que tengo amigos may buenos. (pres. ind.) 57. Yo sé que es increible. (pres. ind.) 58. Sé tambien que puedo terminar este programa bien y continuar a mejorar mi espanol después. (pres.ind.) OA = 5 Pres. Ind. = 5 No Variation of Forms Si___, 59. Es tan dificil decir 'no" cuando ellos preguntan si quieres bailar. (pres. ind.) 60. :Ellos desempefian insultados si rm) quieres bailar con ellos. (pres. ind.) 61. Si el amor o...gcomo se dice?...rencor... dura, es que la persona escoge eso cada dia, cada minuto. (pres. ind.) 62. vamos por avion, porque si vamos en camion, tres dias de la vacacion estaria en la calle. (pres. ind.) 63. 81 yo no estoy'mAS enfermo pienso que voy a disfrutarme 241 en‘México los ultimas dias. (pres. ind.) 64. Si las abejas pican como los moscos mexicanos, entonces, debo ser el primer victima. (pres. ind.) 65. Si no soy una artista en el futuro quiero ser un...una abogada. (pres. ind.) 66. No lo hago porque si quiero mejorarme en espanol, necesito escuchar con cuidado y hablar las palabras que yo sé. (pres. ind.) 67. Si no comprendo, necesito hacer preguntas. (pres. ind.) OA = 9 Pres. Ind. = 9 No Variation of Form Donde 68. Hay un restaurante pequeno cerca de la Alameda en la calle donde se compra tacos muy barrotos. (pres. ind.) 69. El es el padre de la familia donde Carine esta viviendo. (pres. cont. ind.) 70. _ En la noche, caminamos cerca de la playa, donde hay muchos bars, discos de playa, y cosas asi. (pres. ind.) 71. Santa Clara de Cobre es un pueblo, como su nombre dice, donde hay mucho cobre. (pres. ind.) 72. Después de la comida, fuimos a Landa en donde hay un pueblo y un misién. (pres. ind.) 73. Nadamos en donde hay un rio caliente y un rio frio. (pres. ind.) OA = 6 Pres. Ind. = 6 No variation of Form Aunque___ 74. Estaba may bien aunque tienen los menus en inglés. (pres. ind.) 242 75. .Aunque siempre hay mucha gente, nos divertimos mucho. (pres. ind.) OA = 2 Pres. Ind. = 2 No Variation of Form Quien/quienes 76. Encontré su hija quien tiene diez y siete afios. (pres. ind.) 77. El museo es la casa de una mujer quien se llama Gertrude Duby Blom. (pres. ind.) 78. ...repugnante, porque hay gente completamente ciega a los pobres quienes sufren en los mercados y las calles. (pres. ind.) 79. & 80. No pueden pensar sobre contrastes espantosos entre los ricos quien viven en la colina y los pobres quien viven en los callejuelas estrechas. (pres. ind.) (pres. ind.) 81. Yo pienso sobre la gente quien tienen solamente un poco. (pres. ind.) OA = 6 Pres. Ind. = 6 No variation of Form Parecer(se) que . 82. Me parece que las tormentas siempre ocurren cuando yo siento sola. (pres. ind.) 83. Me parezco que el mundo es mas chico ahora. (pres. ind.) OA = 2 Pres. Ind. = 2 No variation of Form Como___ 84. a 85. Pero, como el tiempo vuela y como me falta solamente 5 trimestres, no sé si sea posible. (pres. ind.) 243 (pres. ind.) 86. Como hay muchas turistas, hay tanta gente que quiere ser tu guia. (pres. ind.) OA = 3 Pres. Ind. = 3 No Variation of Form LO que_ 87. Lo que es impresionante en Monte Alban es la simetria...simetria de los piramides. (pres. ind.) 88. Es casi todo lo que necesito. (pres. ind.) 89. El tema sobre lo que me pregunto mucho, es esto. . .gporqué la gente tira su basura? (pres. ind.) 90. Pero sobre todo, lo que mas me gusta es la gente. (pres. ind.) 91. No me molesta pero acabo de descubrir en Guanajuato lo que los mexicanos hacen con el muerto. (pres. ind.) 92. Lo que sé ahora es que solo tengo una semana mas aqui. (pres. ind.) OA=5 Pres. Ind. = 6 No Variation of Form Porque___ 93. IMe gusta mucho este lugar porque la comida esta'muy buena siempre. (pres. ind.) 94. Ademas, México parece mas pobre que es porque esta e1 vecino del los E.U. (pres. ind.) 95. Me gusta mucho México aunque, porque la familia es muy importante aqui. (pres. ind.) 96. Me gusta México tambien porque la vida aqui es muy tranquilo. (pres. ind.) 244 97. Fue mejor porque surfeadores pueden hablar solamente de las olas. (pres. ind.) 98. El empezo a sudar mucho porque no le gusta el picante. (pres. ind.) 99. Podemos reconocer a los santos porque tienen cosas que les representan. (pres. ind.) 100. Hemos tenido suerte porque siempre tenemos asientos. (pres. ind.) 101. Bailé mucho con muchachos mexicanos porque es tan dificil decir 'no' cuando ellos preguntan si quieres bailar. (pres. ind.) 102. La semana que viene es pesada porque empiezan los examenes finales. (pres. ind.) 103. Me gusto las ruinas y fue muy interesante porque hace como dos semanas que estudiamos Tula en la clase de historia. (pres. ind.) 104. Nada mas que una persona puede caminar alli porque es tan chica. (pres. ind.) 105. ...repugnante, porque hay gente completamente ciega a los pobres quienes sufren en los mercados y las calles. (pres. ind.) 106. Es may divertido porque alli esta todos los miembros de la familia. (pres.ind./WF) 107. Me gusta hablar con Vania porque ella tiene casi dos afios y todavia no aprende mucho espanol. (pres. ind.) 108. Habia edificios alli que tenian lugares especiales porque significan o corresponden a las estrellas. (pres. ind.) 109. Compramos las verduras y la fruta para nueve familias porque estamos en un cooperativa. (pres. ind.) 110. Me cae bien mi madre porque ella es muy simpatica. (pres. ind.) 2111. Nb me gusto la ciudad de Oaxaca porque pienso que la ciudad es muy aburrida. (pres. ind.) 112. No puedo viajar mas porque no tengo mmcho dinero. (pres. irui.) 113. Tenia problemas con mis pulmones cuando estabamos jugando porque es muy dificil respirar en México por la 245 altitud. (pres. ind.) 114. Me diverti mucho porque me cae muy bien mi hermana. (pres. ind.) 115. Porque no puedo nadar muy bien, no nadé en el mar. (pres. ind.) 116. Para yo, fue una experiencia buena, porque en el futuro quiero ser una artista profesional. (pres. ind.) 117. Acali es como “Gilligan's Island“ porque hay bungalos, muchos arboles con fruta, y en el fondo, hay un lugar con hamacas y asombra. (pres. ind.) 118. Recuerdo Paco y Poncho mas porque ellos son hombres muy amistosos. (pres. ind.) OA=26 Pres. Ind. 2 Pres.Ind./WP = l ( 4%) 25 (96%) Variation of Forms = Nominal and Adjectival Que 119. Podemos reconocer a los santos porque tienen cosas que les representan. (pres. ind.) 120. Fuimos a La Quebrada para ver los clavadistos que saltan muy alto. (pres. ind.) 121. Fui en un camién de segunda clase que va de Pochutla a Oaxaca. (pres. ind.) ' 122. Hay muchas personas que son contentas. (pres. ind.) 123. Tengo muchas experiencias interesantes que no puedo describir todas. (pres. ind.) 124. Bailé con un muchacho que se llama Pablo por mucho tiempo. (pres. ind.) 125. Adentro hay muchos murales que describen la historia de la independencia de México. (pres. ind.) 126. La semana que viene es pesada porque empiezan los examenes finales. (pres. ind.) 246 127. Taxco es un pueblo que me gusta mucho. (pres. ind.) 128. & 129. Tiene muchas calles que suben y bajan. (pres. ind.) (pres. ind.) 130. Encontramos muchos gringos que estén estudiando en Cuernavaca. (pres. cont. ind.) 131. Pienso que los misiones son muy interesante. (pres. ind.) 132. El sabado en la manana, fui a un mercado (que) se llama Mercado de la Cruz. (pres. ind.) 133. Quedamos en un hotel (que) se llama Hotel Oviedo.(pres. ind.) 134. En un calle (que) se llama Juarez, tenia cotorros. (pres. ind.) 135. Quedamos en un hotel (que) se llama “Hotel Regional“. (pres. ind.) 136. Encontramos un hotel barrato que se llama Ramos. (pres. ind.) 137. Mateo y yo fuimos a un museo (que) se llama NaBalom. (pres. ind.) 138. Vive alla un grupo de indiginas (que) se llama Tzotzil. (pres. ind.) 139. Muchas de las personas en El Tech. son fresas pero las personas que conozco alla son simpaticas. (pres. ind.) 140. Fuimos a la biblioteca para ver un fresco de Juan O'Gorman que explica toda la historia de Michoacan. (pres. ind.) 141. También encontramos un hombre que trabaja para 'The Foreign Ministry“ en México. (pres. ind.) .142. Hay un callejon may famoso (que) se llama ‘Callején del Beso.‘ (pres. ind.) 143. Sabado por la maflana fuimos en autobfis a una alberca jllamada 'Taboada' que esta1muy conocido en Guanajuato. (pres. ind.) 144. Fuimos a la Plaza de las Americas para ver una pelicula cnme se llama 'Pasién Otofial'. (pres. ind.) 145. Otra vez, doy Gracias, que las peliculas me cuestan aqui 247 :mero $4,000. (pres. ind.) 146. Domingo, supi que 3mis padres mexicanos son poetas! (pres. ind.) 147. Hay nmchas cosas que me gustan de este lugar. (pres. ind.) 148. San Miguel de Allende esta un lugar que tiene’mucho para los estudiantes. (pres. ind.) 149. Estoy agracedida que puedo tener tantas experienceas buenos. (pres. ind.) 150. Pienso que cuando Octavio Paz dice que quiere solamente momentos de alegria, significa la alegria de un vida sin remordimientos. (pres. ind.) 151. Me gustan los anuncios para PRI/Solidaridad, y un anuncio de una compafiia de pintura que dice “mas vidas de un gato." (pres. ind.) 152. Siempre pasé por el Hotel Villavirgen, que tiene un arbol maravilloso con flores violetas brillantes en el patio. (pres. ind.) 153. No me gusta esta escrutinio que ocurre cada dia. (pres. ind.) 154. No tengo un idea qué voy a hacer. (pres. ind.) 155. Tengo solamente un amigo que puedo hablar sobre esto y estoy agradecido a él. (pres. ind.) 156. La cosa sola que destruye este actitud es indiferencia. (pres. ind.) 157. Quiero ayudar la gente en las ciudades interiores que hablan espanol. (pres. ind.) 158. aHay alguien que falta? (pres. ind.) .159. Cada dia, descrubri nuevos calles y lugares que son ciiferentes y especiales. (pres. ind.) 1160. Hay muchas personas que corren las olas. (pres. ind.) 161. Este es el ultimo fin de semana que estamos en Querétaro. (pres. ind.) 162. Las cosas que el sabe son increibles. (pres. ind.) 163. Pero el nunca recuerda que nosotros no somos como él. 248 (pres. ind.) 164. Siempre hay un bano (que) huele horrible. (pres. ind.) 165. Frecuentamente no entiendo nada que dice. (pres. ind.) 166. Pero ahora es tiempo que regreso a los Estados Unidos. (pres. ind.) 167. Fuimos a una playa muy pequefia que se llama Zipolite. (pres. ind.) 168. Fui a la club 0 disco (que) se llama 'Qui'. (pres. ind.) 169. Ella tiene una hermana (que) se llama Ana. (pres. ind.) 170. Hay muchos personas Americanos que viven en México. (pres. ind.) 171. Después baliamos un poco y encontramos con Patricio y los Americanos que estudian espanol en TEcnologico. (pres. ind.) 172. Como hay muchas turistas, hay tanta gente que quiere ser tu guia. (pres. ind.) 173. Puerto Escondido es una playa pequefia con mucha gente que les gusta surfear. (pres. ind.) 174. En Puerto Escondido, fuimos a un bar que se llama “Coco's“. (pres. ind.) 0A = 56 Pres. Ind. = 56 No Variation of Form APPENDIX D APPENDIX D PROTOTYPE FEATURES Part I. Interview One Sentences in which the subjunctive is used: 1. Cuando me vuelva a mi casa, yo voy a almorzar con mi familia. [+OF, +fut, —des, +irreg] 2. Cuando yo salga. gsi? [+OF, +fut, —des, +irreg] 3. Mis padres quieren que haga que quiero hacer, gsi? [+OF, +fut, +des, +irreg] 4. Quieren que yo termine mis estudios. [+OF, +fut, +des, —irreg] S. Quieren que me divierta. [+OF, +fut, +des, +irreg] 6. que se alegre [+OF, +fut, +des, -irreg] 7. Quiero que ellos sepan que yo Eston muy feliz. [+OF, +fut. +des, +irreg] 8. Quieren que yo estudie. [+OF, +fut, +des, -irreg] Sentences in which the subjunctive is not used: Es logico/obvio/necesario are [-fut], {-des]. 1. Es logico que yo estudia derecho. [+OF, -£ut, -des, -irreg] 2. y consiguiré abogado, pero, no sé. [-OF, -fut, -des, -irreg] 3. Es necesario que yo voy a otro escuela. [OF, -fut, -des, +irreg] 4. Es necesario voy a otra preparacion. {-OF, -fut, -des, +irreg] 249 250 5. No es necesario que esperamos por todos. [+OF, —fut, -des, —irreg] 6. gNo es obvioso que yo necesito estudiar mucho? [+OF, -fut, -des, -irreg] Es posible is [+fut], {-des]. 7. Es posible que mis companeros y yo, vamos a San Miguel Allende. [+OF, +fut, —des, +irreg] 8. Es posible que yo estudio la medicina. [+OF, +fut, -des, -irreg] 9. Es posible que (la situacién) cambiaré en el futuro. [+OF, +fut, -des, -irreg] Cuando___ is [+fut],[—des]. 10. and 11. gCuando yo graduar...gradué? [+OF, +fut, -des, -irreg] [+OF, +fut, -des, —irreg] 12. Cuando me graduarse de Dartmouth, voy a escuela de abogado. [+OF, +fut, -des, -irreg] 13. gCuando yo tengo sesenta anos? [+OF, +fut, -des, +irreg] 14. Porque me gustaria trabajar en Espafla cuando graduo. [+OF, +fut, -des, —irreg] 15. No sé que quiero hacer cuando todo es término. [+OF, +fut, -des, -irreg] 16. No sé. gCuando yo llegaré? [+OF, +fut, -des, -irreg] 17. Cuando yo me graduo, voy a estudiar mas. [+OF, +fut, -des, -irreg] 18. Cuando regresaré a los E.U., voy a comenzar estudiar una vez mas. [+OF, +fut, -des, -irreg] 19. Cuando hablar espanol muy, mmy bien, me gustaria leer libros de espanol. [+OF, +fut, -des, -irreg] 251 20 and.21. Yo regresa, gregresaré? a Mexico, quiero que nos visitan. [-OF, +fut, -des, -irreg] {-OF, +fut, -des, —irreg] 22. En el futuro, cuando haceré, haré tres o cuatro semanas. [+OF, +fut, -des, +irreg] Querer que is [+fut], [+des]. 23. Mi parentes no demand nada de mi; quieren estar, quieren mi estar alegre. [-OF, +fut, +des, -irreg] 24 and 25. Quiero que ellos saben, que creemos, si nos intentamos. [+OF, +fut, +des, +irreg] [+OF, +fut, +des, -irreg] 26. Quieren que yo estudiando. [+OF, +fut, +des, -irreg] 27. Mi padres quieren que yo estudiar. [+OF, +fut, +des, -irreg] 28. Mis abuelos quieren que practica como una abogado. [+OF, +fut, +des, -irreg] 29. Mi novia me quiere ir a Tequisquiapan. [—OF, +fut, +des, +irreg] 30. zone quieres que tus maestros te ensefien? Solamente me ayudan con mi espanol y especialmente con mi comprehension porque no puedo entender. [-OF, +fut, +des, -irreg] 31. Mis padres me quieren que estudiar ahora. [+OF, +fut, +des, -irreg] 32. Ellos, es que quieren que yo soy en la universidad. [+OF, +fut, +des, +irreg] 33. Quieren que yo terminé mis estudios. [+OF, +fut, +des, -irreg] 34. Tambien, quieren que me adivertirse? [+OF, +fut, +des, +irreg] 35. ...que me adivertirsa? [+OF, +fut, +des, +irreg] 36. quieren que yo busca un trabajo. 252 [+OF, +fut, +des, -irreg] 37. Ellos me quieren continuar mis estudios. {-OF, +fut, +des, -irreg] 38. Ellos me quieren tener una vida mejor que ellos. [-OF, +fut, +d-OFs, +irreg] 39. Quiero que nos visitan. [+OF, +fut, +des, -irreg] 40. Mis padres quieren me estudiar. [—OF, +fut, +des, -irreg] 41.and 42. Quiero que los maestros nos dicen como es el Mexico de hoy y nos ayudana el espanol correcto. [+OF, +fut, +des, +irreg] [-OF, +fut, +des, -irreg] 43. Mis padres quieren que yo hago lo que yo quiero. [+OF, +fut, +des, +irreg] 44. Pero no quieren que estudia para toda la vida. [+OF, +fut, +des, -irreg] 45. Ellos quieren que trabajar por el 'Peace Corps" por poco tiempo. [+OF, +fut, +des, -irreg] 46. Quiero que (nosotros) ver mucho de Querétaro. [+OF, +fut, +des, -irreg] No pensar___ is [-fut],[-des]. 47. Yo no pienso ellos son divorciados. [-OF, -fut, -des, +irreg] 48. No pienso que entiendo; no sé si entiendo o no entiendo la pregunta. [+OF, -fut, -des, +irreg] Recomendar que is [+fut], [+des]. 49. Recomendo tu, tu vas a..., I don't know. Es dependiente. [-OF, +fut, +des, +irreg] 50. Yo recomiendo que visitar' Nueva York. y Boston. y Washington, D.C. [+OF, +fut, +des, -irreg] 51. and 52. Yo recomiendo que tu aves?, eves? gvistes? (given 253 the correct form) veas la capital, Washington, D.C.; es muy interesante [+OF, +fut, +des, +irreg] [+OF, +fut, +des, +irreg] 53. Yo recomiendo que tu viajar a New Hampshire y Chicago porque soy de Chicago. [+OF, +fut, +des, -irreg] Espero que___ is also [+fut], [+des]. 54. En el futuro yo espero que (la clase) cambiar. [+OF, +fut, +des, ~irreg] 55. Solamente que ellos me gustan a mi. [-OF, +fut, +des, —irreg] 56. y que comprenden que yo voy con mis amigos mucho tiempo [—OF, +fut, +des, -irreg] 57. y (que) ellos permiten [-OF, +fut, +des, -irreg] 58. que yo voy a la cocina [-OF, +fut, +des, +irreg] 59. y (que) preparar mi own comida [-OF, +fut, +des, -irreg] 60. (Ellos esperan de me) de que hablar mejor l-OF, +fut, +des, -irreg] 61. y de (que) escribir mejor {-OF, +fut, +des, -irreg] 62. y de que tratar y [-OF, +fut, +des, -irreg] 63. y de (que) aéchole ganas? [-OF, +fut, +des, -irreg] 64. (Creo que ellos esperan) que (yo) aprender mas espanol. [+OF, +fut, +des, -irreg] 65. (Los maestros esperan de mi) de que ameliorar mi espanol [-OF, +fut, +des, -irreg] 66. de (que) hacer mejor mi espanol [-OF, +fut, +des, -irreg] 67. de (que) aprender dichos {-OF, +fut, +des, -irreg] 254 68. De que escribir mejor en espanol {-OF, +fut, +des, -irreg] 69. y de (que) hablar mejor. [-OF, +fut, +des, -irreg] 70. De que placticar conmigo [-OF, +fut, +des, -irreg] 71. de que me ayuda [-OF, +fut, +des, -irreg] 72. de (que) hacer mi experiencia aqui grata [—OF, +fut, +des, +irreg] 73. de (que) hacer memorias lindas de Queretaro, de Mexico [-OF, +fut, +des, +irreg] 74. Que aprenda. [-OF, +fut, +des, -irreg] 7S. Espero que conocer' con otros estudiantes de esta universidad y de Queretaro también. [+OF, +fut, +des, +irreg] 76. Espero que ellos me visitan tambien. [+OF, +fut, +des, -irreg] 77. Espero que ellas ayudarme hablar espanol mejor. [+OF, +fut, +des, -irreg] 78. Espero que ellos reciben mis cartas. [+OF, +fut, +des, -irreg] 79. Que yo esté aqui todas las dias [-OF, +fut, +des, —irreg] 80. y que escucha [-OF, +fut, +des, -irreg] 81. y que yo hago mi mejor trabajo [-OF, +fut, +des, +irreg] 82. Que ellos hacen todo lo que hacen ahora. [-OF, +fut, +des, +irreg] 83. Espero que aprender (espanol). [+OF, +fut, +des, -irreg] 84. ... me ayudé espanol. [-OF, +fut, +des, -irreg] 85. Ellos esperan que yo estoy aprendi-, aprend-, apren- 255 [+OF, +fut, +des, +irreg] 86. Espero que platicar mucho con ellos. [+OF, +fut, +des, -irreg] Antes/después de/para que is [+fut],[-des]. 87. Habra una semana antes de que nosotros debemos regresar a la universidad. [+OF, +fut, -des, -irreg] 88. Manejaré a Dartmouth despues de yo visito con mi familia un poco tiempo. [—OF, +fut, -des, -irreg] 89. Espero platicar con ellos, para que luego puedo apender espanol. [+OF, +fut, -des, -irreg] 90. Es para que los estudiantes trabajar con metales. [+OF, +fut, -des, -irreg] Gustarse que___ is [+fut],[+des]. 91. A mi, me gusta que me enseflar hablar espanol. [+OF, +fut, +des, —irreg] 92. A mi, me gusta para las maestras enseflar algo dificil y ’nuevo. [-OF, +fut, +des, -irreg] Lo que/a donde___ (depends on context) 93. Mis padres quieren que, (lo) que es la mejor para mi. [+OF, +fut, +des,, +irreg] 94. Quieren (lo) que yo quiero. [+OF, +fut, +des, -irreg] 95. It dependiente on donde quieres ir. [-OF, +fut, +des, +irreg] 96. Mis padres quieren que haga (lo) que quiero hacer. [+OF, +fut, +des, +irreg] 97. Ellos 1e gustan cual que yo me gusta. [—OF, +fut, +des, —irreg] 98. Mi padre me dijo, 'Lo que tu quieres.“ [+OF, +fut, +des, +irreg] 256 99. y que escucha todo (lo) que los maestros dicen {-OF, +fut, -des, +irreg] 100. Mis padres quieren que yo hago lo que yo quiero. [+OF, +fut, +des, -irreg] Part II. Interview Two Sentences in which the subjunctive is used: Es posible/necesario que is [+fut], {-des]. 1. Es posible que yo vaya otra vez antes de salir México. [+OF, +fut, +des, +irreg] 2. No es importante que no sea la verdad o la realidad. [+OF, -fut, -des, +irreg] 3. Y es, siempre es mmy util que uno sepa una lengua como espanol. [+OF, -fut, -des, +irreg] Cuando___ is [+fut],[-des]. 4. Cuando gtermine? 1a escuela de Dartmouth, voy a vivir en la playa. [+OF, +fut, -des, —irreg] 5. Las extrafiaré muchas cosas cuando salga de Querétaro. [+OF, +fut, -des. +irreg] 6. Cuando vuelva a los Estados Unidos, voy a dormir y comer mucho, finalmente. [+OF, +fut, -des, +irreg] 7. Necesito primero comprar regalos para.nd,famdlia, y luego, cuando los tenga, voy a comprar una chamarra de piel. [+OF, +fut, -des, +irreg] 8. Los extrafio todavia pero mAS cuando vaya de Querétaro. [+OF, +fut, -des, +irreg] 9. Y luego me dijo, 'Ven aqui cuando quieres.“ [+OF, +fut, +des, +irreg] 10. Cuando regresa, regrese a los Estados Unidos, teneré mucho prisa. [+OF, +fut, -des, -irreg] 257 '11. Cuando sea otofio, voy a regresar a México. [+OF, +fut, - des, +irreg] 12. Cuando yo vuelva a mi casa, los voy a sorprender. [+OF, +fut, ~des, +irreg] 13. Cuando regrese a los Estados Unidos, voy a mi casa. [+OF, +fut, -des, -irreg] 14. gCuando termine aqui? [+OF, +fut, —des, —irreg] 15. No sé que haré cuando termine la universidad. [+OF, +fut, -des, -irreg] Querer/desear que___ is [+fut], [+des]. 16. Mi mama quiere que 'Los Pistones' ganen. [+OF, +fut, +des, —irreg] 17. Quiero que se hagan amigos otra vez. [+OF, +fut, +des, +irreg] 18. Quiero que terminen los clases pero no quiero regresar. [+OF, +fut, +des, -irreg] 19. Quiero que terminen mis clases para regresar pronto. [+OF, +fut, +des, -irreg] 20. Si, quiero que mis clases terminen pronto. [+OF, +fut, +des, -irreg] 21. Solo deseo que le guste comer los moscos. [+OF, +fut, +des, -irreg] 22. Ellos no quieren que yo vaya tampoco. [+OF, +fut, +des, +irreg] 23. Mis papas aqui quieren que yo aprenda todo.. [+OF, +fut, +des, -irreg] 24. Quiero que mi familia, mis amigos aqui me visiten. [+OF, +fut, +des, -irreg] 25. Quiero nos hablen y nos escriben cartas y tarjetas de posta. [+OF, +fut, +des, -irreg] 26. Ella no quiere que yo viva aqui para siempre. [+OF, +fut, +des, -irreg] 27. Ellos no quieren que yo vaya. 258 [+OF, +fut, +des, +irreg] 28. Yo quiero que mi familia me visite en el futuro. [+OF, +fut, +des, -irreg] 29. Quiero que ellos vayan alla. [+OF, +fut, +des, +irreg] Esperar que is [+fut], [+des]. 30. Que tenga un buen dia. [-OF, +fut, +des, +irreg] 31. Espero que sea e1 ultimo tambien. [+OF, +fut, +des, +irreg] 32. Espero que pueda tener todos los dias y todos los horas con ellos. [+OF, +fut, +des, +irreg] 33. Espero que pueda usar mi espanol otra vez. [+OF, +fut, +des, +irreg] 34. (Espero) Que siga aprendiendo del mundo. [-OF, +fut, +des, +irreg] 35. ...que no olvide mi espanol... [-OF, +fut, +des, —irreg] 36. y (que) pueda regresar a México pronto. [-OF, +fut, +des, +irreg] 37. (Espero) Que todo salga bien. [-OF, +fut, +des, +irreg] Alegrarse que___ is [+des], {-fut]. 38. & 39. Me alegro que clases estén casi terminadas y que no haya mas trabaj o . [+OF, ~fut, +des, +irreg] [-OF, -fut, +des, +irreg] 40. Me alegro que pueda regresar a mu pais pronto. [+OF, -fut, +des, +irreg] Ojala ._.is [+des],[¢fut]. 41. Ojala que él y yo podamos ir a la playa algun fin de semana. [+OF, +fut, +des, +irreg] 42. Ojala que ellos les gustan México. 259 [+OF, +fut, +des, -irreg] 43. Ojala que ganen los 'Toros' de Chicago. [+OF, +fut, +des, -irreg] 44. Ojala que no tengamos mas tarea ahora. [+OF, +fut, +des, +irreg] 45. Ojalé que Querétaro y México estén mas cerca. [+OF, -fut, +des, +irreg] 46. Ojala pueda regresar a México en el futuro cercano. [+OF, +fut, +des, +irreg] 47. ...ojalé puedan. [+OF, +fut, +des, +irreg] ’Other' Syntactic Frames vary. 48. Pienso que no hay nadie en nuestro grupo que guste la clase. [-OF, -fut, +des, -irreg] 49. ...no sé si sea posible. [+OF, +fut, -des, +irreg] 50. Quien sabe si, si lo realice. [-OF, +fut, -des, -irreg] Sentences is which the subjunctive is not used: Es posible/necesario que___ is [+fut], {-des]; es importante___ is {-fut]. {-des]. 1. Posiblemente, es posible que él tiene demasiado transparencias. [+OF, +fut, -des, +irreg] 2. Es posible que tenemos cuatro mas. [+OF, +fut, -des, +irreg] 23. Cada dia la cosa mas importante es que él termina los transparencias. [+OF, -fut, -des, -1rreg] 4. Es necesario que terminamos 1a lectura. [+OF, +fut, -des, —irreg] 5. Es necesario que manejar, de, que maneja de Kentucky hasta, hasta Dartmouth. [+OF, +fut, -des, -irreg] 260 {-OF, +fut, -des, -irreg] 6. Pero, desafortunadamente, es necesario que regresarémos. [+OF, +fut, -des, -irreg] 7. Es posible que me queda una semana mas. [+OF, +fut, -des, -irreg] 8. Es posible que regresa la proximo afio. [+OF, +fut, -des, -irreg] 9. Es necesario que nosotros continuarémos, pero alla en los E.U. [+OF, +fut, -des, —irreg] Cuando is [+fut], {-des]. 10. Cuando llegarémos al aeropuerto, vamos a sentir tristes. [+OF, +fut, -des, -irreg] 11. Voy a comprar una fabrica de sillas y papeles cuando ganaré mucho dinero. [+OF, +fut, -des, -irreg] 12. Cuando regreso a los Estados Unidos, puedo agradecer a Dios. [+OF, +fut, —des, -irreg] 13. VOy a sentir alegre una vez mas cuando regresaré a mi casa. [+OF, +fut, -des, -irreg] 14. Cuando salgo‘México, llevaré con md muchas cosas. [+OF, +fut, —des, +irreg] 15. Ellos no entienden que cuando yo regresaré azmi casa, que no voy a descansar bastante. (+OF, +fut, -des, -irreg) 16. Cuando terminan clases aqui, voy a ser muy feliz y. [+OF, +fut, -des, -irreg] 17. Cuando regresaré a Dartmouth, quiero mostrar bien. [+OF, +fut, -des, -irreg] 18. No pienso acerca de mis experiencias en Mexico ahora, solamente sobre mis experiencias cuando regresa a casa. [+OF, +fut, -des, -irreg] I Querer que is [+fut], [+des]. 19. No quiero que la comida me hace enfermo. 261 [+OF, +fut, +des, +irreg] 20. iMis hermanos siempre quieren que yo los ayudé con su inglés. [+OF, +fut, +des, —irreg] 21. Mis papés aqui quieren que yo aprenda todo pero mas que disfruta las experiencias aqui. [—OF, +fut, +des, —irreg] 22. Quiero nos hablen y nos escriben cartas y tarjetas de posta. [-OF, +fut, +des, -irreg] 23. No es seguro pero quieren que, que yo se doctora en medicina. [+OF, +fut, +des, -irreg] 24. El siempre quiere que nosotros no hacemos caso a el tiempo. [+OF, +fut, +des, +irreg] No creer que___ is {-fut], {-des]. 25. Pero esto era antes y ahora no creo que piensan asi. [+OF, -fut, -des, -irreg] 26. No creo que él sabe todo que el piensa. [+OF, -fut, -des, +irreg] 27. No creo que es la mejor cosa para aprender espafiol. [+OF, -fut, —des, +irreg] 28. No creo que vuelvo a Acapulco otra vez. [+OF, +fut, -des, +irreg] 29. Y no creo que se encuentra hospitalidad asi en los Estados Unidos. [+OF, -fut, -des, +irreg] 30. No creo que me gusta. [+OF, -fut, -des, -irreg] 31. Pero ellos no creen que es la verdad, nunca. [+OF, -fut, -des, +irreg] 32. No creo que los mexicanos entienden esto. [+OF, -fut, -des, +irreg] 33. El siempre dice esto y cree que yo hablo mejor que él. [+OF, -fut, -des, -irreg] Esperar que is [+fut], [+des]. 262 34. Espero que tengo la gusta a verle otra vez. [+OF, +fut, +des, +irreg] 35. Espero que no voy a estar enfermo otra vez. [+OF, +fut, +des, +irreg] 36. Espero que podré tener todos los dias y todos los horas con ellos. [+OF, +fut, +des, +irreg] 37. Espero que la semana proxima todo va a ser bien. [+OF, +fut, +des, +irreg] 38. ( Espero) ... que poder ser amigas otra vez. [~OF, +fut, +des, +irreg] 39. Mafiana es el examen de historia y espero que no seré tan dificil que el otro. [+OF, +fut, +des, +irreg] 40. En la fiesta voy a comer mucho pero espero que no me voy a enfermar. [+OF, +fut, +des, +irreg] 41. espero que de aprender mucho aqui. [-OF, +fut, +des, -irreg] 42. E1 espera que nosotros escuchar a él todo el tiempo. [+OF, +fut, +des, -irreg] 43 . (Espero) ...que no olvido mi espanol. [-OF, +fut, +des, -irreg] ' 44. (Espero) Que yo...que salgo adelante. [-OF, +fut, +des, +irreg] Hasta/para que___ is [+fut], {-des]; sin que___ is {—fut], [- des]. 45. & 46. No puedo esperar hasta que very digo este a mi novio. [+OF, +fut, -des, -irreg] [-OF, +fut, -des, +irreg] 47. ... tengo una semana mas hasta que puedo regresar a Los Angeles. [+OF, +fut, -des, +irreg] 48. No sé hasta cuando tenemos clases pero creo que hasta en agosto. [+OF, +fut, -des, +irreg] 49. & 50. No puedo andar en la playa sin que muchas personas me piden dinero o que compro cosas. [+OF, -fut, -des, -irreg] [-OF, -fut, -des, -irreg] 51. Es necesario que manejar, de, que maneja de Kentucky hasta Dartmouth en dos dias, porque...para que llego para la 263 escuela en tiempo. [+OF, +fut, -des, -irreg] 52. Todos vamos a ir para que celebramos e1 despedido. [+OF, +fut, -des, -irreg] Alegrarse (de) que , estar alegre/contento de que is [- fut], [+des]. 53. Acapulco es un lugar interesante pero estoy alegre de que estoy en Querétaro ahora. [+OF, —fut, +des, +irreg] 54. Estoy alegre que nuestro ensayo de historia esté terminado. [+OF, -fut, +des, +irreg] 55. Estoy muy contenta de que hablo espafiol. [+OF, -fut, +des, -irreg] 56. Estoy contenta de que terminar todo bién. [+OF, ~fut, +des, -irreg] ’Other’ Syntactic Frames vary. 57. Regresé un poco desilusionada; no porque no me gusta Acapulco, sino porque he creado Acapulco como paraiso excepcional. [-OF, -fut, -des, -irreg] 58. Yo creo que le gusta el hecho de que vivo en un rancho también. [+OF, -fut, -des, -irreg] 59. Pero como dicemos en inglés, I'Haga todo lo que flota su barco.‘ [+OF, -fut, —des, —irreg] LIST OF REFERENCES LI ST OF REFERENCES Adamson, Hugh D. 1988. variation Theory and Second Language Acquisition. Georgetown University Press. Adamson, Hugh D. 1989. 'Variable Rules as Prototype Schemas.I In Susan Gass, Carolyn Madden, Dennis Preston and Larry Selinker (eds.), variation in Second Language Acquisition Vblume II: Psycholinguistic Issues, pp. 219- 232. Clevedon/Philadelphia: Multilingual Matters. .Adamson, Hugh D. 1990. 'Prototype Schemas, variation Theory, and the Structural Syllabus.’ IRAL 28: 1—25. Al-Kasey, Tamara S. 1993. Sentential Complement.Acquisition in L1 and L2: (Non)Conjunction of Factivity and.Mbod. Unpublished doctoral dissertation, University of Massachusetts. Bardovi-Harlig, Kathleen. 1987. 'Markedness and Salience in Second Language Acquisition.‘ Language Learning 37: 385-407. Bardovi—Harlig, Kathleen. 1992. 'The Relationship of Fonm and Meaning: .A Cross—Sectional Study of Tense and Aspect in the Interlanguage of Learners of English as a Second Language.‘ Applied Psycholinguistics 13: 253-278. Beebe , Leslie, (ed . ) 1988 . Issues in Second Language Acquisition: .Mhltiple Perspectives. New York: Newbury House Publishers. Bell, Anthony. 1980. 'Mood in Spanish: .A Discussion of Some Recent Proposals.‘ Hispania 63: 377-390. Bello, Andrés. 1984. Gkamética de la Lengua castellana. Coleccion Edaf Universitaria. Bialystok, Ellen. 1990. 'The Competence of Processing: Classifying Theories of Second Language Acquisition.‘ TESOL Quarterly 24: 635-648. Blake, Robert. 1991. 'La Adquisicion del Subjuntivo: Intersticio entre la Sintaxis y la Pragmética.‘ In 264 265 Humberto Lopez Morales (ed.), La Ensefianza del ESpafiol como Lengua Materna, pp. 231-241. San Juan, P.R.: Editorial de la Universidad de Puerto Rico. Bolinger, Dwight. 1968. 'Postposed Main Phrases: An English Rule for the Romance Subjunctive.‘ Canadian JOurnal of Linguistics 14: 3-30. Cazden, D., H. Cancino, E. Rosansky, and J. Schumann. 1975. 'Second Language Acquisition Sequences in Children, Adolescents, and Adults.’ Final Report, U.S. Dept. of Health, Education, and welfare, National Institute of Education; Office of Research Grants, Project #730744, Grant #NE-6-00-3-0014. Collentine, Joseph. 1995. 'The Development of Complex Syntax and Mood-Selection Abilities by Intermediate-Level Learners of Spanish.‘ Hispania 78: 122-135. Corder, S. Pit. 1992. 'A Role for the Mother Tongue.‘ In Susan M. Gass and Larry Selinker (eds.), Language Transfer in Language Learning. Amsterdam/Philadelphia: John Benjamins. Crookes, Graham. 1989. 'Planning and Interlanguage Variation.‘ Studies in Second Language Acquisition 11: 367-383. De Villiers, J. 1980. 'The Process of Rule Learning in a Child's Speech: A New Look.‘ In K. Nelson (ed.) Children's Language, Vol 2, pp. 1—44. New York: Gardner. Dickerson, L. 1975. 'The Learner's Interlanguage as a System of variable Rules.‘ TESOL Quarterly 9: 401-407. Dickerson, I“ and VL. Dickerson. 1977. 'Interlanguage Phonology: Current Research and Future Directions.‘ In S. Pit Corder and E. Roulet (eds.), The Notions of Simplification, Interlanguages and Pidgins. Actes du Cinquiéme Colloque de Linguistique Applique de .Neufchatel, pp. 18-30. Neufchatel: Libraire Droz. Dulay, Heidi, Marina Burt and Stephen Krashen (eds.). 1982. Language TWO. Rowley: Newbury House. Eckman, Fred R. 1984. 'Universals, Typologies and Interlanguage.‘ In William.E. Rutherford (ed.), Language Universals and Second Language Acquisition, pp. 79- 105. Amsterdam/Philadelphia: John Benjamins. Ellis, Rod. 1985. 'Sources of variability in Interlanguage.‘ Applied Linguistics 6: 118-131. 266 Ellis, Rod. 1986. understanding Second Language Acquisition. Oxford: Oxford University Press. Ellis, Rod. 1988. 'The Effects of Linguistic Environment on the Second Language Acquisition of Grammatical Rules.‘ Applied Linguistics 9: 257-274. Ellis, Rod. 1990. 'A Response to Gregg.‘ Applied Linguistics 11: 384-391. Floyd, Mary Beth. 1990. 'Sentence Complexity and Clause Subordination in Children's Spanish.‘ Hispania 73: 488- 497. Flynn, Suzanne. 1989. 'The Role of the Head-Initial/Head- Final Parameter in the Acquisition of English Relative Clauses by Adult Spanish and Japanese Speakers.’ In Susan M. Gass and Jacquelyn Schachter (eds.), Linguistic Perspectives on Second Language Acquisition, pp. 89-108. New York: Cambridge University Press. Garcia, Mary Ellen and Tracy Terrell. 1977. 'Is the Use of Mood in Spanish Subject to variable Constraints?‘ In Michio Peter Hagiwara (ed.), Studies in Romance Linguistics: Proceedings of the Fifth Linguistic Symposium on Romance Languages, pp. 214-226. Rowley, Mass.: Newbury House. Gass, Susan. 1984. 'A Review of Interlanguage Syntax: Language Transfer and Language Universals.‘ Language Learning 34: 115—131. Gass, Susan. 1987. 'The Resolution of Conflicts among Competing Systems: A Bidirectional Perspective. ' Applied Psycholinguistics 8: 329-350. Gass, Susan. 1989. 'Language Universals and Second-Language Acquisition.‘ .Language Learning 39: 497-534. Gili y Gaya, Samuel. 1969. Curso Superior de Sintaxis Espafiola.. Barcelona: Bibliograf. Gili y Gaya, Samuel. 1972. Estudios de Lenguaje Infantil. Barcelona: Bibliograf. Goldin, Mark G. 1974. 'A Psychological Perspective of the Spanish Subjunctive.‘ Hispana 57: 295-301. Gonzalez, Gustavo. 1975. 'The Acquisition of Grammatical Structures by Mexican-American Children.‘ In E. Hernandez-Chavez, A.B. Cohen and A.F. Beltramo (eds.), .El Lenguaje de Los Chicanos, pp. 220-237. Arlington, Virginia: Center for Applied Linguistics. 267 Gregg, Kevin R. 1989. 'Second Language Acquisition Theory: The Case for a Generative Perspective.‘ In Susan M. Gass and Jacquelyn Schachter (eds.), Linguistic Perspectives on Second Language Acquisition, pp. 15-40. Cambridge and New York: Cambridge University Press. Gregg. Kevin R. 1990. 'The Variable Competence Model of Second Language Acquisition, and Why it Isn't.‘ Applied Linguistics 11: 364-383. Huebner, Thom. 1983. A Longitudinal Analysis of the .Acquisition of English. Ann Arbor: Karoma Publishers. Huebner, Thom. 1985. 'System and Variability in Interlanguage Syntax.‘ Language Learning 35: 141-163. Hulstijn, Jan H. 1989. 'A Cognitive View on Interlanguage variability.' In‘MiriamLR. Eisenstein (ed.), The Dynamic Interlanguage: Empirical Studies in Second Language variation, pp. 17-31. New York: Plenum Press. Klein, Flora. 1975. 'Pragmatic Constraints on Distribution: The Spanish Subjunctive.‘ In Papers from the Regional Mbeting, Chicago Linguistic Society: 353-365. Klein, Flora. 1980. 'Experimental Verification of Semantic Hypotheses Applied to Mbod in Spanish.‘ In Papers in Language and Linguistics 17: 15-34. Krakusin, Margarita and Artistofanes Cedeno. 1992. 'Seleccién del Modo después de el hecho de que.‘ Hispania 75: 1289-1293. Krashen, Stephen. 1981. Second Language Acquisition and Second Language Learning. Oxford: Pergamon. Krashen, Stephen. 1982. Principles and Practice in Second Language Acquisition. Oxford: Pergamon. Krashen, Stephen. 1985. The Input Hypothesis: Issues and Implications. New York: Longman. Krashen, Stephen and Tracy Terrell. 1983. The Natural .Approach: Language Acquisition in the Classroom. New York: Pergamon/Alemany. Labov, William" 1970. 'The Study of Language in its Social Context.‘ Studium Generale 23: 30-87. Labov, William. 1972. Sbciolinguistic Patterns. Oxford: Basil Blackwell. Lantolf, James R. 1978. 'The variable Constraints on Mood in 268 Puerto Rican-American Spanish.’ In Margarita Sufier (ed.), Contemporary Studies in Romance Linguistics, pp. 193-217. Washington, D.C.: Georgetown University Press. Larsen-Freeman, Diane and Michael Long. 1991. An Introduction to Second Language Acquisition Research. New York: Longman. Lavandera, Beatrice. 1983. 'Shifting Moods in Spanish Discourse.‘ In Discourse Perpectives on Syntax, pp. 209- 236. New York: Academic Press. Lavandera, Beatrice. 1984. 'El Cambio de Modo como Estrategia de Discurso.‘ In Variacién y Significado. Coleccion Hachete Universidad. Lee, James F. 1987. 'Comprehending the Spanish Subjunctive: An Information Processing Perspective.‘ The Modern Language JOurnal 71: 50-57. Liceras, Juana M. 1986. Linguistic Theory and Second Language.Acquisition: The Spanish.annative Grammar of English Speakers. Gunter Narr Verlag Tubingen. Liceras, Juana M. 1989. 'On Some Properties of the 'Pro- drop' Parameter: Looking for Missing Subjects in NOn- native Spanish.‘ In Susan M. Gass and Jacquelyn Schachter (eds.), Linguistic Perspectives in Second Language Acquisition, pp. 109-133. Cambridge and New York: Cambridge University Press. Lunn, Patricia. 1989a. 'The Spanish Subjunctive and I'Relevance"'. In Carl Kirschner and Janet DeCesaris (eds.), Studies in Romance Linguistics: Selected Papers from the Seventeenth Linguistic Symposium on Romance Languages Rutgers university, 27-29.March 1987, pp. 249- 260. Amsterdam/Philadelphia: John Benjamdns Publishing Co. Lunn, Patricia. 1989b. 'Spanish Mbod and the Prototype of Assertability.‘ Linguistics 27: 687-702. Lunn, Patricia and Janet DeCesaris. 1992. Investigacién de Gramatica. Boston: Heinle and Heinle. Mejias-Bikandi, Errapel. 1994. 'Assertion and Speaker's Intention: A Pragmatically Based Account of Mood in Spanish.‘ Hispania 77: 892-902. Olivella de Castells, Matilede. 1981. .MUndo Hispano: Lengua .y Culture. New York: John Wiley & Sons. Onieva Morales, Juan Luis. 1993. La Gramatica de la Real 269 Academia Espafiola. Madrid: Editorial Playor. Preston, Dennis. 1989. Sociolinguistics and Second Language Acquisition. Oxford: Basil Blackwell. Preston, Dennis. 1993. 'Variation Linguistics and SLA.‘ Second Language Research 9: 153-172. Reider, Michael. 1989. 'Clitic Promotion, The Evaluated Proposition Constraint, and Mood in Spanish ‘Verbal Complements.‘ Hispania 72: 283-294. Reider, Michael. 1990. 'Neg—Transportation, Neg-Trace, and the Choice of Mood in Spanish'. Hispania 73: 212-222. Rivero, Maria Luisa. 1971. 'Mood and Presupposition in Spanish'. Foundations of Language 7: 305-336. Rutherford, William. 1982. 'Markedness in Second Language Acquisition.‘ Language Learning 32: 85-107. Rutherford, William. 1989. 'Interlanguage and Pragmatic WOrd Order.‘ in Susan M. Gass and Jacquelyn Schachter (eds.), Linguistic Perspectives on Second Language Acquisition, pp. 163-182. Cambridge and New York: Cambridge University Press. Schachter, Jacquelyn. 1986. 'In Search of Systematicity in Interlanguage Production.‘ Studies in Second Language Acquisition 8: 119-133. Schachter, Jacquelyn. 1989. 'Testing a Proposed Universal.’ In Susan M. Gass and Jacquelyn Schachter (eds.), Linguistic Perspectives on Second Language Acquisition, pp. 134-158. Cambridge and New York: Cambridge University Press. Seco, Rafael. 1954, 1975, 1989. Manual de Gramatica Espafiola. (Undécima Edicion, revisado y ampliado por Manuel Seco), Buenos Aires: Editorial Aguilar. Sperber, Dan and Deirdre Wilson. 1986. Relevance: Communication and Cognition. Cambridge MA: Harvard University Press. Stokes, J. 1988. 'Some Factors in the Acquisition of the Present Subjunctive in Spanish.‘ Hispania 71: 705-710. Stokes, J. 1990. 'Some Factors in the Acquisition of the Present Subjunctive in Spanish: A Re-Analysis.‘ Hispania 73: 805-806. Studerus, Lenard H. 1981. 'A Spanish Twilight Zone: Mood, 270 Syntax, and Past Temporal Reference.‘ Hispania 64: 97- 103. Studerus, Lenard H. 1992. 'Some Unresolved Issues in Spanish Mood Use.‘ .Hispania 78: 94-104. Takagaki, Toshihiro. 1984. 'Subjunctive as the Marker of Subordination.‘ Hispania 67: 248-256. Tarone, Elaine. 1982. 'Systematicity and Attention in Interlanguage.’ Language Learning 32: 69-82. Tarone, Elaine. 1983. 'On the Variability of Interlanguage Systems.‘ Applied Linguistics 4: 142-163. Tarone, Elaine. 1985a. 'The Interlanguage Continuum.‘ In Barbara Wheatly, Ashley Hastings, Fred Eckman, Lawrence Bell, Gary Krukar, and Rita Rutkowski (eds.), Current Approaches to Second Language Acquisition: Proceedings of the 1984 university of Wisconsin-Milwaukee Linguistics Symposium, pp. 31-40. Bloomington: Indiana Linguistics Club. Tarone, Elaine. 1985b. 'Variability in Interlanguage Use: A Study of Style-Shifting in Morphology and Syntax.‘ Language Learning 35: 373-404. Tarone, Elaine. 1988. variation in Interlanguage. London: Edward Arnold. Tarone, Elaine. 1989. 'On Chameleons and Monitors.‘ In Miriam R.Eisenstein (ed.), The Dynamic Interlanguage: Empirical Studies in Second Language variation, pp. 3-15. New York: Plenum Press. Tarone, Elaine. 1990. 'On Variation in Interlanguage: A Response to Gregg.‘ Applied Linguistics 11: 392-400. Terrell, T., B. Baycroft and C. Perrone. 1987. 'The Subjunctive in Spanish Interlanguage: Accuracy and Comprehensibility.’ In Bill vanPatten, Trisha Dvorak and James Lee (eds.), Foreign Language Learning: A Research Perspective, pp. 19-32. Rowley, Mass.: Newbury House. Terrell, Tracy, Magdalena Andrade, Jeanne Egasse and Elias Miguel Munoz. 1986. Dos Mundos: A Communicative Approach. New York: Random House, Inc. Terrell, Tracy D. 1976. 'Assertion and Presupposition in Spanish Complements'. In Maria Lujan and Frank Hensey (eds.), current Studies in Romance Linguistics, pp. 221- 245. Washington. D.C.: Georgetown University Press. 271 Terrell, Tracy and Joan HoOper. 1974. 'A Semantically Based Analysis of Mood in Spanish.’ Hispania 57: 484-94. White, Lydia. 1987. 'Markedness and Second Language Acquisition: The Question of Transfer.‘ Studies in Second Language Acquisition 9: 261-286. White, Lydia. 1989. 'The Adjacency Condition on Case Assignment: Do L2 Learners Observe the Subset Principle?‘ In Susan M. Gass and Jacquelyn Schachter (eds.), Linguistic Perspectives on Second Language Acquisition, pp. 134-158. Cambridge and New York: Cambridge University Press. Wolfram, Walt. 1989. 'Systematic Variability in Second- Language Tense Marking.‘ In Miriam R. Eisenstein (ed. ) . The Dynamic Interlanguage: Empirical Studies in Second Language Variation, pp. 187-197. New York: Plenum Press. Young, Richard. 1988. 'Variation and the Interlanguage Hypothesis.‘ Studies in Second Language Acquisition 10: 281-302. Young, Richard. 1989. 'Ends and Means: Methods for the Study of Interlanguage Variation.’ In Susan Gass, Carolyn Madden, Dennis Preston and Larry Selinker (eds. ) . Variation in Second Language Acquisition Volume II: Psycholinguistic Issues, pp. 65-90. Clevedon/ Philadelphia: Multilingual Matters . Young, Richard. 1991. Variation in Interlanguage Morphology. New York: Peter Lang. Young, Richard. 1993. 'Functional Constraints on variation in Interlanguage Merphology.' Applied Linguistics 14: 76-97. "lilillllllllllllll“