Variable Use of Plural Address Forms in Andalusian Spanish

Home , Andalusian Spanish, Castilian Spanish, Peninsular Spanish, Spanish pronouns

Variable use of plural address forms in Andalusian Spanish

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Elena Jaime Jiménez, M.A.

Graduate Program in Spanish and Portuguese

The Ohio State University

2018

Dissertation Committee

Professor Scott A. Schwenter, Adviser

Professor Rebeka Campos-Astorkiza

Professor Terrell A. Morgan

Elena Jaime Jiménez

2018

Abstract

In Central and Western Andalusian Spanish, both pronominal and verbal variation are found in the use of plural address forms (Lara Bermejo 2015). The pronominal variation involves alternation between the 2PL pronouns ustedes and vosotros as subjects, e.g. ustedes coméis and vosotros coméis ‘you (plural) eat.’ The verbal variation involves the alternation between the 2PL and 3PL verb forms when ustedes is the subject pronoun, e.g. ustedes coméis and ustedes comen ‘you (plural) eat.’ Previous work addresses pragmatic differences between these variants only in terms of contextual (in)formality and (a)symmetric interpersonal relationships, arguing that there are no pragmatic differences between the variants (Alvar 1973; Lara Bermejo

2015; Narbona et al. 1998). Building upon prior research on pragmatic differences between singular address forms (Raymond 2016; Sinnott 2010), in this study I show that plural address form variation in Andalusian Spanish is likewise closely tied to pragmatic differences.

I collected a total of 7,852 tokens from online fora, from Twitter, and from the corpus PRESEEA Málaga. Data also come from an oral elicitation task and an online forced-choice survey with different types of contexts. Data were analyzed using multivariate logistic regression in R, random forests and conditional inference trees.

Results indicate that pre-verbal as opposed to post-verbal placement leads to significantly more use of ustedes over vosotros. Ustedes is also more likely to occur in non-neutral contexts, as shown by the contexts from the elicitation task and online survey, and by the co-occurrence of ustedes with graphic features like emojis in the online sources. Furthermore, ustedes is more frequently produced by males, lower

ii education speakers, and speakers from rural areas. Imperative mood and priming lead to more use of the 3PL verb form over the 2PL with ustedes.

This dissertation reveals that use of the 3PL with ustedes is limited to formulaic expressions of the type juzguen ustedes ‘you guys be the judges.’ It also reveals that ustedes plus the 2PL verb form follows the pattern of its predecessor vuestras mercedes,

‘your honors,’ which also occurred in pre-verbal position more frequently and with second person agreement instead of third person agreement before it was fully stabilized in the system (Fernández Martín 2012). This pattern suggests a possible change in progress, supported by the association of ustedes with males and lower education and the 3PL with rural areas, as well as with priming, which has been identified as conditioning obsolescing linguistic elements (Rosemeyer & Schwenter

2017). Furthermore, ustedes with the 2PL or the 3PL, as opposed to vosotros, conventionally implicates non-neutrality, where the specific attitude toward the interlocutors is context-dependent. This study therefore sheds new light on pronominal shifts in Spanish, which are typically assigned to static social categories (cf. Raymond

2016). It shows how address form switching is used by the same speaker in the same interaction to convey distinct pragmatic meanings. More generally, it contributes to the analysis of plural address forms, which are highly understudied from a cross-linguistic perspective.

Key words: address form switching, second person plural pronouns, Andalusian

Spanish, change in progress, variationist sociolinguistics, pragmatics.

iii

Acknowledgments

First of all, I would like to thank my advisor, Dr. Scott Schwenter. It has been an honor for me to be his advisee. His genius teachings in sociolinguistic variation and pragmatics, and in linguistic research in general, his help, guidance, and expertise have qualified me to conduct solid research and become a professional researcher in all facets, including writing and reviewing abstracts and research papers, and presenting at conferences.

I owe a debt of gratitude to Dr. Rebeka Campos Astorkiza. Thanks to her, I am now also able to conduct research in phonology/phonetics. In addition, she has taught me how to write research papers professionally. Crucially, without her tireless and invaluable continuous help and support in all the aspects of graduate life and work, the writing of this dissertation and my completion of the PhD would not have been possible.

I would like to express my great appreciation to Dr. Terrell Morgan. Taking his Dialectology course and teaching the Spanish Pronunciation class with his fantastic and clever methodology and materials have been inspirational for me. Not only have they strengthened my expertise and knowledge as a linguist, but they have also greatly helped me become a linguistics professor. Furthermore, Terrell has always provided me help when needed.

Lastly, I am eternally grateful to Dr. Juliana De la Mora, who introduced me to linguistics when I was completing a literature-based Master’s degree. This was a turning point in my life; without which I wouldn’t have switched my career to linguistics and without which I wouldn’t have continued to a doctoral program either.

Juliana gave me all the needed inspiration and help that made me enter a doctoral program in linguistics.

Vita

2006...... B.A. English, Arabic and French Translation and Interpreting,

University of Granada, Spain

2008………………………....M.A. Arabic and Hebrew Cultures, Past and Present,

University of Granada, Spain

2013………………………….M.A. Spanish/Linguistics, and Graduate Teaching

Assistant, Department of World Languages, Literatures,

and Linguistics, West Virginia University

2013 to present……………….PhD Hispanic Linguistics, and Graduate Teaching

Associate, Department of Spanish and Portuguese, The

Ohio State University

Awards

2010………………Fulbright Scholarship

2016………………Global Gateway Graduate Student Research Abroad Grant, OSU

2017………………Alumni Grant for Graduate Research and Scholarship, OSU

Publications

Jaime Jiménez, Elena. 2012. Lo que ustedes queráis: A first approach to variation

of second person plural pronouns in Andalusian Spanish. Refugio Poético 2.

Jaime Jiménez, Elena. 2011. Arabic Translations in Spain. Proteus 20 (2). 1-2.

Fields of Study

Major Field: Spanish and Portuguese vi

Table of Contents

Abstract ...... ii Acknowledgments...... iv Vita ...... vi List of Tables ...... x List of Figures ...... xii Chapter 1: Introduction ...... 1 1.1. Definition of the Phenomenon ...... 1 1.2. Outline of the Dissertation ...... 8 Chapter 2: Background ...... 11 2.1. Historical background ...... 11 2.2. Current 2PL Address Form System in Peninsular Spanish ...... 18 2.2.1. Implications for this Dissertation ...... 26 2.3. The Role of Frequency, Priming, and Clause Type in Changes in Progress .... 31 2.3.1. Frequency and Entrenchment Effects ...... 31 2.3.1.1. Frequency and the Autonomy Effect ...... 34 2.3.2. Priming ...... 35 2.3.3. Main and Subordinate Clauses...... 37 2.3.4. Implications for this Dissertation ...... 39 2.4. Pragmatic Factors...... 40 2.4.1. Pragmatic Effects on Historical Address Forms ...... 40 2.4.2. Pragmatic Effects on Current Singular Address Forms ...... 41 2.4.3. Implications for this Dissertation ...... 47 2.4.4. The Pragmatic Meaning of Online Special Features ...... 48 2.5. Research Questions and Hypotheses ...... 51 Chapter 3: Methodology ...... 60 3.1. Data from Online Sources and PRESEEA Málaga...... 64 3.1.1. Corpus and Tokens ...... 64 3.1.2. Data Analysis ...... 67 3.2. Oral Elicitation Task and Online Forced-Choice Survey ...... 75 3.2.1. Online Forced-Choice Survey...... 75 3.2.1.1. Task, Participants and Tokens ...... 75 vii

3.2.1.2. Data Analysis ...... 79 3.2.2. Oral Elicitation Task ...... 84 3.2.2.1. Task, Participants and Tokens ...... 84 3.2.2.2. Data Analysis ...... 89 Chapter 4: Results ...... 96 4.1. Online and PRESEEA Data Results ...... 96 4.1.1. Ustedes/Vosotros Alternation ...... 96 4.1.1.1. Ustedes/Vosotros Alternation in Western Andalusia vs. Central Andalusia ...... 105 4.1.1.2. Ustedes/Vosotros Alternation. Frequent vs. Infrequent Verbs ...... 109 4.1.1.3. Conclusion ...... 111 4.1.2. 2PL/3PL Verb Form Alternation ...... 112 4.1.2.1. 2PL/3PL Verb Form Alternation in Western Andalusia...... 126 4.1.2.2. 2PL/3PL Verb Form Alternation in Central Andalusia ...... 130 4.1.2.3. 2PL/3PL Verb Form Alternation with Frequent Verbs ...... 133 4.1.2.4. Conclusion ...... 139 4.2. Online Forced-Choice Survey Results ...... 141 4.2.1. Ustedes/Vosotros Alternation ...... 141 4.2.1.1. Ustedes/Vosotros Alternation in Western Andalusia ...... 144 4.2.1.2. Ustedes/Vosotros Alternation in Central Andalusia ...... 147 4.2.1.3. Conclusion ...... 151 4.2.2. 2PL/3PL Alternation ...... 152 4.2.2.1. 2PL/3PL Alternation in Western Andalusia ...... 157 4.2.2.2. 2PL/3PL Alternation in Central Andalusia ...... 160 4.2.2.3. Conclusion ...... 163 4.3. Oral Elicitation Task Data Results ...... 164 4.3.1. Ustedes/Vosotros alternation ...... 164 4.3.1.1. Ustedes/Vosotros Alternation in Western Andalusia ...... 168 4.3.1.2. Ustedes/Vosotros Alternation in Central Andalusia ...... 171 4.3.1.3. Conclusion ...... 174 4.3.2. 2PL/3PL Alternation ...... 175 4.3.2.1. Conclusion ...... 178

viii

Chapter 5: Discussion and Conclusions ...... 179 5.1. Change in Progress ...... 180 5.1.1. Pronominal Alternation ...... 180 5.1.2. Verbal Alternation ...... 185 5.2. Discourse-Pragmatic Differences ...... 192 5.2.1. Pronominal Alternation ...... 192 5.2.2. Verbal Alternation ...... 198 5.3. Final Conclusions and Contributions ...... 201 References ...... 207 Appendix A: Online Forced-Choice Survey ...... 215 Appendix B: Oral Elicitation Task 1 ...... 222 Oral Elicitation Task 2 ...... 229

List of Tables

Table 1. Plural address form in different Spanish varieties...... 3

Table 2. Address forms in later Latin and Old Spanish (Penny 2002: 138)...... 11

Table 3. Address forms in the Golden Age (Penny 2002: 138)...... 12

Table 4. Modern pronominal address system in Peninsular Spanish (Penny 2002:

138)...... 12

Table 5. Dependent variables and their variants...... 64

Table 6. Distribution of oral elicitation task participants from Málaga, Central

Andalusia...... 86

Table 7. Best-fit model. Ustedes/vosotros alternation. Online and PRESEEA data. .. 99

Table 8. Best-fit regression model. Ustedes/vosotros alternation. Online and

PRESEEA data...... 103

Table 9. Best-fit regression model for the ustedes/vosotros alternation in Western

Andalusia. Online and PRESEEA data...... 109

Table 10. Best-fit model for the ustedes/vosotros alternation. Frequent verb lemmas.

Priming. Online and PRESEEA data...... 110

Table 11. Best-fit model. 2PL/3PL alternation. Placement. Online and PRESEEA data...... 116

Table 12. Best-fit model for the 2PL/3PL alternation. Intervening material. Online and

PRESEEA data...... 121

Table 13. Best-fit model. 2PL/3PL alternation. Overt pronoun. Online and PRESEEA data...... 124

Table 14. Best-fit model. 2PL/3PL alternation. Frequent verbs. Overt pronoun. Online and PRESEEA data...... 138

Table 15. Best-fit model. 2PL/3PL alternation. Frequent verbs. Intervening material.

Online and PRESEEA data...... 138

Table 16. Best-fit model after collapsing. Ustedes/vosotros alternation. Online forced- choice survey...... 143

Table 17. Best-fit model. Ustedes/vosotros alternation. Western Andalusia. Online forced-choice survey...... 146

Table 18. Best-fit model. Ustedes/vosotros. Central Andalusia. Online forced-choice survey...... 149

Table 19. Best -fit model for the 2PL/3PL alternation. Online forced-choice survey.

...... 156

Table 20. Best-fit model for the 2PL/3PL alternation. Western Andalusia. Online forced-choice survey...... 159

Table 21. Best-fit model for the 2PL/3PL alternation. Central Andalusia. Online forced-choice survey...... 162

Table 22. Best-fit model for the ustedes/vosotros alternation. Oral elicitation task. . 167

Table 23. Best-fit model for the ustedes/vosotros alternation. Western Andalusia. Oral elicitation task...... 170

Table 24. Best-fit model for the ustedes/vosotros alternation. Central Andalusia. Oral elicitation task...... 173

Table 25. Best-fit model for the 2PL/3PL verb form alternation. Oral elicitation task.

...... 177

Table 26. Summary of results...... 179

List of Figures

Figure 1. Geographical distribution of the use of ustedes with the 2PL or 3PL verb form in informal contexts. Adapted from Alvar (1973)...... 5

Figure 2. Implicational hierarchy followed by the progress of the 3PL morphology. . 23

Figure 3. se/os alternation with ustedes plus the 2PL form of the verb...... 29

Figure 4. % of tokens for each variant in the pronominal and the verbal alternations.

Online sources and PRESEEA Málaga...... 66

Figure 5. Distribution of online forced-choice survey participants by education group.

...... 77

Figure 6. Distribution of online forced-choice survey participants by age group...... 78

Figure 7. Rates of pronominal and verbal variants. Online forced-choice survey...... 79

Figure 8. Distribution of participants by age. Oral elicitation task. Western Andalusia.

...... 87

Figure 9. Distribution of participants by education. Oral elicitation task. Western

Andalusia...... 88

Figure 10. Rates of pronominal and verbal variants. Oral elicitation task...... 89

Figure 11. Conditional inference tree. Ustedes/vosotros alternation. Online and

PRESEEA data...... 97

Figure 12. Conditional inference tree. Ustedes/vosotros alternation after collapsing.

Online and PRESEEA data...... 101

Figure 13. Random forest. Ustedes/vosotros alternation. Online and PRESEEA data.

...... 102

xii

Figure 14. Rates of ustedes and vosotros with pre-verbal placement and post-verbal or null placement. Online and PRESEEA data...... 104

Figure 15. Rates of ustedes and vosotros with non-neutral and neutral or absence of online features. Online and PRESEEA data...... 105

Figure 16. Ustedes/vosotros rates in Western Andalusia and in Central Andalusia.

Online and PRESEEA data...... 106

Figure 17. Conditional inference tree. Ustedes/vosotros alternation. Western

Andalusia. Online and PRESEEA data...... 107

Figure 18. Random forest. Ustedes/vosotros alternation. Western Andalusia. Online and PRESEEA data...... 108

Figure 19. Conditional inference tree. 2PL/3PL verb form alternation. Online and

PRESEEA data...... 113

Figure 20. Conditional inference tree for the 2PL/3PL alternation. Factor placement included. Online and PRESEEA data...... 115

Figure 21. Random forest. 2PL/3PL alternation. Placement. Online and PRESEEA data...... 116

Figure 22. Rates of 3PL and 2PL with imperatives and non-imperatives. Online and

PRESEEA data...... 118

Figure 23. Rates of 3PL and 2PL with pre-verbal or null placement and post-verbal placement of the subject. Online and PRESEEA data...... 119

Figure 24. Conditional inference tree. 2PL/3PL alternation. Intervening material.

Online and PRESEEA data...... 120

Figure 25. Random forest. 2PL/3PL alternation. Intervening material. Online and

PRESEEA data...... 121

xiii

Figure 26. Conditional inference tree. 2PL/3PL alternation. Overt pronoun. Online and PRESEEA data...... 123

Figure 27. Random forest for the 2PL/3PL alternation. Overt pronoun. Online and

PRESEEA data...... 124

Figure 28. Conditional inference tree for the 2PL/3PL alternation. Western Andalusia.

Placement. Online and PRESEEA data...... 127

Figure 29. Conditional inference tree. 2PL/3PL alternation. Western Andalusia.

Intervening material. Online and PRESEEA data...... 128

Figure 30. Conditional inference tree. 2PL/3PL alternation. Western Andalusia. Overt pronoun. Online and PRESEEA data...... 129

Figure 31. Conditional inference tree. 2PL/3PL alternation. Central Andalusia.

Placement. Online and PRESEEA data...... 131

Figure 32. Conditional inference tree. 2PL/3PL alternation. Central Andav lusia.

Intervening material. Online and PRESEEA data...... 132

Figure 33. Conditional inference tree. 2PL/3PL alternation. Central Andalusia. Overt pronoun. Online and PRESEEA data...... 133

Figure 34. 2PL/3PL rates with frequent and infrequent verbs. Online and PRESEEA data...... 134

Figure 35. Conditional inference tree. 2PL/3PL alternation. Frequent verbs.

Placement. Online and PRESEEA data...... 135

Figure 36. Conditional inference tree for the 2PL/3PL alternation. Frequent verbs.

Intervening material. Online and PRESEEA data...... 136

Figure 37. Conditional inference tree for the 2PL/3PL alternation. Frequent verbs.

Overt pronoun. Online and PRESEEA data...... 137

xiv

Figure 38. Conditional inference tree. Ustedes/vosotros alternation. Online forced- choice survey...... 142

Figure 39. Random forest. Ustedes/vosotros alternation. Online forced-choice survey.

...... 143

Figure 40. Conditional inference tree. Ustedes/vosotros alternation. Western

Andalusia. Online forced-choice survey...... 145

Figure 41. Random forest. Ustedes/vosotros alternation. Western Andalusia. Online forced-choice survey...... 146

Figure 42. Conditional inference tree. Ustedes/vosotros alternation. Central

Andalusia. Online forced-choice survey...... 148

Figure 43. Random forest. Ustedes/vosotros alternation. Central Andalusia. Online forced-choice survey...... 149

Figure 44. Conditional inference tree. 2PL/3PL verb form alternation. Online forced- choice survey...... 154

Figure 45. Random forest. 2PL/3PL alternation. Online forced-choice survey...... 155

Figure 46. Conditional inference tree. 2PL/3PL verb form alternation, Western

Andalusia. Online forced-choice survey...... 158

Figure 47. Random forest. 2PL/3PL alternation. Western Andalusia. Online forced- choice survey...... 158

Figure 48. Conditional inference tree. 2PL/3PL alternation. Central Andalusia. Online forced-choice survey...... 161

Figure 49. Random forest for the 2PL/3PL alternation. Central Andalusia. Online forced-choice survey...... 161

Figure 50. Conditional inference tree. Ustedes/vosotros alternation. Oral elicitation task...... 165

Figure 51. Random forest for the ustedes/vosotros alternation. Oral elicitation task data ...... 166

Figure 52. Conditional inference tree. Ustedes/vosotros alternation. Western

Andalusia. Oral elicitation task...... 169

Figure 53. Random forest for the ustedes/vosotros alternation. Western Andalusia.

Oral elicitation task...... 170

Figure 54. Conditional inference tree for the ustedes/vosotros alternation. Central

Andalusia. Oral elicitation task...... 172

Figure 55. Random forest for the ustedes/vosotros alternation. Central Andalusia. Oral elicitation task...... 173

Figure 56. Conditional inference tree. 2PL/3PL verb form alternation. Oral elicitation task...... 176

Figure 57. Random forest. 2PL/3PL verb form alternation. Oral elicitation task. .... 177

Figure 58. Pronominal and verbal alternation in Western and Central Andalusia.

Online forced-choice survey...... 186

Figure 59. Productivity, markedness and pragmatic restriction continuum...... 202

xvi

Chapter 1: Introduction

1.1. Definition of the Phenomenon

This dissertation addresses the variable use of plural address forms in Andalusian

Spanish. According to the Diccionario panhispánico de dudas (RAE 2017), in standard

Castilian Spanish, the deferential plural form of address is ustedes ‘you (plural),’ which is morphologically third person plural (3PL) and thus must be followed by a verb conjugated in the 3PL form, e.g. ustedes comen ‘you (plural) eat.’ On the other hand, the non-deferential plural form of address is vosotros ‘you (plural),’ morphologically second person plural (2PL), which must be followed by a verb conjugated in the 2PL form of the verb, e.g. vosotros coméis ‘you (plural) eat.’ In Peninsular Spanish, including Eastern Andalusia but excluding Central and Western Andalusia, vosotros plus the 2PL verb form is used in informal contexts and ustedes plus the 3PL verb form in formal contexts.

In Latin America and the Canary Islands, ustedes plus the 3PL verb form is used in both formal and informal contexts, making it the only form used for addressing a group of interlocutors (Alvar 1973; Fernández Martín 2012; Jiménez Fernández

1999). In the Spanish variety of Andalusia, and, more precisely, in Central and

Western Andalusia, both ustedes and vosotros alternate as subject pronouns of the

2PL verb form in informal contexts, e.g. ustedes coméis and vosotros coméis ‘you

(plural) eat.’ There is also alternation of the 2PL and 3PL verb forms in informal contexts when ustedes is the subject, e.g. ustedes coméis and ustedes comen ‘you

(plural) eat.’ On the other hand, the combination of vosotros with a 3PL verb form does not occur, e.g. vosotros comen ‘you (plural) eat’ (Alvar 1973; Fernández Martín

2012; Narbona et al. 1998). The alternation between the pronouns ustedes/vosotros is illustrated in example (1), whereas the alternation between the 2PL and 3PL verb forms is illustrated in example (2). The examples are taken from the corpus I created specifically for this dissertation and are reproduced as they appear in their original sources, as are all examples that follow. This means that examples are included with lack of diacritics or punctuation, or with the incorrect diacritics or punctuation if that is how they appeared in their original source:

(1) a. Ustedes quereis que sea el mejor central del mundo.

‘You [ustedes] want [2PL] him to be the best center back in the world.’

b. No se que opinareis vosotros.

‘I don’t know what you [vosotros] think [2PL] about it.’

(2) a. por lo menos den ustedes la cara.

‘at least show [3PL] your face, you guys [ustedes].’

b. ¿Ustedes que pensáis?, yo es que nunca me he comprado una consola

recién salida al Mercado.

‘What do you [ustedes] think [2PL]? The thing is, I have never bought a

video game platform which just came out on the market.’

Spanish variety Address form Formality Northern and Central Ustedes coméis Formal Peninsular, and Eastern Vosotros coméis Informal Andalusian Latin American and Canary Ustedes comen Formal and informal Islands Vosotros coméis Informal Central and Western Ustedes coméis Informal Andalusian Ustedes comen Formal and informal

Table 1. Plural address form in different Spanish varieties.

Additionally, in Central and Western Andalusia, there is also a clitic alternation, between 3PL reflexive clitic se and 2PL reflexive clitic os, which co- occur with ustedes and the 2PL form of the verb. The prescriptive rule, on the other hand, dictates that the clitic se must co-occur with a verb in the 3PL or third person singular form and with the subject pronoun usted(es) (RAE 2017). In Central and

Western Andalusia, se/os alternate in informal contexts in co-occurrence with ustedes and the 2PL verb form (Alvar 1973; Jiménez Fernández 1999). Lara Bermejo (2015) argues that 3PL reflexive clitics are more widespread in Cádiz and Seville, in Western

Andalusia. However, Bustos Tovar (2001) argues that the use of se with ustedes and the 2PL verb form, i.e. ustedes se peináis ‘you guys comb your hair’ is an archaism.

He further argues that the form is rural, associated with low education, and stigmatized. The phenomenon is illustrated in (3). The clitic alternation, however, will not be addressed in the present dissertation. This is because of the lack of enough tokens with the reflexive clitic se co-occurring with ustedes and the 2PL verb form,

3 which makes a variationist analysis of the variants se versus os with ustedes plus the

2PL verb form not possible.

(3) a. Ustedes se quejáis pero yo estoy de niñero con mi sobrina!!

‘You guys [ustedes] complain [2PL verb form with clitic se], but I am

babysitting my niece!!’

b. Pero las exigencias que ustedes os hacéis en vuestra cabeza no?

‘But the demands that you guys [ustedes] create [2PL verb form with

clitic se]in your own mind, am I right?’

Central Andalusia includes the provinces of Málaga and Córdoba; Western

Andalusia includes the provinces of Huelva, Seville and Cádiz; Eastern Andalusia includes the provinces of Granada, Jaén and Almería. As illustrated in Figure 1 below, both the pronominal and the verbal alternations occur in Central Andalusia, including the province of Málaga and Southern areas of the province of Córdoba. Both alternations also occur in Western Andalusia, including the provinces of Cádiz,

Seville, and Huelva. On the other hand, those alternations do not occur in Eastern

Andalusia (Alvar 1973). The map in Figure 1 illustrates the geographical distribution of the use of ustedes with the 2PL or the 3PL verb form in informal contexts.

Figure 1. Geographical distribution of the use of ustedes with the 2PL or 3PL verb form in informal contexts. Adapted from Alvar (1973).

The use of vosotros and ustedes is considered one of the most significant linguistic phenomena differentiating Eastern Andalusia from Central and Western

Andalusia. The plural address forms’ alternations in informal contexts only occur in

Western and Central Andalusia (Alvar 1973). Fernández Martín (2012) states that ustedes is used in Western and Central Andalusia in contexts where vosotros is used instead in Eastern Andalusia, i.e. informal contexts. In Eastern Andalusia, only vosotros plus the 2PL verb form is used in informal contexts and ustedes plus the 3PL verb form is reserved for formality and distance, thus following the Castilian pattern.

Andalusian Spanish historically derives from Castilian Spanish, like the other regional Spanish dialects, and has followed its own evolutionary path, which is innovative and differentiated from that of Castilian Spanish (Hernández Campoy &

Villena Ponsoda 2009; Jiménez Fernández 1999; Narbona Jiménez et al. 2011).

However, as mentioned by Narbona Jiménez and colleagues (2011), it is argued that we cannot talk about one single Andalusian dialect, but about different Andalusian

5 dialects because of the lack of uniformity of the Andalusian varieties, which presents different features in the two main differentiated dialectal areas, Western versus

Eastern Andalusia. Nowadays, Eastern Andalusian Spanish is not moving away but converging toward current Castilian Spanish. On the other hand, there is an innovative divergence from the Castilian norm occurring in Western Andalusian Spanish, which has its own center of prestige in the city of Seville. Seville is, in fact, the center of a prestigious vernacular dialect in Western Andalusia, as mentioned by Villena Ponsoda

(2000) and García Mouton (1992). On the other hand, there is a third dialectal area,

Central Andalusia, which has an intermediate position in linguistic terms, i.e. it shows an intermediate linguistic stage between Eastern Andalusian, closer to the Castilian norm, and Western Andalusian, more divergent from the Castilian norm (Villena

Ponsoda 2000).

Previous work on the alternation of ustedes/vosotros and 2PL/3PL verb forms in

Western and Central Andalusia includes brief definitions of the phenomenon and of its geographical distribution (Alvar 1973; Narbona et al. 1998), or offers an historical account of plural address forms in Spanish (Fernández Martín 2012). Jiménez

Fernández (1999) also mentions that the 3PL verb form with ustedes is used by educated speakers whereas the 2PL verb form with ustedes is used by uneducated speakers. Lara Bermejo (2015) associates the use of the pronoun ustedes with the 2PL verb form to rural areas, old speakers and a low level of education. Previous work thus assigns pronominal shifts in Andalusian Spanish solely to static social categories.

Contra authors such as Jimenez Fernández (1999) or Penny (2002), the production of the 2PL or the 3PL forms of the verb with ustedes, as well as the use of the pronouns vosotros/ustedes with the 2PL in Central and Western Andalusia, is not

6 categorical (Narbona et al. 1998). This means that there is inter-speaker and even intra-speaker variation. This is proved by the data collected for a previous study of mine, where the same speaker was found to use both pronominal or both verbal variants (Jaime Jiménez 2014). In example (4) below, the same speaker uses first the

2PL form of the verb, dais, and then the 3PL verb from dan ‘you guys give’.

(4) ¿porque no dais la cara? ¿como abonado del club tengo derecho a saber? Y

por ultimo ¿porque no dan ustedes la cara para con ellos o mejor dicho para con

nosotros.

‘why don’t you guys show [2PL] your face? As a subscriber of the club, I have

the right to know? And last, why don’t you guys [ustedes] show [3PL] your face

for them, or rather, for us.’

In example (5), the same speaker uses both vosotros and ustedes:

(5) Ahora ustedes teneis vuestra fiesta grande que es un coto cerrado para

vosotros y a dia de hoy sigue cerrado para el resto.

‘Now you guys [ustedes] have [2PL] your big party, which is a reserve for you

guys [vosotros], and still today, is closed for others.’

Therefore, social factors alone, including gender, age, education, and origin, are insufficient to account for the variation we observe in Central and Western Andalusia between ustedes and vosotros and between the 2PL and the 3PL verb forms.

Furthermore, whereas level of formality does not explain the variation, no previous study considers whether and how address form switching in Andalusian Spanish is used to convey different pragmatic meanings (cf. Raymond 2016).

There are not, in fact, previous systematic variationist studies on the plural address forms in Central and Western Andalusia. One of the most significant aspects of the

7 present dissertation is that it is the first systematic variationist study of 2PL address forms in Andalusia using inferential statistics and analyzing linguistic and extra- linguistic factors, including social and pragmatic factors.

1.2. Outline of the Dissertation

In Chapter 2 of this dissertation, I review the general historical development of singular and plural Spanish address forms. The goal of this review is to contextualize and let the reader better understand the use of the plural address forms in Central and

Western Andalusia within their historical frame. Then, I address current understanding of the phenomenon based on work up until now. I first review work conducted on plural pronouns, ustedes and vosotros, in Peninsular Spanish, and then I review the methodology and findings from previous work specifically on Andalusian plural address forms. Findings of previous work suggest that the plural address form system in Andalusian Spanish is in a change in progress, but the plural address form system in Andalusian Spanish and the fact that it could be in a change in progress have not been analyzed systematically using variationist methodology.

I then move on to discuss previous findings regarding linguistic factors that have been found to play a role in processes in change, including frequency and priming. In this dissertation, I explore these factors as independent variables in the variationist analysis of the plural address forms in Andalusian Spanish. I then review previous work on singular address forms, which finds that pragmatic differences condition singular address form switching in Spanish. I expand on those pragmatic analyses of address form switching to include plural address forms in Central and

Western Andalusia. Chapter 2 concludes with a discussion of the research questions and hypotheses that drive the present dissertation. These refer to the different patterns of pronominal and verbal variation in Central and Western Andalusia, regarding linguistic, social, and pragmatic factors.

In Chapter 3 of this dissertation, I discuss the methodology. I include a description of the sources from which I gathered the data analyzed in this dissertation, the tokens extracted, the definition of the dependent variables under study, the independent variables or factors considered in this project, and the statistical methods used for the analysis. For collecting data, I used different types of sources, and gathered both already-existing data and elicited data from participants. More precisely, I collected already-existing data from online sources from Central and

Western Andalusia, and from the oral corpus from Málaga (Central Andalusia): the

PRESEEA Málaga. I also elicited data from participants from Central and Western

Andalusia using two tasks, namely, an oral elicitation task and an online forced- choice survey. The dependent variables are the pronominal variable, i.e. ustedes or vosotros, and the verbal variable, i.e. 2PL or 3PL verb forms. Each token was coded for a number of independent variables, including linguistic, pragmatic, and social factors. The methodology of this dissertation is innovative in different ways: it uses naturally-occurring data from conversations in online fora to study plural address forms, and it analyzes the pragmatic effect of the use of online symbols and features, such as emojis, on the variable use of plural address forms. It is also one of the few studies that analyzes the effect of different pragmatic contexts, e.g. anger, apology, neutral contexts, in a variationist study. Furthermore, this dissertation analyzes the effects of priming and frequency on plural subject pronoun expression in a novel way.

In Chapter 4 of this dissertation, I discuss the results. I analyze the results for the pronominal variation and for the verbal variation separately. For each dependent variable, I include the general results for both Central and Western Andalusia, and then the results for each region. I include conditional inference trees showing predictor interactions, random forests showing the hierarchical ordering of the predictors in terms of their statistical significance, and tables showing the best-fit models for each alternation, i.e. the verbal, and the pronominal.

In Chapter 5, I include a final discussion and conclusions. Based on the results obtained, I describe, more generally, how the verbal and the pronominal alternation and how the patterns of variation are different and similar in Central versus Western

Andalusia. I demonstrate that the plural address form system is in a process of change in both Central and Western Andalusia, where there is dialect leveling toward the standard Castilian norm, i.e. vosotros plus the 2PL verb form. I also demonstrate that plural address form switching in Central and Western Andalusia is closely tied to pragmatic differences.

Finally, I discuss the implications and contributions of this dissertation, which looks at plural address forms in Andalusian Spanish using a variationist methodology for the first time, offering a first complete and systematic picture of the phenomenon, by considering different types of sources for data gathering, and by including linguistic and extra-linguistic factors. Furthermore, this study shows that the plural pronominal and verbal variation in Andalusian Spanish is an intricate and heterogeneous phenomenon, not explainable by a few stative categories, as has been assumed before. This dissertation opens the door to more research on plural address forms.

Chapter 2: Background

2.1. Historical background

In Latin, the forms of address were tu for addressing one interlocutor ‘you’ and vos for several interlocutors ‘you (plural).’ Vos was also used as a formal/deferential address form for both singular and plural in later Latin and in Old Spanish, as shown in Table 1. With vos, both singular and plural, formal or informal, the verb form used was the 2PL, ending in -des, from Latin -tis, e.g. cantades ‘you (singular or plural) sing.’ With tú, second person singular verb forms were used, with the second person singular marker ending -s, e.g. cantes ‘you (singular) sing.’

Non-deferential Deferential Singular tú vos Plural vos vos

Table 2. Address forms in later Latin and Old Spanish (Penny 2002: 138).

In the fifteenth century, vos widened its range of use and lost its deferential value. As a result, new forms to show deference appeared, such as merced ‘grace’ in vuestra merced and its plural form vuestras mercedes ‘your grace(s);’ these forms appeared along with morphological third person forms of the verb, to address the interlocutor or interlocutors in a non-direct way, for more deference (Fernández

Martín 2012). Vos was then restricted to non-deferential use and evolved into vosotros for the plural, due to the combination of vos+otros ‘you+others,’ which also took the

2PL form of the verb, like vos. The new system looked as shown in Table 2 below at the beginning of the Golden Age, from the fifteenth to the seventeenth centuries.

Non-deferential Deferential Singular tú / vos vuestra merced Plural vosotros vuestras mercedes Table 3. Address forms in the Golden Age (Penny 2002: 138).

Then, during the Golden Age and the eighteenth century, only tú came to be used in Peninsular Spanish and in certain areas of Latin America in Peru, Bolivia, Mexico for non-deferential, informal use. In many areas of Latin America, however, vos and tú continue alternating nowadays as non-deferential singular pronouns, along with usted in some varieties. Vuestra merced underwent a series of contractions

(vuesarced, vucé, vested, etc.) until it was reduced to usted, and vuestras mercedes to ustedes. The modern system emerged in Castilian Spanish, as shown in Table 3.

Non-deferential Deferential Singular tú usted Plural vosotros ustedes

Table 4. Modern pronominal address system in Peninsular Spanish (Penny 2002:

138).

The generalization of the use of vuestra(s) merced(es) and the third person morphology of the verb for expressing deference may be partially due to the fact that vos verbal forms (2PL) ending in –des, as in cantedes ‘you guys sing (subjunctive),’ left the system during the fourteenth and fifteenth centuries. More precisely, the intervocalic d of-des was dropped starting at the end of the fourteenth century. This

12 resulted in verbal forms for the plural address form vos, e.g. amades> amás, that were different from the verbal forms for tú only in the position of the stress, e.g., tú amas, vos amás. For other forms, like conditionals, the verb would be the same for vos and for tú after dropping the intervocalic d of-des -des; for example, enseñaríades became enseñarías ‘you would teach;’ enseñarías would be used for the singular and the plural address forms, tú and vos, respectively; another example is cantabas ‘you used to sing,’ from cantávades, which, also after dropping the intervocalic d of -des, i.e. cantabas, would be used for both tú and vos (singular and plural).

Thus, the use of the third person verbal morphology would avoid the confusion.

More precisely, after dropping the intervocalic d of -des, the resulting hiatus could be resolved through assimilation in a monophthong, as in ades>aes>ás, or through dissimilation in a diphthong, as in ades>aes>áis. Since the sixteenth century, in

Peninsular Spanish, the dissimilated forms were used, e.g. cantáis ‘you guys sing,’ for the 2PL verb forms. In American voseo, i.e. the use of vos as a non-deferential singular address form, the assimilated forms were adopted mostly, e.g. cantás ‘you sing,’ although the dissimilated option still occurs with voseo too (Penny 2000). It was the assimilation solution what resulted in verbal forms that could create confusion between the verbal forms for tú and vos in certain tenses and moods, since the end of the fourteenth century (Fernández Martín 2012).

However, when vuestra merced was not yet consolidated in the pronominal system, syntactic agreement of vuestra merced was not always with third person forms but there were mixed patterns between vuestra merced>usted/vos. During the sixteenth and seventeenth centuries, there were mixed patterns in the use of vuestra(s) merced(es) and then usted(es) (singular and plural), in Peninsular Spanish. This

13 means that vuestra(s) merced(es), and then usted(es), was first used in agreement with

2PL forms before it adopted agreement with third person forms, as illustrated in the following example, taken from Fernández Martín (2012: 143), who analyzed different sources, including CORDE, literary sources, manuscripts, administrative documents:

“vuestra merced fuéredes (2PL verb form)” ‘your grace would go.’ In Latin American

Spanish, when vosotros was still used, and its use alternated with that of ustedes, since the sixteenth century until the first half of the nineteenth century, there were mixed patterns for vosotros and ustedes. Therefore, still in the nineteenth century, in colonial writings from some areas of Latin America, like Buenos Aires, Argentina, one finds examples of vosotros with the 3PL verb form and ustedes with the 2PL verb form, e.g. vosotros quieren, ustedes queréis ‘you (plural) want’ (Sánchez Méndez

2003). Cases of vuestras mercedes with the 2PL verb form are also found, as illustrated in the following example, taken from Fernández Martín (2012: 127): contemplad (2PL verb form) vuesas mercedes ‘look, your graces.’ Sánchez Méndez

(2003) argues that a distinction in formality between vuestras mercedes and vosotros was lost in Latin America since the beginning of Spanish in Latin America, and that mixed paradigms between the two can be observed since the sixteenth century. He further argues that those mixed paradigms could have been exported from Andalusia to Latin America. However, in Andalusia, these mixed paradigms have lasted until the present, whereas in Latin America they disappeared; more precisely, by the eighteenth century, in colonial writings from the majority of Latin America, only vuestras mercedes and the 3PL paradigm can be found, even though in some writings from the nineteenth century there were still some examples of mixed patterns, i.e. of vosotros with the 3PL verb form and ustedes with the 2PL verb form. Vosotros fell into disuse

14 in Latin America due, to a large extent, to the fact that vos and vosotros co-existed and shared the same verb paradigm, i.e. the 2PL form of the verb, the clitic os ‘you,’ the possessive vuestro ‘your.’ Whereas this was solved in Spain in favor of vosotros and vos fell into misuse, the opposite happened in Latin American Spanish, where vos was maintained and vosotros lost (Nieuwenhujsen 2003).

Similarly, in the Canary Islands, vosotros and the 2PL verb form were lost in favor of ustedes plus the 3PL verb form, which is now the form used to address a group of interlocutors in all contexts. However, in some rural areas of La Gomera, el

Hierro, and La Palma, Tenerife, some examples of archaizing vosotros plus the 2PL verb form and vos instead of os, as reflexive pronoun, are found, e.g. vosotras vos laváis ‘you guys wash yourselves.’ Furthermore, examples of ustedes plus the 2PL verb form are found in rural areas of Tenerife, more precisely, in the village of Arafo.

Similarly, in rural areas of Salamanca, in the region of La Ribera, until the first half of the twentieth century, there were examples of mixed patterns which included use of vos 2PL verb forms, corresponding to Old Spanish vos, with usted and with ustedes, and with él/ella ‘he’, ‘she’ e.g. ¿cantáis usted? ‘do you sing?’, ¿cantáis ustedes? ‘do you guys sing’, ¿cantáis él? ‘does he sing?’ In the Philippines too, cases of ustedes plus the 2PL verb forms were found until the first half of the twentieth century

(Fernández Martín 2012). In rural areas of Aragón, in a village in the Pyrenees, examples of mixed patterns are found, more precisely, of usted with the 2PL verb form, as illustrated in the following example taken from Fernández Martín: usted no le llegasteis a conocer ‘you didn’t get to meet him’ (2012: 193). Also in Aragón, some cases of ustedes and vosotros yuxtaposed, with ustedes always placed first, have been found, e.g. ustedes vosotros no lo conocéis (Fernández Martín 2012: 193).

Based on the presence of mixed patterns in the history of Spanish, both in the singular and plural, Fernández Martín (2012) argues that the current use of ustedes plus the 2PL verb form is an archaism. More specifically, the use of ustedes/vuestras mercedes plus the 2PL verb form instead of the 3PL verb form has been present in both Peninsular Spanish and Latin American Spanish since the sixteenth century.

Similarly, Bustos Tovar (2001) argues that ustedes plus the 2PL verb form is a grammatical archaism kept alive since the sixteenth century. Furthermore, vuestra(s) merced(es) (singular and plural) appeared before the verb more frequently than after the verb when it was not yet fully established as an address form and the system was in a process of change (Fernández Martín 2012). Due to the placement of vuestra(s) merced(es) before the verb and its initial agreement with 2PL instead of 3PL verb forms, vuestra(s) merced(es) is analyzed in the literature as an apposition where the real subject was vos, with which vuestra(s) merced(es) co-occurred at first. This is illustrated in an example taken from Fernández Martín (2012: 131): ¿vuestra merced, vos cantáis? ‘do you sing, your grace?’ Later on, vuestra(s) merced(es) was reanalyzed as the subject pronoun and vos was dropped (Fernández Martín 2012).

Similarly, Lara Bermejo argues that the mixed pattern nowadays, i.e. the use of ustedes with 2PL verb forms, is explained by considering ustedes as topicalized, where vosotros is the real, non-explicit, “silent” subject (2015: 461), e.g. ustedes

(silent vosotros) queréis ‘you (plural) want.’ As was the case with vuestra(s) merced(es), ustedes is therefore possibly being reanalyzed as the subject in a grammaticalization process (Lara Bermejo 2015). This is also suggested by the occurrence of examples of ustedes and vosotros yuxtaposed in some areas, i.e. rural

16 areas of Aragón, with ustedes always placed in front of vosotros, e.g. ustedes vosotros queréis.

Moreno de Alba (2010) argues that vosotros fell into disuse in Latin American

Spanish, and ustedes expanded in Spain from 2% to 52% use during the eighteenth century for formal contexts, as shown by data collected by Alba Moreno from Corpus diacrónico del español (CORDE). Then, during the 20th century, vosotros spread in

Spain, except for Western and Central Andalusia, after democracy arrived post-

Franco, which favored more solidarity in interpersonal relationships (Fernández

Martín 2012).

Previous findings and observations regarding the development of the plural address forms in Spanish, namely vuestras mercedes>ustedes and vosotros, guide my query into the current status of the variable use of plural address forms in Andalusian

Spanish as a change in progress. I hypothesize that not only is ustedes in combination with the 2PL form an archaic construction, more precisely, an old construction from the sixteenth century, but is also an obsolescing linguistic element, which is an old linguistic element falling into disuse (cf. Rosemeyer and Schwenter 2017). I base this hypothesis on the presence of mixed patterns in the past in Peninsular Spanish and

Latin American Spanish, i.e. the use of vuestras mercedes and ustedes in agreement with 2PL forms. More precisely, I base this hypothesis on the fact that mixed patterns involving 2PL verb forms and vosotros/ustedes have been common when the address form systems were in a process of change; then, the mixed pattern was eventually lost in some cases, when the change was completed. That is the case of Latin American

Spanish, where, eventually, the 2PL verb form and vosotros, and the mixed patterns of ustedes plus 2PL or vosotros plus 3PL were lost in favor of ustedes plus the 3PL

17 verb form, which is now the only form to address a group of interlocutors in Latin

American Spanish. I also base my hypothesis on the fact that, in other cases, mixed patterns with 2PL verb forms, i.e. ustedes plus 2PL verb forms, are kept as obsolescing linguistic elements, with limited use, limited to certain rural areas, in regions where a change of the address form system has been completed. That is the case of some examples found of ustedes plus 2PL verb form in rural areas of

Salamanca, where, on the other hand, vosotros plus the 2PL verb form is the form generally used to address a group of interlocutors. It is also the case of the Canary

Islands, where the plural address form chiefly used for a group of interlocutors is ustedes plus the 3PL verb form, while vosotros and the 2PL verb form fell into disuse.

In this system, however, cases of vos as reflexive pronoun, and of the subject pronouns vosotros and ustedes plus 2PL verb forms occur only in certain limited rural areas, thus suggesting that they are obsolescing linguistic elements (Fernández Martín

2012).

2.2. Current 2PL Address Form System in Peninsular Spanish

Standard grammatical descriptions of the current Castilian pronominal system include two second person singular forms, one deferential, for distancing and formality, usted, and one non-deferential, tú ‘you (singular).’ It also includes two plural forms, one deferential, for distancing and formality, ustedes, and one non- deferential, vosotros ‘you (plural),’ as illustrated in Table 3, Section 2.1. These descriptions make the address form system look like it is symmetrical, with one singular non-deferential address form and its deferential singular counterpart, and one

18 plural non-deferential address form and its deferential plural counterpart (Morgan and

Schwenter 2016). However, the system is actually not symmetrical. Morgan and

Schwenter (2016) analyzed the Castilian system of pronominal address. They constructed an online survey in which, first, participants had to self-report on their frequency of use of usted and ustedes. Morgan and Schwenter found that usted was reported by participants to be used more frequently than ustedes in situations where a deferential pronoun was needed, more precisely, for a single addressee in the case of usted, and for a group of addressees in the case of ustedes. This means that usted, as a singular address form, is used for deferential purposes more frequently than ustedes, which is frequently replaced by vosotros for referring to a group of addressees with deferential purposes. Next, participants had to choose a 2PL pronoun for different real-life scenarios, i.e. vosotros or ustedes, in situations where usted had been imposed upon the participants as the singular address form, by the wording of the scenarios. Moreover, in one of the scenarios, participants had to choose one second person singular pronoun, i.e. usted or tú, as well as vosotros or ustedes. Morgan and

Schwenter found that even if participants use the singular deferential form usted for an interlocutor, they may use the plural non-deferential form vosotros for a group of interlocutors including the interlocutor previously addressed by usted. This means that vosotros often serves as the plural for both tú and usted. Therefore, vosotros is in fact currently the most productive 2PL address form for speakers of Peninsular Spanish

(Morgan and Schwenter 2016). The literature suggests that vosotros is replacing ustedes in contexts where ustedes had previously replaced vosotros, i.e. formal contexts (Fernández Martín 2012; Morgan and Schwenter 2016). In Central and

Western Andalusia, as was explained in section 1.1., ustedes alternates with vosotros

19 as the plural form of tú, i.e. in informal contexts. This means that the address form system in Central and Western Andalusia is not symmetrical either. On the other hand, in other dialects, ustedes and vosotros never alternate in informal contexts

(Alvar 1973; Fernández Martín 2012; Narbona et al. 1998). Below, in chronological order, I focus on previous work on the current 2PL address form system in Central and Western Andalusia. This previous work informs the hypotheses of this dissertation. This dissertation, however, goes beyond the limitations of these same previous studies, which I discuss in section 2.2.1.

In a previous study, I analyzed the second person plural variation in Central and

Western Andalusia. I conducted interviews about perceptions about the use of ustedes plus the 2PL form of the verb expressed by Andalusian speakers. Results indicated that speakers from Central Andalusia considered the phenomenon stigmatized and typical of uneducated people and rural areas. On the other hand, speakers from

Western Andalusia didn’t perceive the use of ustedes plus the 2PL form of the verb as stigmatized (Jaime Jiménez 2011). In another previous study, I analyzed both the 2PL pronominal and verbal alternation in Central and Western Andalusia, by analyzing data from online fora (Jaime Jiménez 2014). Based on the existence of mixed patterns of vuestras mercedes and ustedes in the past, as argued by Fernández Martín (2012), I explored the status of ustedes plus the 2PL verb form as an obsolescing linguistic element. To test this hypothesis, I analyzed frequency effects on the 2PL variation in

Andalusian Spanish, based on the findings and observations made by Erker and Guy

(2012) about frequency. Erker and Guy state that “high frequency is related with an archaic rate of use, defying a change in progress,” and with “conservation of older patterns” (2012: 531). Erker and Guy researched the effects of frequent verb forms on

20 variable subject pronoun expression in Spanish, more precisely, in relation to null versus overt subject expression. Erker and Guy considered local frequency, which means that a threshold of 1% of their own corpus was the cut-off for considering a verb form as frequent. They found that frequency has no independent effect but instead interacts with other factors. Frequency magnifies the effect of certain factors which, without frequent forms, do not emerge as significant. Erker and Guy offer an independent explanation of the effect of frequency by arguing that the magnifying effect of frequency reflects the linguistic experience of the speakers, which means that

“speakers cannot formulate hypotheses about individual lexical items until they have sufficient evidence” (2012: 526). Following Erker and Guy (2012), I analyzed the effects of local frequency on the 2PL pronominal and verbal variation in Andalusian

Spanish. I found that pre-verbal placement, following the pattern of vuestras mercedes, leads to significantly more use of ustedes over vosotros with the 2PL verb form but only with frequent verbs. I argued that this suggests that ustedes plus the

2PL verb form represents an obsolescing linguistic element that is kept alive only through the effect of frequent verbs. Regarding the verbal alternation, I found that the

3PL verb form with ustedes is more likely to occur when the verb is in the imperative mood and with stative and mental verbs (Jaime Jiménez 2014).

Lara Bermejo (2015) also analyzed the 2PL pronominal and verbal system in

Andalusian Spanish. In order to do that, he collected data from the ALPI (Atlas

Lingüístico de la Península Ibérica), from the corpus CORDIAL-SIN, as well as from oral interviews he conducted in Andalusia. For the interviews, Lara Bermejo used what he called dubbing of TV sitcoms, to elicit the production of 2PL address forms.

The TV sitcoms used were the Spanish series Aquí no hay quien viva ‘No one could

21 live here’ and the American series Friends. Lara Bermejo showed the videos to participants and presented them with contextualizing sentences as a stimulus. Then, participants had to dub a scene, with the volume turned off, where a character addresses a group of interlocutors. The participants had to dub the character addressing the group of interlocutors. On the other hand, older speakers were presented with daily life scenarios and not with the TV sitcoms, because they were less motivated by that type of stimuli. In both tasks, the participants had to address a group of characters presented in the videos or in the scenarios, so that participants had to use the 2PL or 3PL pronominal and verbal forms. Different symmetrical and asymmetrical interpersonal relationships were present in the scenarios and the TV scenes. Therefore, the characters of the scenarios and TV scenes were imaginary family, friends, and unknown people the interviewee had to address. Using logistic regression and chi squares, Lara Bermejo analyzed the effect of several social factors including sex, age, gender, education, province of origin, area from the ALPI atlas, and population size. Regarding linguistic factors, Lara Bermejo included stressed pronoun

(subject and object of preposition), unstressed pronoun (clitics), possessives, tense, mood, sentence type (declarative, interrogative, negative, affirmative, imperative), communicative situation (symmetrical vs. asymmetrical interpersonal relationships), subordinate verb tense, type of subordinate clause, and 2PL pronominal morphology

(clitics os, se, sus). Lara Bermejo then completed a descriptive statistical analysis of the linguistic variables, but no inferential statistics like with the social factors. He argued that the use of ustedes and the 2PL or the 3PL is more widespread in the

Western Andalusian provinces of Seville and Cádiz than in the Western Andalusian province of Huelva and the Central Andalusian provinces of Málaga and Córdoba.

Lara Bermejo further claims that the use of ustedes plus the 2PL or the 3PL in informal contexts originated in Seville and Cádiz, and then spread toward the other

Andalusian provinces where it is less common, i.e. Huelva, Málaga, and Córdoba.

Lara Bermejo (2015) considers that ustedes plus the 2PL or the 3PL, is innovative in informal contexts in Central and Western Andalusia. More precisely, Lara Bermejo argues that 3PL morphology spreads following an implicational hierarchy and that this spread of 3PL forms is more advanced in Cádiz and Seville. This suggests a change in progress where the verbal and pronominal 3PL forms are the innovation, progressively spreading. Lara Bermejo bases his arguments on Blake’s implicational hierarchy related to Latin cases and markedness (2004). Blake argues that Latin cases show implicational relationships with each other according to the following hierarchy: subject>direct object>indirect object>oblique case. As an example, the unmarked order of words in many languages follows the order of this hierarchy from left to right, where the most marked case is the oblique. According to Lara Bermejo, the innovative 3PL morphology in Andalusian Spanish follows a similar implication hierarchy, which is illustrated in Figure 2.

Stressed pronouns> reflexives>verbs> accusative clitics>dative clitics>possessives

Figure 2. Implicational hierarchy followed by the progress of the 3PL morphology.

Thus, according to Lara Bermejo, 3PL morphology in Andalusian Spanish is first adopted by the linguistic element on the top of the implicational hierarchy because the higher an element is in the hierarchy, the easier the access the speaker has to it and the

23 less marked the element is. The linguistic element at the top of the hierarchy is the stressed pronoun, i.e. ustedes, which functions as a subject, e.g. ustedes coméis ‘you guys eat’ or as an object of a preposition, e.g. para ustedes ‘for you guys.’ For example, in vosotros os coméis el pollo ‘you guys eat the chicken,’ the first linguistic element adopting the 3PL form would be the subject, which results in ustedes os coméis el pollo ‘you guys eat the chicken’. Then, the 3PL form spreads down in the hierarchy so that it is adopted by the reflexive clitic, i.e., se. Therefore, the previous example would become ustedes se coméis el pollo ‘you guys eat the chicken.’ In this example, both the subject pronoun and the reflexive clitic have adopted the 3PL morphology. Then, the verb adopts the 3PL morphology, so that the result would be ustedes se comen el pollo ‘you guys eat the chicken.’ In this example, the subject pronoun, the reflexive clitic and the verb have adopted the 3PL morphology. Then, the 3PL form is adopted by accusative clitics, followed by dative clitics, e.g., los, les instead of the 2PL clitic os, in a ustedes los vi el otro día ‘I saw you guys the other day’, and les doy esto a ustedes ‘I give you guys this.’ Finally, the 3PL morphology is adopted by the possessives, the most marked case, so that 3PL possessives are used, e.g. su instead of the 2PL possessive vuestro ‘your’ (Lara Bermejo 2015).

Furthermore, according to Lara Bermejo, if an element at a given point in the hierarchy has adopted the 3PL morphology, e.g., dative clitics, then elements located higher in the hierarchy will have necessarily adopted the 3PL morphology too, e.g. the verb. Lara Bermejo finds 3PL morphology expansion to follow this hierarchy, which would thus suggest that 3PL forms are in fact the innovation. Moreover, Lara

Bermejo discusses the case of Argentinian voseo. In Argentinian Spanish, voseo morphology was first adopted also by the stressed pronoun, i.e. vos ‘you (singular).’

The rest of the linguistic elements, including verb, clitics and possessives, at first did not have voseo morphology but rather tuteo morphology in Argentinian Spanish, e.g. comes (second person singular tú verb form, ‘you eat’), te (tuteo clitic, ‘to you’), tu

(tuteo possessive, ‘your’). Later on, voseo morphology emerged on the verb in the imperative mood, in commands, e.g. ¡cantá! ‘sing!’, and then emerged on verbs in the present tense and in the subjunctive. Clitics and possessives, as well as the rest of verb tenses, still have tuteo rather than voseo morphology (Lara Bermejo 2015). Lara

Bermejo argues that the 3PL morphology in Andalusia is following a similar pattern to voseo in Argentina. According to Lara Bermejo, like in Argentinian Spanish, the innovative morphology in Andalusian Spanish, i.e. the 3PL morphology, has been adopted first with the stressed pronoun, i.e. ustedes, functioning as a subject, e.g. ustedes coméis ‘you (plural) eat’ or as an object of a preposition, e.g. para ustedes

‘for you (plural).’ Similarly, just as the voseo verb forms were used at first only in the imperative mood, the combination of the 3PL verb form with ustedes in informal contexts in Andalusian Spanish is occurring more frequently in the imperative mood, as found by in a previous research of mine (Jaime Jiménez 2014) and by Lara

Bermejo (2015). Alternation of verb forms in cases of mixed paradigms is in fact found most frequently in the imperative mood than in other moods, as is the case of

2PL/3PL verb form alternation with the subject pronoun ustedes in Central and

Western Andalusia, and the case of voseo and tuteo verb forms with the subject pronoun vos in Argentinean voseo (Fernández Martín 2012).

In addition, beyond the hierarchical adoption of 3PL morphology, Lara Bermejo found that there are no pragmatic differences between ustedes and vosotros or between the 2PL and 3PL verb forms. More precisely, he included in his analysis

25 types of situations, namely, asymmetric and symmetric power relationships, and didn’t find a significant difference between symmetric and asymmetric contexts for the choice of ustedes or vosotros, or of 2PL or 3PL verb form or clitics, i.e. se, le, les, lo/la, los/las (3PL), versus os (2PL). Similarly, in other previous studies, where the phenomenon is briefly described, it is mentioned that the variants ustedes/vosotros and 2PL/3PL verb forms occur in the same pragmatic contexts in terms of

(a)symmetric power relationships. More precisely, previous studies briefly mention that all the variants occur in the same informal contexts, with symmetry of interpersonal relationships (Alvar 1973; Fernández Martín 2012; Jiménez Fernández

1999).

2.2.1. Implications for this Dissertation

Previous work suggests that ustedes with the 2PL verb form is an obsolescing linguistic element. More precisely, this is suggested by the fact that ustedes plus the

2PL occurs more frequently in pre-verbal position, as was the case of vuestras mercedes at first, when the system was in a process of change, toward, eventually, vuestras mercedes>ustedes plus the 3PL verb form. Furthermore, the pre-verbal placement of ustedes plus the 2PL verb form is occurring with frequent verbs. That ustedes plus the 2PL is an obsolescing linguistic element is also suggested by the existence of mixed patterns in the past, when both vuestras mercedes and ustedes occurred in agreement with 2PL forms, when the system was in a process of change, in both Peninsular and Latin American Spanish. In the latter, the mixed pattern involving the 2PL verb form fell into disuse and the plural address system adopted,

26 eventually, only ustedes plus the 3PL verb form to address a group of interlocutors.

That ustedes with the 2PL verb form is an obsolescing linguistic element is also suggested by the current presence of mixed patterns, e.g. usted and/or ustedes plus

2PL verb forms, maintained in the present as archaizing linguistic elements, limited to some specific rural areas of regions where vosotros plus the 2PL verb form, or ustedes plus the 3PL verb form has been adopted as the plural address form, like in the

Canary Islands. That ustedes plus the 2PL verb form is an obsolescing linguistic element is also suggested by the fact that the 3PL morphology in Andalusia, i.e. ustedes and the 3PL verb form, is following a similar pattern to voseo in Argentina, first occurring more frequently in the imperative mood, indicating a possible change in progress toward 3PL morphology (Jaime Jiménez 2014; Fernández Martín 2012;

Lara Bermejo 2015). That ustedes plus the 2PL verb form is an obsolescing linguistic element is also suggested by the fact that in a previous study I found that the rate of vosotros with 2PL verb forms is higher than the rate of ustedes plus 2PL in Central and Western Andalusia. More precisely, out of 2,004 tokens extracted from online fora, Twitter, and PRESEEA Málaga, tokens of the variant ustedes comprised 38.3% of all the tokens, and tokens of the variant vosotros comprised 61. 7% of all the tokens

(Jaime Jiménez 2016).

Despite these findings and observations, previous work on plural address forms in

Andalusian Spanish does not include a systematic variationist study of the phenomenon. Lara Bermejo (2015) analyzed what factors affect the use of ustedes/vosotros; however, he conducted his analysis without regard for verb form variation and therefore could not determine whether ustedes is being used in combination with the 2PL or the 3PL verb form. Moreover, he assumes that all

27 infinitives used as commands correspond to the 3PL verb form. Counting instances of infinitives as instances of the 3PL form of the verb is called into question in the present dissertation. Lara Bermejo bases his assumption on the fact that the infinitive has no person agreement but is expressed by default with the third person clitic se.

However, infinitives cannot be counted as 3PL verb forms since occurrences of vosotros with 3PL form of the verb are in fact very rare, and it is also the case that infinitives used as commands occur with vosotros. Infinitives, including infinitives with enclisis of se, can actually co-occur both with ustedes and vosotros. As

Fernández Martín (2012) argues, sentarse ‘you guys sit down’ is the preferred form for a command in Andalusia, both for ustedes and vosotros, in all Andalusian provinces. An example with an infinitive with enclisis of se appears associated with vosotros in (6), an example taken from the oral elicitation task I conducted in

Andalusia for this dissertation. The example was produced by a participant from

Málaga, Central Andalusia.

(6) Vosotros mirarse al espejo que tenéis por donde callar más que los demás.

‘You guys [vosotros] look [infinitive] yourselves [clitic se] at the mirror; you

can’t show off more than others.’

Moreover, Lara Bermejo proposed that the progress in the use of the 3PL morphology follows the implicational hierarchy illustrated in Figure 2 above. In such hierarchy, the first element is the stressed pronoun, then reflexives, then the verb.

However, in a previous study I found that 3PL reflexives are rare, and less frequent overall than 3PL verb forms (Jaime Jiménez 2016). My results showed that there was

89% os, versus only 11% se from a total of 63 tokens with presence of a reflexive clitic with ustedes and the 2PL form of the verb (63 out of 2,004 for the pronominal

28 alternation and 1,049 for the verbal alternation). On the other hand, for the verbal alternation there were 1,049 tokens, with 24.5% of the 3PL form of the verb and

75.5% of the 2PL verb form. All se tokens came from Western Andalusia only, whereas tokens with the 3PL verb form are found in Central Andalusia as well (Jaime

Jiménez 2016). Therefore, the most common form, overwhelmingly, is ustedes with the clitic os, e.g. ustedes os laváis ‘you guys wash yourselves,’ as illustrated in Figure

3 below. This suggests that the use of the clitic se with ustedes and the 2PL verb form is actually a non-productive variant, and possibly an obsolescing linguistic element, as suggested by the findings of Bustos Tovar (2001), who argued that se with ustedes and the 2PL verb form is stigmatized, and it is an archaism, associated with rural areas, and low education. This calls into question that the progress in the use of the

3PL morphology follows the implicational hierarchy illustrated in Figure 2 as argued by Lara Bermejo.

100% 90% 80% 70% 60% 50% Total % of clitic 40% 30% 20% 10% 0% se os

Figure 3. se/os alternation with ustedes plus the 2PL form of the verb.

In addition, Lara Bermejo completed a descriptive statistical analysis but no inferential statistical analysis of the linguistic factors, so he was unable to determine the hierarchical ordering of the linguistic factors affecting the variation in terms of their statistical significance.

In order to overcome these limitations of previous studies, in this dissertation I analyze pronominal choice considering ustedes/vosotros as subjects of the 2PL verb form, and infinitives are not counted as instances of the 3PL verb form, but instead as a separate verb category, as an independent variable. Furthermore, in the current study, I determine the hierarchical ordering of the linguistic factors in terms of their statistical significance. Building off previous findings, I explore the status of the system of plural address forms in Andalusian Spanish as a change in progress where ustedes plus the 2PL verb form is an obsolescing linguistic element and ustedes with the 3PL verb form is the innovation. In order to do that, I examine the conserving effects of priming, frequency, and clause type, since these have been identified as conditioning obsolescing linguistic elements (Bybee and Thompson 1997; Bybee

2001; Rosemeyer and Schwenter 2017; Schwenter 2015). Therefore, I discuss previous work on frequency, priming, and clause type, in that order, below, in section

2.3., and then, in section 2.3.4., I discuss how their conclusions related to changes in progress apply to the hypotheses of this dissertation.

2.3. The Role of Frequency, Priming, and Clause Type in Changes in Progress

2.3.1. Frequency and Entrenchment Effects

There are two types of frequency effects, a reduction effect and a conservation effect. In the first type, the reductive change, words with high token frequency are affected before low-frequency words (Bybee and Thompson 1997). Using examples from sounds, deletion of Spanish intervocalic /d/ is higher in high token frequency words. In American English, loss of syllabicity in post-stress schwa is more advanced in high token frequency words such as family, camera, memory (Bybee and

Thompson 1997). On the other hand, in the second type of frequency effect, which has a conserving effect, analogical leveling or regularization affects low-frequency words before high frequency words (Bybee and Thompson 1997; Bybee 2002). For example, relatively low-frequency verbs in English such as weep/wept, creep/crept are regularizing to weeped, creeped in present-day English; however, high-frequency verbs such as keep/kept, sleep/slept, are not regularizing. Exceptionally high- frequency words or phrases can be acquired without following the general patterns of the language, due to their high frequency in use, high availability in the input and thus stronger lexical representation (Bybee 2002). The more a form is used, the stronger its lexical representation is, which makes the form easier and faster to be accessed

(Bybee and Thompson 1997). Therefore, more frequent words are more resistant to morphophonemic change in the paradigm, keep their morphophonemic irregularity, and are not replaced by new forms, since they are more autonomous forms (Bybee and Brewer 1980). Old constructions thus resist analogical leveling, which first affects low frequency words. Low-frequency words are less frequently practiced, less

31 present in experience, so that they are more subject to the strong patterns of the language due to their weaker representations (Bybee 2002).

Conservation of old patterns also affects the syntactic level. For example, English pronouns still maintain case distinction, i.e. for nominative versus accusative/dative cases, while case distinctions are not maintained in nouns. This is because pronouns are much more frequent than noun phrases and thus resist change (Bybee and

Thompson 1997). Another example is that of the Canadian French subjunctive, which is an old structure preserved by high frequency verbs (Poplack 1992). More precisely, in Canadian French, the subjunctive has been lost, it is no longer productive, and there is no functional or meaningful difference between indicative and subjunctive.

However, residues of the subjunctive remain with high token frequency verb forms and in the most frequent syntactic contexts, in entrenched phrases with particular matrix verbs, and with particular irregular embedded verbs. Therefore, the use of the subjunctive in Canadian French is limited to specific routinized constructions with specific lexical items which are accessed in production, such as il faut que + subjunctive of faire/aller/avoir/être (‘it is necessary that + subjunctive of do/go/have/be’). The subjunctive is therefore preserved in these entrenched sequences which behave more like lexical items than hierarchical syntactic structures. The subjunctive is therefore a “lexically-arbitrary residue of formerly productive patterns”

(Bybee and Thompson 1997: 384).

Erker and Guy (2012) studied the effects of lexical frequency on variable Spanish subject expression, i.e. the variable use of overt and null subjects. They argue that high frequent verb forms should lead to more use of null subject pronoun expression.

This is because lower rates of overt subject pronoun expression are archaic, and the

32 general trend is toward more overt subject pronoun expression in general in Spanish.

Erker and Guy analyzed frequency as a categorical binary variable, i.e. a verb form was classified either as frequent or as infrequent. Furthermore, following a usage- based approach, they defined frequency locally, which means that verbs that constituted 1% or more of their corpus were classified as frequent verbs, and verbs which constituted less than 1% of their corpus were considered infrequent. Erker and

Guy found that frequency does not have an independent effect but interacts with the other conditioning factors in their study. Erker and Guy compared the effects of the different factors on subsets of the data, namely, on frequent verb forms and on infrequent verb forms and found that the effects were not consistent in both groups.

For example, they found that frequent verb forms lead to lower rates of overt pronouns when they are also irregular verb forms. On the other hand, they found that mental activity verbs lead to more overt pronoun usage when they are also frequent and regular, but, when considering frequent and infrequent forms separately, the effect of the semantic content of verbs is significant only with frequent forms. They also found that the predictive power of tense-mood-aspect was stronger with frequent forms than with infrequent verb forms. Therefore, the effect of factors conditioning subject pronoun expression does not operate across the whole lexicon, but only with frequent forms. Frequency thus has a potentiating or amplifying effect, i.e. factors that didn’t have an effect with infrequent forms have a predictive effect with frequent forms only; and factors with a weak predictive effect with infrequent forms have a stronger effect on subject expression with frequent verb forms only.

2.3.1.1. Frequency and the Autonomy Effect

There is a third effect of frequency, along with the reduction effect and the conserving effect, which is the autonomy effect, related to the conserving effect of frequency (Bybee 2006). Bybee and Brewer (1980) argue that high frequent sequences, the ones that have become entrenched in their structure and are not likely to follow the most productive patterns of the language, have separate lexical entries, i.e. they are stored in the lexicon as a whole and independent unit. This means that highly frequent sequences are non-segmented, autonomous, and they are prefabricated word combinations, whereas less frequent words maintain the identity of their parts and are not autonomous (Bybee 2006). Autonomous sequences can be learned unanalyzed and unsegmented (Bybee and Brewer 1980). Erker and Guy (2012) also discuss this autonomy effect, and consider the possibility of accessing word sequences or phrases as units with a single meaning, which implies that speakers possess knowledge of collocational information. For instance, idioms are an extreme case of highly frequent co-occurring elements. Therefore, “particular high-frequency pairings of pronoun with verb are treated as ‘chunks’ and conserved in that relationship”

(Erker and Guy 2012: 531). Speakers may store such collocations of overt/null subject pronoun expression plus verb form, since high frequency provides them with enough evidence about the context of occurrence and the probable collocation (Erker & Guy

2012). This means that a frequent verb form that commonly occurs with a null pronoun is entrenched in a collocation with the null pronoun; on the other hand, a frequent verb that commonly occurs with an overt subject is equally entrenched in a collocation with the overt subject (Erker & Guy 2012).

In the case of the subjunctive in Canadian French, if there is intervening material between the main verb and the subordinate verb, the indicative is more likely to occur.

On the other hand, if there is no intervening material, the subjunctive is more likely to occur. This is because, as mentioned before, the subjunctive is limited to entrenched, routinized sequences of the type il faut que (‘it is necessary that) + embedded irregular frequent verb in the subjunctive. Therefore, intervening material leads to more use of the productive variant, which is the indicative, whereas the subjunctive is limited to automated sequences (Bybee and Thompson 1997).

2.3.2. Priming

The studies of frequency, showing a conserving effect in changes in progress, follow usage-base approaches to linguistics, which assume that changes are dependent on cognitive processes based on usage. In the case of frequency, the cognitive process is entrenchment, resulting from repetition (Rosemeyer and Schwenter 2017). Another cognitive process with conserving effects in changes in progress is morphosyntactic persistence or priming, as argued by Rosemeyer and Schwenter (2017). The term priming is typically associated with psycholinguistics while corpus linguistics prefers the term persistence, but priming is also used frequently by variationists (Schwenter

2015). Therefore, the term priming is used in the present dissertation as a synonym of persistence. Priming is defined as the tendency to use the same variant recently used in the discourse (Schwenter 2015). If a speaker faces a linguistic variable with the possibility of choosing between different variants with the same meaning, he/she will be influenced by the exposure to one of the variants present in previous discourse,

35 produced by himself/herself or another speaker; more precisely, he/she will be more likely to use the most recently-produced variant, all other things being equal. For instance, in the variable use of passive and active sentences, it was found that speakers would produce more readily the structure, i.e. passive or active, previously presented in discourse as a prime (Szmrecsanyi 2005). Szmrecsanyi argues that

“persistence plays an important role in language use. For one thing, there is a sizable body of psycholinguistic, experimental research demonstrating that language users are hard-wired to go for recently used (or activated) linguistic patterns whenever they can;” “persistence thus influences the speaker’s choice between linguistic elements that compete within a certain envelope of variation” (Szmrecsanyi 2005: 114).

Rosemeyer and Schwenter (2017) also argue that priming has a role in the conservation of the old variant of a morphosyntactic variable. Because priming of a linguistic element activates the linguistic element in question, it also strengthens its mental representation, which makes it easier to be accessed. This is related to the significance of repetitiveness in discourse, which is also related to the role of frequency in conserving old patterns (Szmrecsanyi 2005). Priming, like frequency, results in the conservation of the obsolescing element in processes of change.

Furthermore, less frequent linguistic elements prime more strongly, so that priming effects are inversely correlated to overall frequency of usage of the linguistic element in question. This is because marked forms are more unexpected, uncommon, whereas the unmarked forms are the most productive ones, the default forms (Bybee and

Brewer 2980; Thorsos 2002). Therefore, priming leads to more use of the less frequent and more marked variant, which is the obsolescing variant. The obsolescing variant is falling into disuse and thus is less frequent. This variant, by virtue of being

36 less common and less expected, creates a greater surprisal effect when being used, leading to higher cognitive activation. Therefore, the less probable a variant is, the stronger the priming effects. Thus, when infrequent forms are used, priming is stronger and helps to prevent obsolete, infrequent variants from leaving the system

(Jaeger and Snider 2008; Rosemeyer 2015; Schwenter 2015).

Rosemeyer and Schwenter (2017) analyzed the alternation between the past subjunctive endings -ra and -se in Spanish, e.g. cantara and cantase ‘had sung.’ The older form is -se and it has been progressively replaced by -ra, which has higher frequency of occurrence overall. Rosemeyer and Schwenter found that -se tokens correspond to a few highly frequent verbs, so that -ra tokens have a higher type/token ratio than -se tokens. Furthermore, they found that priming was relevant for both -ra and -se but much stronger for -se; more precisely, the probability of a prior -se in discourse leading to a subsequent -se form is higher than the probability of a prior -ra leading to a subsequent -ra case. They also found that the paradigm of -se is more limited than that of -ra, so that, even though both are more common with the third person singular forms of the verb, this association is stronger for -se forms, which is the less productive, less frequent form. They conclude that obsolescing linguistic variants prime more strongly than productive variants, so that priming conditions obsolescing, less frequent linguistic elements in language change.

2.3.3. Main and Subordinate Clauses

Bybee (2001) argues that innovations occur first in main clauses but older patterns are maintained longer in subordinate clauses when there is an ongoing change. While

37 main clauses are better controlled by the speaker, subordinate clauses are more difficult to process and less likely to be subject to incipient changes. Subordinate clauses are thus processed in large chunks, like irregular verbs, as an automated sequence, and their constituents are less independent and not so likely to change

(Bybee 2001). Furthermore, whereas topicalization, contrast, and presentative focus are found in main clauses, subordinate clauses contain old information, backgrounded information, low focus, and complementary information to the information shared earlier in the main clause. Therefore, there are more complex pragmatic relations and content in main clauses, while subordinate clauses are pragmatically flat. This helps to explain why changes and newly grammaticizing constructions occur typically in the main clause first. For example, in Latin and Old Spanish, the verbal -ra ending had a past anterior meaning, e.g. pudiera meaning ‘had been able.’ However, then, new forms appeared for perfect tenses, with the auxiliaries haber and ser plus past participle. The -ra forms were restricted to subordinate clauses, dependent on a main verb, and with a past subjunctive meaning. This is because subordinate clauses are associated with low focus, which is defined as related to negation, description of states, and relative clauses. On the other hand, the new periphrastic past anterior appeared in main clauses because it was associated with contexts of high focus, i.e. those which describe transitive events and have animate subjects and objects (Bybee

2001). Similarly, in the case of Canadian French, not only is the subjunctive, as mentioned before, associated with high frequent irregular verbs and with no intervening material between main and subordinate verb, and in entrenched sequences with particular verbs, but also it occurs in the embedded, subordinate clause.

2.3.4. Implications for this Dissertation

Based on the previous discussion of the effects of frequency, priming, and clause type in processes of change, in the present dissertation, I hypothesize that ustedes plus the 2PL verb form is more likely to occur with frequent verb forms, as I hypothesize that it is the obsolescing linguistic construction. Furthermore, I further hypothesize that speakers possess collocational information about ustedes plus the 2PL verb form at some level. More precisely, I argue that ustedes plus the 2PL verb form is accessed as a single unit. This means that, as a variant of low-productivity, it is accessed as a non-segmented word sequence, although not as a construction, where there is only one slot to be filled with little to no variation regarding the lexical elements that can fill it (Posio 2015). Ustedes plus the 2PL verb form will be more likely to occur with no linguistic elements between the subject ustedes and verb, but with different 2PL verbs that can fill the 2PL verb form slot.

I also hypothesize that ustedes plus the 2PL verb form is favored by priming, as the obsolescing linguistic variant. When ustedes is used, this should lead to more surprisal because it is the less expected variant, less common variant, with the rate of vosotros being double the rate of ustedes (Jaime Jiménez 2016). Surprisal makes ustedes more accessible in the ongoing discourse. The effect of clause type is also analyzed in the present dissertation, where it is hypothesized that ustedes plus the 2PL verb form occurs more frequently in subordinate clauses, since ustedes plus the 2PL verb form is hypothesized to be the obsolescing sequence.

2.4. Pragmatic Factors

Previous work on 2PL address forms in Andalusian Spanish argues that there are no pragmatic differences between ustedes/vosotros and 2PL/3PL verb forms in

Central and Western Andalusia (Alvar 1973; Fernández Martín 2012; Jiménez

Fernández 1999; Lara Bermejo 2015). However, research on Spanish historical second person address forms and research on current Spanish second person singular address form do find pragmatic differences between the various address forms that alternate in use. This is because they look at pragmatic factors going beyond the traditional ones of (in)formality or (a)symmetry of interpersonal relationships. I discuss these studies below, in section 2.4.1. and Section 2.4.2., in chronological order, as well as what the implications for my dissertation are, in section 2.4.3.

Furthermore, due to the nature of this dissertation, for which a large amount of the data analyzed was gathered from online sources, an additional section, 2.4.4., on pragmatic uses of online graphic features like emojis, is included.

2.4.1. Pragmatic Effects on Historical Address Forms

In analyses of the historical evolution of second person address forms in Spanish, pragmatic differences are found between the forms. Bentivoglio (2003) analyzed letters written by Andalusian immigrants in Latin America for their relatives in Spain, during the sixteenth century. She argues that vuestra merced was the most formal form of address in the sixteenth century in Peninsular Spanish, and tú the less formal, whereas vos occupied an intermediate position. Bentivoglio found that the pronominal

40 alternations found in the letters respond to pragmatic factors, to switch from formality, expressed with vuestra merced, to intimacy, expressed with vos; from closeness and affection, expressed with vos, to distance, expressed with vuestra merced; from distance, expressed with vuestra merced, to anger, expressed with vos.

Bentivoglio concludes that pronominal switching between vuestra merced and vos is motivated by changes in attitude toward the interlocutor, where vuestra merced is used generally when the attitude toward the interlocutor is changed to indicate distance, formality, recommendations, or complaints; vos is used when the attitude toward the interlocutor is changed to indicate affection, familiarity, or even anger.

Fernández Martín (2012) argues that in the sixteenth century, the singular address form vuestra merced, the predecessor of usted, was the marked form, as opposed to vos, and that vuestra merced was used to alleviate face-threatening situations in complaints, criticism, anger, or to be extremely deferential to capture the goodwill of the interlocutor. Regarding plural address forms, Fernández Martín (2012) argues that ustedes, as opposed to vosotros, was used in the eighteenth century not only in formal contexts and in asymmetric interpersonal relationships, but also in informal contexts, thus alternating with vosotros, for distancing when expressing anger or criticism toward the interlocutors.

2.4.2. Pragmatic Effects on Current Singular Address Forms

Research on singular Spanish address forms also finds an effect of pragmatic factors on the variable use of second person pronouns in Spanish. For example, Castro

(2001) studied the alternation of usted, tú, and vos ‘you (singular)’ in Honduras. She

41 conducted a questionnaire, where she asked speakers to reflect explicitly about their pronominal use. Castro also conducted oral interviews to elicit the production of second person singular address forms. Castro argues that analyzing address form switching by considering the effect of static binary social categories such as gender, socioeconomic status, and static binary pragmatic categories such as asymmetric and symmetric interpersonal relationships is not enough to determine the precise causes of address form switching. This is because the same speaker may use, for example, vos and usted in a similar situation.

More precisely, Castro found that usted is used for social distance and vos for intimacy, while tú is used as an intermediate form between usted and vos in terms of social distance and intimacy. Vos is used in informal contexts in symmetric interpersonal relationships, for familiarity, but also for expressing aggressiveness, or anger toward the interlocutor, and, on the other hand, for expressing affection too.

Usted is used for expressing social distance, with unknown people, in formal situations, in an asymmetric power relationship, to show deference. However, usted is also used to express anger toward the interlocutor, and also for intimacy, affection, or, on the other hand, to mitigate the force of criticism, complaints, promises, persuasive acts, and threats. Furthermore, Castro found that there is meaning encoded in the pronoun switch from vos to usted, more precisely to mitigate criticisms, and petitions, and to congratulate. On the other hand, a temporary switch from usted to vos can indicate anger. Therefore, Castro argues that pronominal switching is:

a systematic process that is motivated by pragmatic factors such as the need to: (1)

encode emotional meanings (i.e., anger, empathy, dislike, rejection, love and

tenderness); (2) aggravate or soften the illocutionary force of speech acts, such as,

commands, insults, criticism, questions, request and offenses; and, (3) protect the

speaker’s face when performing acts likes apologizing, congratulating, making an

offer, or complementing (Castro 2002: 93, as cited in Fernández Martín [2012:

80]).

Sinnott (2010) studied the Spanish second person singular address forms tú and usted ‘you’ in Castilian Spanish and provided an account of their pragmatic flexibility. By performing qualitative and quantitative analyses of data collected from oral interviews and written questionnaires conducted in Madrid and Manzanares

(Spain), Sinnott demonstrated that both conventional and conversational implicatures play a role in the use of tú and usted address forms. Conventional implicatures are those “attached to a particular linguistic expression, regardless of context,” whereas conversational implicatures are context dependent (Birner 2013: 62, 66). Sinnott found that usted is the marked form which conventionally implicates social distance, understood as “differentiation felt by the speaker toward the addressee” (2010:198).

The conventional implicature of distance entailed by usted is underspecified enough to include the different possibilities that may cause distance; therefore, whether this social distance is based on age, respect, relationship, emotion, formality, or manners is context-dependent. This means that the reason behind social distance is conversationally implicated, i.e. “exactly which is meant is conveyed via CVI,” so that “distance is entailed by the V form; the motivation for acknowledging or imposing distance can vary by the context” (Sinnott 2010: 212-213). On the other hand, tú is the unmarked form that, unlike usted, does not entail social content. Tú is the unmarked form used to implicate lack of social distance or when nothing about distance is to be expressed. Furthermore, tú conventionally implicates its meanings, so

43 that all meanings contributed by tú to the utterance are context dependent. Usted is then the more pragmatically restricted form, since there has to be a reason to use usted, but not to use tú; therefore, tú is the default form in Madrid and Manzanares

Spanish, i.e. the most usual and the unspecified one. Furthermore, “the V form entails specific information whereas the T form does not” (Sinnott 2010: 210). According to

Schwenter and Torres Cacoullos (2008), as cited in Sinnott (2010: 246), “the default expression is the one that is preferred in the most frequent and, crucially, the least specified contexts.” The interaction of usted and tú also brings about different conversational implicatures, including respect, anger, and solidarity. More precisely, the switches of address forms, between tú and usted, convey, via conventional implicature, distance or lack of distance, which can be temporary or permanent. For example, a switch from tú to usted may be motivated to create temporary distance toward the interlocutor, conversationally implicating emotions like anger in a specific context. On the other hand, a switch from usted to tú may be used to remove distance, conversationally implicating, for example, disrespect toward the interlocutor in a specific context (Sinnott 2010).

Millán (2011) conducted research on the second person singular pronouns that alternate in informal contexts in different areas of Colombian Spanish, i.e. tú, vos, and usted ‘you’, in what she calls a system of “mixed-use” (2011: 34). Millán collected data from participants from Cali and Medellín, by using sociodemographic and written questionnaires, as well as oral interviews. Millán found that emotional closeness is a significant factor in the pronominal alternation. More precisely, she found that, in Medellín, usted is used when negative feelings toward the interlocutor are involved; but when the emotional relationship between the interactants is positive

44 or neutral, vos or usted are used. For Cali, Millán found that usted is preferred to express negative feelings, like anger, toward the interlocutor; while tú is preferred when the feelings toward the interlocutor are positive or neutral.

Johnson (2016) studied the alternation of voseo and tuteo verb forms in negative commands in Argentinian Spanish. She also addressed pragmatic differences between the variants. In order to do that, she used an online forced-choice survey which included immediate and neutral contexts. An immediate context was defined as one in which the addressee must react to the speaker’s desires quickly, and the actions are already in progress; therefore, the goal of the command is to change a state of affairs.

Immediate contexts were further divided into anger and non-anger contexts. Neutral contexts were defined as those in which actions are not yet in progress; consequently, commands in neutral contexts are not aimed at changing a state of affairs. Participants had to choose a negative command response for each context. Then, the author included type of context as a factor in her statistical analysis. Johnson found that voseo verb forms were used significantly more in immediate contexts than in neutral contexts, and, more precisely, more in anger contexts than in other immediate contexts.

Raymond (2016) addresses Spanish second person singular pronouns that alternate in different dialects of Spanish, tú, vos, usted ‘you.’ He analyzed data from institutional talk, e.g. political talks, and from everyday conversation. Raymond argues that studies of Spanish pronouns typically “assign one pronoun to one (type of) interlocutor” (2016: 641). Whereas previous studies argue that usted occurs in a specific community in specific contexts or is produced by specific interlocutors, social distance/intimacy is not absolute. Raymond argues that Spanish pronouns are

45 semantically coded for social distance but that social distance has an in-the-moment pragmatic meaning which is not dependent upon static social categories such as gender, sex or age, but depends instead on moment-to-moment negotiable identity, which determines pronoun choice. This means that during a single interaction, speakers take different identity stances which are not pre-established or permanent, and this is expressed through address form switching. Therefore, on one hand, interlocutors possess identity statuses, i.e. a more permanent category under which they are classified, like a mother, a lawyer, a young speaker, a high education speaker; at the same time, different moment-by-moment identity stances are taken and these are conveyed through address form switching. Furthermore, this pronoun switching to take different identity stances in a single interaction occurs in the different Spanish dialects, in different situations and contexts, across genders, ages, social statuses, and interpersonal relationships, i.e. it is not bound to static categories. For example, in a conversation between a Guatemalan mother and her daughter, during the same interaction, the mother shifts from the unmarked, familiar vos, to the marked usted, in order to switch from intimacy to distance, to give a negative assessment and reprimand her daughter. In another example, in Los Angeles Vernacular Spanish, a speaker shifts from the unmarked and default tú, which he has been using during an interaction to address his interlocutor, to marked usted. Furthermore, he accompanies the address form shift via pre-verbal placement of the pronoun usted, topicalizing it.

The address pronoun shift accompanied by the syntactic fronting of the pronoun serves to indicate that the speaker takes a stance acknowledging that social unfamiliarity or distance has emerged between the interlocutors. In another example, a switch from usted to tú is illustrated. More precisely, a lawyer whose legal

46 knowledge has been put into question by another lawyer, switches from usted to tú to counter that challenge, and thus indicate, by using tú, that both share the same legal knowledge, the same identity stance as lawyers (Raymond 2016).

2.4.3. Implications for this Dissertation

Previous research on current Andalusian second person plural address forms treats pragmatic differences between ustedes/vosotros and 2PL/3PL verb forms only in terms of contextual (in)formality and (a)symmetric interpersonal relationships; thus, previous research on Andalusian plural address forms does not find differences between the variants (Lara Bermejo 2015; Narbona et al. 1998). On the other hand, previous studies on historical plural and singular address forms and studies on current singular address forms go beyond the analysis of the traditional factors of

(in)formality or (a)symmetry of interpersonal relationships. Thus, they do find an effect of pragmatic factors on the variable use of second person pronouns in Spanish; they find that address form switching is used to convey different attitudes and feelings toward the interlocutor(s). More precisely, previous research finds that usted nowadays is the marked form, as opposed to vos or tú, as well as vuestra(s) merced(s) and usted(es) were historically the marked forms, as opposed to vos/ tú and vosotros, associated with non-neutrality, to express negative feelings toward the interlocutor or to alleviate face-threatening situations (Fernández Martín 2012; Millán 2011; Sinnott

2010). Therefore, in the present dissertation, I hypothesize that 2PL pronominal and verbal switching in Andalusian Spanish is also motivated in part by pragmatic factors that go beyond the (in)formality of the context and power relationships. In order to

47 test that hypothesis, I consider and analyze context type, e.g. anger, praise, or apology context, and I hypothesize that context type is an important predictor of plural address forms in Central and Western Andalusia.

2.4.4. The Pragmatic Meaning of Online Special Features

Data gathered from online sources, more precisely, from online fora and Twitter, make up a large amount of the data used for the present dissertation. Therefore, pragmatic differences for data from online sources are considered by examining the co-occurrence of what I call online special features, which have also been called computer-mediated communication cues (Kalman and Gergle 2014). Online features or computer-mediated communication cues (CMC) are what in CMC substitute for traditional non-verbal and paralinguistic cues indicating affective states which are present in speech, such as facial expressions or other bodily indicators (Danet 2010;

Dresner and Herring 2010). Online special features or CMC cues can substitute too for specific intonation and/or gestures (Kalman and Gergle 2014). Thus, online features are discourse-level features of CMC which modify the meaning of the written message, since the information conveyed by online features is not conveyed by the written words alone. These CMC cues or online features include examples like multiple punctuation, e.g. ???; eccentric non-standard spelling like letter repetition, used frequently to represent an extended phoneme, most frequently articulable in speech, and encoded by the repeated letter, e.g. a in Holaaaa ‘Helloooo’; all capital letters, e.g. HOLA ‘HELLO’; all lower-case, e.g. hola, ¿qué quieres? ‘hi, what do you want?’; emojis, i.e. pictographs of faces, objects or symbols, e.g.  , ☹;

48 emoticons, emotion icons, or smileys, i.e. graphic icons representing human faces to express emotions and which are read sideways, with the head tilted toward one’s left shoulder, e.g. :-) (smile), :-( (frown); written-out laughter, e.g. jajaja ‘hahaha’; spaces between letters, ellipsis, run-together words, blank spaces in line, interjections such as

¡ah! ‘ah!,’, vocal segregates to fill pauses such as eh ‘er.’ All these examples are used to emulate non-verbal features of speech and compensate for the lack of visual information in written communication. With the use of online features, CMC can be as effective as traditional modes of communication, thus making it possible to convey socio-emotional cues through CMC. On the other hand, unlike traditional nonverbal cues, which don’t evolve in a short period of time and which vary from culture to culture, online features are more dynamic and rapidly evolving (Danet 2010; Kalman and Gergle 2014).

The interpretation of the meaning of online features is complex, and, like that of traditional nonverbal cues, is highly context-dependent. Dresner and Herring (2010) argue that the use of online special features such as emoticons goes beyond expressing a single emotion, like happiness, or does not always express an emotion, but instead its main function is to contribute to the pragmatic meaning of the utterance in which they occur. Therefore, online features are not independent from language but form part of the text and are essential to interpret the linguistic information contained in the written message. More precisely, Dresner and Herring argue that online features can be used as indicators of the illocutionary force of the textual utterance where they are produced. Therefore, not only are online features such as smileys, frowns, and winks used to express emotion, but they are also used to indicate something about the force of the message where they are contained. Thus, online special features like

49 smileys, e.g. ☺ or XD, are used to alleviate face-threatening situations in apologies, criticism, requests, and complaints. For example, a wink, expressed with an emoji,  , or with an emoticon, ;-), may be used to indicate that the writer does not intend a literal interpretation of his message, i.e. he/she is not serious about the propositional content of his written message. The same winking online feature can be used to mitigate a face-threatening speech act, e.g. indicating that the force of the utterance is assertive instead of directive. This means that the utterance is not a command but just an expression of the writer’s preferences, as in Me gustaría que termináramos esto hoy ;-) ‘I would like for us to finish this today ;-).’ A smiley, :-) is also frequently used for mitigation purposes instead of merely indicating a happy state. For example, a smiley can be used to mitigate a request, to not impose on the interlocutor, as in Me gustaría recibir más feedback :) ‘I would like to receive more feedback :).’ A smiley can also be used to soften the force of a strong complaint, to not seem rude toward the interlocutor but to appear friendly toward the interlocutor, by expressing just an assertion. This is achieved by modifying the complaint with the smiley, which makes the complaint seem more like a humorous complaint, e.g. ¡odio este clima de mierda!

:-) ‘I hate this damn weather! :-).’ A smiley can be also used to achieve a non-literal interpretation of the message, to indicate that the propositional content of the written message should not be taken seriously, thus reducing its force, as in me voy a tener que cortar estos pelos :) ‘I am going to have to get this hair cut :).’ On the other hand, special online features are also used conventionally, i.e. to map directly onto facial expressions. This means that a smiley can be also used to indicate happiness, and a sad emoji or emoticon to indicate sadness, or an anger emoticon or emoji can be used to indicate anger in a context where the writer means to be openly rude or aggressive,

50 e.g. (Dresner and Herring 2010). Whether online special features are used to mitigate the force of a speech act or whether they are used to conventionally map directly onto facial expressions to indicate emotions like anger, sadness or happiness, is, as aforementioned, highly context-sensitive. This means that their interpretation depends highly on the context and, therefore, the same feature does not have a unique meaning; for instance, a smiley can indicate affection or happiness, but it can also indicate mitigation of a face-threatening situation.

Summarizing, special online features are used to express different attitudes toward the interlocutor(s), like anger, or are used for apologizing, for criticizing, for praising, or for mitigation, which are functions involving temporary distancing from the interlocutors, or non-neutrality. In this dissertation, I hypothesize that 2PL pronominal and verbal switching in Andalusian Spanish is motivated in part by pragmatic factors.

Therefore, for the data gathered from online sources, I hypothesize that special online features, as contributors to the pragmatic meaning of the utterance, are an important predictor in the variation.

2.5. Research Questions and Hypotheses

The general research question of this dissertation is: what are the linguistic, social and pragmatic factors that condition the 2PL pronominal alternation between ustedes and vosotros and the verbal alternation between 2PL and 3PL verb forms with ustedes, in Central and Western Andalusian Spanish? More specifically, research question 1 of this dissertation is:

1) Is the system in a change in progress, with ustedes plus the 2PL verb form as

an obsolescing linguistic element and ustedes plus the 3PL verb form as the

innovation?

The sub questions within research question 1 are:

a. What are the different patterns of the change in progress in Central versus

Western Andalusia?

b. How is the subject pronominal variation different from the verbal variation

regarding a possible change in progress?

Research question 2 of this dissertation is:

2) Are there discourse-pragmatic differences between ustedes and vosotros and

between the 2PL and the 3PL verb forms?

The sub questions within research question 2 are the following, similar to those within research question 1:

a) How do discourse-pragmatic factors affect the variation in Central Andalusia

differently from Western Andalusia?

b) How is the subject pronominal variation different from the verbal variation

regarding discourse-pragmatic factors?

In order to consider the patterns of variation in the different areas of Andalusia, to explore regional differences, I included the factor region in the analysis, with a category for each Andalusian province where the pronominal and verbal variation occurs, i.e. Málaga and Córdoba in Central Andalusia, and Cádiz, Seville, and Huelva in Western Andalusia. Based on the findings by Lara Bermejo (2015), who found that

3PL morphology originated and is more common in Western Andalusia, and on the fact that Central Andalusia has an intermediate position in linguistic terms between

Eastern Andalusian and Western Andalusian (Villena Ponsoda 2000), I predict that the standard Castilian variant, i.e. vosotros plus the 2PL verb form as opposed to ustedes with either the 2PL or the 3PL verb form, is more widespread in Central

Andalusia than in Western Andalusia. On the other hand, I hypothesize that ustedes with either the 2PL or the 3PL verb forms is more widespread in Western Andalusia than in Central Andalusia.

Regarding the hypothesis of the change in progress, as seen in section 2.2.1., I predict that the plural address system in Central and Western Andalusia is in a process of change where the mixed pattern of ustedes with the 2PL verb form is an obsolescing linguistic element. That ustedes with the 2PL verb form is an obsolescing linguistic element is based on the fact that there were mixed patterns in the past including the use of vuestra(s) merced(es) in agreement with 2PL verb forms, when the address system was in a process of change toward the generalization of use of vuestra(s) merced(es) plus third person verb forms. I also base my hypothesis on the fact that the mixed pattern, of ustedes plus 2PL verb forms, has been found nowadays in other areas too, but as an archaizing element limited to a few occurrences restricted to some very specific rural areas. I also base my hypothesis on the fact that ustedes with the 2PL occurs more frequently in pre-verbal position than in the post-verbal position, as was the case of vuestra(s) merced(es) before being reanalyzed as the subject, when the address system was in a process of change (Jaime Jiménez 2014;

Fernández Martín 2012). I also based my hypothesis on the fact that the rate of vosotros is double the rate of ustedes with the 2PL verb form in Central and Western

Andalusia (Jaime Jiménez 2016).

Furthermore, findings from previous work suggest that the 3PL verb form with ustedes is the innovation, since the 3PL morphology in Andalusia is following a similar pattern to the voseo morphology spread in Argentina, with more occurrences of the 3PL morphology first in the imperative mood and with stressed pronouns rather than with clitics (Fernández Martín 2012; Jaime Jiménez 2014; Lara Bermejo 2015).

Moreover, instances of mixed patterns, i.e. of ustedes plus the 2PL verb form have been found in the Canary Islands and in Latin America, where 2PL verb forms and vosotros eventually fell into disuse, as also did the mixed pattern of ustedes plus the

2PL verb form, except for a few instances of ustedes plus 2PL verb forms in some rural areas of the Canary Islands (Fernández Martín 2012). In both Latin American and the Canary Islands, the plural address form used is now ustedes plus the 3PL verb form, which could possibly be the direction plural address forms are taking in at least some areas of Andalusia. More precisely, I predict that the change is toward ustedes plus the 3PL verb form in Western Andalusia, where the dialect is more differentiated from the Castilian norm, and where the use of the 3PL morphology in informal contexts originated (Lara Bermejo 2015). On the other hand, I hypothesize that in

Central Andalusia the change is toward vosotros plus the 2PL verb form, since the dialect in Central Andalusian is closer to the Castilian norm. Therefore, I predict that in Western Andalusia, the change in progress which the system is hypothesized to be undergoing is specifically reflected in the verbal alternation of 2PL/3PL verb forms with ustedes, where the change is predicted to be toward the 3PL verb form. In

Central Andalusia, I predict that the change in progress is specifically reflected in the pronominal alternation vosotros/ustedes plus the 2PL verb form, since the change is predicted to be toward vosotros plus the 2PL verb form in Central Andalusia.

The hypothesis of the change in progress is tested by exploring the effect of linguistic factors that have been shown to have a role in processes of change and in conditioning obsolescing linguistic elements. These factors include frequency, priming, intervening material, and clause type (see Section 2.3.). They also include placement (see Section 2.1.), and functionality of the pronouns, ustedes and vosotros, as subject, or as object of preposition. More precisely, I predict that:

(i) Frequent verbs will lead to more use of ustedes plus the 2PL verb form over

vosotros and over the 3PL verb form with ustedes.

(ii) Priming will have a role in the conservation of the old pattern, i.e. the

combination of the 2PL verb form with ustedes, as opposed to vosotros plus

the 2PL verb form and as opposed to ustedes plus the 3PL verb form. This

means that the probability of a prior ustedes plus the 2PL verb form to lead to

a subsequent ustedes plus the 2PL verb form will be greater than the

probability of a prior vosotros plus the 2PL verb form to lead to a subsequent

vosotros plus the 2PL verb form. The probability of a prior ustedes plus the

2PL verb form to lead to a subsequent ustedes plus the 2PL verb form will be

also greater than the probability of a prior ustedes plus the 3PL verb form to

lead to a subsequent ustedes plus 3PL verb form.

(iii) Intervening material between subject and verb will lead to more use of the

productive sequences, i.e. vosotros plus the 2PL verb form and ustedes plus

the 3PL verb form. No intervening material between subject and verb will

lead to more use of the 2PL verb form with ustedes.

(iv) Ustedes plus the 2PL will occur more frequently in subordinate clauses than

in main clauses, and vosotros plus the 2PL verb form, and the 3PL with

ustedes will occur more frequently in main clauses.

(v) Preverbal placement of the subject pronoun ustedes will lead to more use of

ustedes with the 2PL verb form over vosotros and over the 3PL verb form

with ustedes; the effect of pronominal placement is explored again in this

dissertation, to determine whether ustedes appears more frequently in

preverbal position than in post-verbal position, like vuestra(s) merced(es)

when the pronominal system was still in a process of change and vuestra(s)

merced(es) was not fully established as an address form yet.

(vi) Ustedes will appear more frequently as subject than as object of preposition;

for objects of preposition, vosotros will be used more frequently, if ustedes

plus the 2PL is the obsolete, unproductive form conserved as a pre-verbal

subject like vuestras mercedes.

The hypothesis that the system is in a process of change where ustedes plus the

2PL verb form is an obsolete form, and that the change is toward ustedes plus the 3PL verb form in Western Andalusia, but toward vosotros plus the 2PL verb form in

Central Andalusia is also explored through the effect of social factors in the present dissertation. More precisely, social factors include gender, sex, education, and urban or rural origin. I hypothesize that male, older speakers with a rural origin and lower level of education are more likely to use ustedes with the 2PL as opposed to vosotros in Central Andalusia, and as opposed to ustedes plus the 3PL verb form in Western

Andalusia. Furthermore, ustedes plus the 3PL verb will be produced more frequently by women and younger speakers. More precisely, young and female speakers will be

56 more likely to produce ustedes plus the 3PL over the 2PL in Western Andalusia if the change is toward ustedes plus the 3PL in Western Andalusia; in Central Andalusia, on the other hand, young and female speakers will be more likely to produce vosotros plus the 2PL verb form over ustedes plus the 2PL, if there is in fact a leveling toward the Castilian norm of vosotros plus the 2PL in Central Andalusia.

Based on the pragmatic studies on variable use of Spanish address forms reviewed in section 2.4., I hypothesize that there are discourse-pragmatic differences between ustedes and vosotros. More precisely, I expect non-neutrality to play an important role in the use of vosotros/ustedes. I hypothesize that ustedes plus the 2PL verb form, despite being used in informal contexts like vosotros, is still used to express non- neutrality as opposed to vosotros, and that the 3PL verb form with ustedes is used for non-neutrality as well. More precisely, I hypothesize that ustedes plus the 2PL or the

3PL form of the verb is the marked form conventionally implicating non-neutrality.

The specific attitude or feeling toward the interlocutors, either positive or negative, is context-dependent, which means that the specific attitude is a conversational implicature. On the other hand, I hypothesize that vosotros is the non-marked form, used when the speaker desires to express neutrality or when nothing about neutrality is to be expressed. Therefore, a parallel distinction as the one found for tú/usted by

Sinnott (2010) is predicted to be applicable to the 2PL Andalusian address forms in this dissertation. More precisely, I hypothesize that the conventional implicature of non-neutrality entailed by ustedes plus the 2PL or the 3PL verb form is underspecified for the different attitudes, positive or negative, that may be expressed toward the interlocutors, which are instead specified via conversational implicatures. I predict that the ustedes form entails specific information, that of non-neutrality, whereas

57 vosotros does not, so that vosotros can be used to deny that non-neutrality or to not comment on neutrality. This hypothesis is tested by operationalizing pragmatic implicatures of the contexts where ustedes/vosotros plus the 2PL verb form occur, to analyze them as factors that condition the variation. If the hypothesis is true, ustedes plus the 2PL form of the verb, as opposed to vosotros, will be more likely to occur in all contexts where any kind of positive or negative feelings are being expressed toward the interlocutors. Vosotros, on the other hand, will be more likely to occur in contexts where no particular feelings are expressed toward the interlocutors. Due to the association of vosotros with neutrality and ustedes with non-neutrality, if the hypothesis is true, ustedes is the variant used by speakers to bring about pragmatic enrichment that is not brought about when using vosotros, according to Horn's

‘‘Division of Pragmatic Labor’’ (Horn 1984). On the other hand, there shouldn’t be significant differences between 2PL and 3PL with ustedes regarding pragmatic factors, as both variants are hypothesized to indicate non-neutrality similarly.

More precisely, the hypothesis of the discourse-pragmatic differences is tested by exploring the effect of the extra-linguistic factor of online special features (see

Section 2.4.4.), as well as the effect of the linguistic factor mood. I predict that, for data from online sources, online special features used to express non-neutrality will lead to more use of ustedes plus the 2PL or plus the 3PL verb form, as opposed to vosotros. For the rest of the data, I hypothesize that different kinds of non-neutral contexts, including anger, praise, apology, criticism contexts, will lead to more use of ustedes plus the 2PL or the 3PL verb form as opposed to vosotros. Furthermore, if ustedes plus the 2PL or the 3PL verb form are used to express non-neutrality, the variation will be affected by mood, where ustedes with the 2PL or the 3PL verb form

58 will be more likely to occur in the imperative mood, to alleviate the face-threatening situation involved by commands, or to be more openly aggressive or rude in the command.

Chapter 3: Methodology

In order to answer the research questions posed in this dissertation, due to the complexity of the phenomenon, data was extracted from different types of sources.

These include online sources, namely, online fora and Twitter, the oral corpus

PRESEEA Málaga, an elicitation task, and an online forced-choice survey. Different datasets were created and different analyses were conducted for each type of source.

For all the datasets, a token is each occurrence of either ustedes or vosotros as the subject pronoun of a 2PL verb form, as shown in (7).

(7) a. Ustedes queréis que sea el mejor central del mundo.

‘You [ustedes] want [2PL] him to be the best center-back in the world.’

b. No se que opinareis vosotros.

‘I don’t know what you [vosotros] think [2PL] about it.’

The analyses conducted in the present dissertation include ustedes and vosotros as subject pronouns of only a 2PL verb form and not of a 3PL verb form because there is no variation of vosotros with 2PL and 3PL. Vosotros only occurs with the 2PL verb form, except for rare instances. The combination of vosotros with the 3PL form of the verb (7) was attested in the literature only for some occurrences in the nineteenth century in Latin American Spanish (Fernández Martín 2012). Its use has not been attested since then in the literature. In the present study, only six tokens, produced by speakers from Seville (Western Andalusia), Córdoba (Central Andalusia), and Jaén

(Eastern Andalusia), were found of that combination, out of 4,171 tokens where it could have occurred, which means that there is 0.14% of vosotros pus the 3PL form of the verb. Those six tokens were extracted from the online sources and no tokens of

60 the combination of vosotros with the 3PL form of the verb were found in the oral corpus PRESEEA Málaga, or from the oral elicitation task. Therefore, vosotros plus the 3PL verb form was not included in this study, and it was not included in the online forced-choice survey as an option participants could choose. Vosotros plus the 2PL verb form is illustrated in (8), which is one of the six tokens found for the present study.

(8) Saquen vosotros vuestras conclusiones.

‘Draw [3PL] you guys [vosotros] your own conclusions.’

Null pronouns are counted as a token of vosotros or ustedes in contexts where subject unexpression is favored, i.e. in subordinate clauses, see (9) and (10), and coordinated sentences with prior co-referential explicit subjects in the main clause, see (11) (Torres Cacoullos and Travis 2013). In all the other cases, null pronouns were excluded as tokens. Therefore, most of the tokens had overt subjects. For online sources, I also discarded tokens produced by speakers whose origin could not be determined.

(9) Ustedes decid que ø sois cristianos.

‘You [ustedes] say [2PL] that you [null pronoun] are [2PL] Christians.’

(10) Ustedes cuando ø veis Hermano Mayor también soltáis la típica frase.

‘You [ustedes] when you [null pronoun] see [2PL] Big Brother also say

[2PL] the typical phrase.’

(11) Vosotros sois unos luchadores y ø saldréis de esta.

‘You [vosotros] are fighters and [null pronoun] will get [2PL] over this.’

In an example like (9), the explicit ustedes was coded with its verb decís was coded as one token of the variant ustedes, and the null subordinate subject of the verb sois was coded as another token of ustedes. In an example like (10), the explicit ustedes along with its verb soltáis was coded as one token of the variant ustedes, and the null subordinate subject of the verb veis was coded as another token of ustedes. In an example like (11), the explicit vosotros was coded with its verb sois was coded as one token of the variant vosotros, and the coordinate null subject of the verb saldréis was coded as another token of vosotros. Cases of use of both overt ustedes and vosotros in the same sentence in examples of the type of (9), (10), and (11) were not found. This means that examples such as *ustedes decís que vosotros sois; *ustedes cuando vosotros veis; *vosotros sois unos luchadores y ustedes saldréis de esta are ungrammatical.

Additionally, in the oral elicitation task, tokens include occurrences of ustedes and vosotros as objects of prepositions, along with occurrences of ustedes and vosotros as subjects of the 2PL verb form. Example (12) illustrates an occurrence of ustedes as object of the preposition a, from the oral task, as well as an occurrence of vosotros as object of the preposition de, produced by the same speaker in the same utterance.

(12) Si soy buen cocinero es porque gracias a ustedes, que he aprendido de

vosotros.

‘If I am good cook is because thanks to you guys [ustedes], since I have learnt

from you guys [vosotros].’

Also counted as a token is each occurrence of either a 2PL or a 3PL verb form with ustedes as its subject pronoun. As was mentioned before for the pronominal

62 alternation, vosotros as subject pronoun is not included in the verb form alternation because there is no variation of verb forms with vosotros. The verb form alternation is illustrated in (13).

(13) a. Ustedes queréis que sea el mejor central del mundo.

‘You [ustedes] want [2PL] him to be the best midfielder in the world.’

b. Tienen ustedes una magnifica, yo diria extraordinaria afición.

‘You [ustedes] have [3PL] amazing, I would say wonderful fans.’

Like for the pronominal alternation, 2PL verb forms with null pronouns are counted as a token of ustedes plus 2PL verb forms in contexts where an unexpressed subject pronoun is more frequently found, i.e. in subordinate clauses, where a prior explicit ustedes is in the main clause, see (9) and (10) above; and coordinated sentences with co-referential subjects, where the prior subject is an overt ustedes, see

(11) above. All 3PL verb forms were counted as tokens, both with null and overt ustedes, provided the reference was a second person plural, i.e. a group of addressees, and not a third person plural, i.e. a group of people where nobody is part of the audience.

A total of 7,852 tokens were extracted from all the sources. Two dependent linguistic variables are analyzed for each dataset in separate analyses, since there is no verbal variation for vosotros. One dependent linguistic variable is 2PL subject pronoun, with two variants, ustedes and vosotros. The other dependent linguistic variable analyzed in the present study is verb form, which has two variants, the 2PL verb form and the 3PL verb form. The two dependent linguistic variants, along with their variants, are included in Table 5.

Dependent Variable 1: Dependent Variable 2: 2PL subject pronoun verb form Variant 1 Vosotros 3PL (quieren)+ustedes Variant 2 Ustedes 2PL (queréis) +ustedes

Table 5. Dependent variables and their variants.

3.1. Data from Online Sources and PRESEEA Málaga

3.1.1. Corpus and Tokens

I extracted data from the oral corpus PRESEEA Málaga, developed by the research team VUM (Urban Vernacular of Málaga), at the University of Málaga, Central

Andalusia. This corpus has online access and is readily available for analysis. There is, however, a lack of corpora for the other geographical areas where the phenomenon occurs, such as Huelva, Cádiz, and Córdoba, and some of the existing corpora are not up to date, such as the 80s corpus for Seville (Lamíquiz and Pineda 1983). There is also a lack of relevant data in other corpora, such as in the Habla culta de Sevilla corpus of 2005 (Carbonero and Ortiz 2005). In this corpus, only singular address forms tú/usted as subjects of singular verb forms are found, as mentioned by

Fernández Martín (2012). Furthermore, there is a lack of sufficient tokens in existing corpora for the variables of the present study, like in the oral corpus PRESEEA

Málaga. Therefore, I created a corpus from online sources. More precisely, I extracted tokens from online soccer fora from Málaga (Central Andalusia), and from Huelva and Seville (Western Andalusia); from a Carnival online forum from Cádiz (Western

Andalusia); and from Twitter for all the provinces in Central and Western Andalusia,

64 i.e. Huelva, Cádiz, Seville, Málaga, and Córdoba. The kinds of sources provided, i.e. online soccer fora of the local teams, and the online Carnival forum, guaranteed the origin of the participants to a large extent. Local soccer teams from Seville, Huelva, and Málaga are typically followed just by people from those areas. The people participating in the Carnival forum are big fans of this folkloric festivity from Cádiz, and these fans are typically from Cádiz. In any case, the origin of the speakers was also checked on their public online profiles, publicly posted on the fora. For those not available, the origin was extracted from the information they provided in the public conversations on the fora. For Twitter, the origin of speakers was determined through

Twitter geotags (Page et al. 2014). For all the online sources, the tokens from speakers whose origin could not be determined were discarded, which were a total of

50 tokens.

Online sources such as these present important advantages. They make large-scale studies easier, variation can be studied comparing different sources, and change in real time can be analyzed (Tagliamonte 2012). Online sources provide “rich vernacular corpora where people live and breathe and talk to each other.”

Furthermore, by extracting data from online sources, we are avoiding the observer’s paradox (Tagliamonte 2012: 106). In addition, in online fora and Twitter, the language used is still semi-conversational. As argued by Danet (2010), since online data is dynamic, interactive, and ephemeral, it is experienced frequently as speech- like. Furthermore, online sources provide very recent data which allows us to study the current status of the phenomenon. The oldest tokens collected for the present study were produced in 2007, and the most recent ones were produced in October

2017, i.e. over a ten-year timespan.

From online sources and the PRESEEA Málaga, I extracted a total of 2,712 tokens. These include 2,408 tokens for the ustedes/vosotros alternation, and 1,170 tokens for the 2PL/3PL verb form alternation with ustedes as a subject.1 As can be observed on the left side of Figure 4, vosotros is more common than ustedes. As can be observed on the right side of Figure 4, the 2PL verb form with ustedes is much more common than the 3PL verb form with ustedes, in contrast to the normative paradigm which requires 3PL conjugation for ustedes.

Pronominal and verbal alternations 100% 90% 80% 74% 70% 62% 60% 50% 37% 40% 30% 26% 20% 10% 0% Ustedes Vosotros 3PL 2PL

Figure 4. % of tokens for each variant in the pronominal and the verbal alternations.

Online sources and PRESEEA Málaga.

1 The totals for the pronominal and verbal alternation do not add up to the total amount of tokens because some of the verbal tokens are the same tokens that form part of the pronominal tokens, more precisely, the tokens corresponding to ustedes plus the 2PL verb form. 66

3.1.2. Data Analysis

Each token was coded for ten independent variables in the case of the verbal alternation and nine independent variables in the case of the pronominal alternation.

Social factors such as age, gender, and education could not be included in the analysis of this dataset because complete information about speakers is not available in online sources. Some variables were relevant for answering my first research question, and others were relevant for the second research question.

The variables that help me answer my first research question, i.e. whether ustedes plus the 2PL verb form is an obsolescing linguistic element and the 3PL with ustedes is the innovation, include priming, intervening material between subject pronoun and verb, overt/null pronoun, frequency of the lexeme in Davies’s Corpus del Español

(2016-), clause type, and subject pronoun placement, for both dependent variables, i.e. for both the pronominal and the verbal alternation. To account for priming, three levels were considered for the ustedes/vosotros alternation. These include same prior pronoun and different prior pronoun if the token has an overt subject. For example, if the token is ustedes queréis ‘you (plural) want,’ it is coded as having the same prior pronoun if there is a previous overt ustedes with the 2PL verb form; the same token, ustedes queréis, is coded as having a different prior pronoun if there is a prior overt vosotros with the 2PL verb form. The third level is null pronoun, which refers to the token having a null subject, for which priming is not considered. Priming is not considered for null subjects because priming affects overt and null subjects differently. More crucially, priming is not considered for null pronoun tokens because null subjects included in the present dissertation are those occurring in subordinate or

67 coordinate clauses after a prior co-referential overt pronoun (see examples 9, 10 and

11 in Chapter 3). This context for null subjects is not site for priming since it is favoring unexpression (Torres Cacoullos and Travis 2013). Furthermore, by considering priming for these null subjects, collinearity issues would be created where null subjects would always be primed by a same prior pronoun, which is the prior overt pronoun considered in the first place to be able to account for these null subjects as tokens. For the 2PL/3PL verb form alternation, two levels were coded for, including same prior verb form, e.g. quieren (3PL) […] comen (3PL); and different prior verb form, e.g. queréis (2PL) […] comen (3PL) ‘you (plural) want […] you

(plural) eat.’

The factor intervening material was divided into three levels, including (i) the presence of intervening material between the subject pronoun and the verb form, e.g. ustedes es que sois buenos ‘you (plural) really are good,’ (ii) the absence of intervening material between the subject pronoun and the verb form, e.g. ustedes queréis ‘you (plural) want,’ and (iii) null subject pronoun, e.g. queréis ‘you (plural) want.’ The factor overt pronoun includes two levels: overt pronoun if the subject pronoun ustedes or vosotros is explicit, versus non-overt pronoun if the subject pronoun ustedes or vosotros is null.

Based on Erker and Guy (2012), the factor lexical frequency includes two categorical levels, frequent or infrequent lexeme. However, Erker and Guy analyzed local frequency using a 1% threshold of their own data, i.e. a frequency relative to token occurrences in their own corpus, in order to be able “to attribute grammatical patterns to particular speakers’ experiences,” thus using a more usage-based approach

(2012: 530). I also used the 1% threshold of my own data in a previous study (Jaime

Jiménez 2014). However, the threshold of 1% of my own data could be skewing the results, since for instance a verb that is frequently repeated in online soccer fora, such as jugar ‘to play,’ might not be frequent overall. As Martínez Sanz and Van Herk

(2013) argue, with local frequency, topic and speaker could be determining which forms are frequent. Furthermore, unlike Erker and Guy (2012), frequency of lemmas instead of verb forms was considered, since the 2PL verb form is very infrequent as compared to the 3PL form of the verb (Bybee and Brewer 1980; Tiersma 1982).

Furthermore, according to Martínez Sanz and Van Herk (2013), lemmas should be considered since this allows for a more balanced dataset, where lemmas fill cells left empty by verb forms, and frequency effects are then related to the whole verbal paradigm. To consider global frequency of lexemes, I searched in the Spanish corpus

Davies’ Corpus del Español (2016-) to determine if a lexeme is frequent or infrequent. I established 900/1,000,000 to be the point of division since there was a large cut-off at that point, i.e. whereas just a few verbs from my dataset had a frequency of over 900 occurrences per million in the Davies corpus, most of the verbs from my dataset had a frequency below 900 occurrences per million in Davies’ corpus. Therefore, lexemes from my dataset with a frequency of 900 or more occurrences per million in Davies’ corpus were coded as frequent, whereas verb lexemes with a frequency below 900 occurrences per million in Davies’ corpus were coded as infrequent.

The factor clause type includes main clause, subordinate clause, and coordinated clause. The factor pronoun placement was divided into pre-verbal, e.g. ustedes queréis, and post-verbal, e.g. queréis ustedes ‘you (plural) want.’ A third level was used to code for null subjects, whose placement cannot be determined. The factors

69 placement, priming, intervening material and overt pronoun were included in separate analyses for the pronominal alternation since the four of them have a level for null subject pronouns, which refers to the same subset of tokens in the four cases. For the verbal alternation, placement, overt pronoun, and intervening material were included in separate analyses. Priming, on the other hand, was included in all the analyses in the verbal alternation, because priming does not have a level for null pronouns in the verbal alternation.

The variables that help answer my second research question, regarding whether there are discourse-pragmatic differences between ustedes and vosotros and between the 2PL and the 3PL verb forms, include tense-mood-aspect (TMA), and special online features. Initially, twenty-two levels were considered within the factor TMA, according to the different aspects, moods, and tenses encountered in the data. However, due to insufficient token counts throughout these categories, these were collapsed into a four-way distinction including indicative, subjunctive, imperative, and conditional.

The factor special online features was included to test the co-occurrence of online special features with ustedes and vosotros, and with the 2PL and the 3PL verb forms.

The levels, of this factor, which were initially seven, were established following online features identified in the literature analyzing the discourse-pragmatic functions of those special features (Danet 2010; Dresner and Herring 2010; Kalman and Gergle

2014). The levels of special online features were established based also on the online features observed in the data for this study. More precisely, the factor special online features was initially coded for smileys, including smiling emoticons, ☺, as well as similar smiling symbols, such as xD, which are all symbols showing positive feelings

(Danet 2010; Kalman and Gergle 2009). Another level included displeased emoticons,

70 including any kind of emoticon or similar symbol depicting negative emotions, such as anger emoticons, , frowns, , or bewilderment symbols, :S. The levels ‘!’ and

‘?’ included repetition of exclamation and question marks, respectively, with any number of repetitions. Multiplied characters was another level, including any letter repetition. The level capital letters included capital letters as well as bold and colored letters, which are, as capital letters, a way of highlighting and stressing a word or phrase. Laughter included every type of laughter expression, such as jajaja ‘hahaha,’ or LOL.

Due to poor data distribution throughout all these categories all levels were collapsed into four: (i) online special features showing positive feelings toward the interlocutors, see (14a); (ii) online features depicting negative emotions toward the interlocutors, see (14b) and (14c); (iii) absence of online features, see (15a); (iv) presence of online features in neutral contexts, see (15b). To classify online features, an analysis of the context was conducted, due to the complexity and high context sensitivity of online features. Thus, a smiley may not always indicate positive feelings toward the interlocutors. Therefore, other online features were also included in the categories of presence of negative online features and of presence of positive online features, and not only emoticons or emojis showing an anger face or smileys, respectively. More precisely, other special features used to indicate anger or sarcasm were included within the category of online features showing negative feelings toward interlocutors. Examples are capital letters representing yelling, in an anger context, or special online features having a mitigation function, such as smileys used to alleviate a face-threatening situation, see (14c) (Dresner and Herring 2010). Profanity as an indicator of an anger context was also included within this category of emojis

71 showing negative feelings toward the interlocutors. Online features other than smileys, but also indicating positive feelings toward the interlocutors, were included within the category of online features showing positive feelings toward the interlocutors. Examples are exclamation mark repetition used to praise the interlocutors, see (14a).

(14) a. enorme sois ustedes!!! Gracias y felicidades!!!

‘You [ustedes] are [2PL] great!!! Thank you and congratulations!!!’

b. Ustedes dos sobre favoritismos sois los que más tenéis callar .

‘Especially you two [ustedes] are [2PL] the ones who can’t have [2PL] a word

about favoritisms .’

c. Pues entonces disculpas xD no recuerdo haber entrado aquí a comentar

agrupaciones como tal, pero bueno, si ustedes lo decís xDD pues será

‘So then I apologize xD, I don’t remember having logged in here to really

comment on [Carnival] bands …but well, if you [ustedes] say [2PL] so xDD,

then it must be true.’

In example (14a) above, the repetition of exclamation marks is utilized to help the writer express positive feelings towards his readers, more precisely, praise. In (14b), on the other hand, an anger emoticon, , is used in an anger context. Finally, in

(14c), the online symbol representing a smiling face, xD, with repetition of D, xDD, is used. The writer does not totally agree with the other forum participants, or he is not sure about the truth of the other forum participants’ assertion. However, the writer alleviates the face-threatening situation of expressing such doubt or disagreement by apologizing as well as by using the online feature xD or xDD continuously during his contribution.

Neutral contexts are those in which online features don’t indicate a particular feeling, neither anger nor affection or positivity toward the interlocutors. Neutral online features may be used only to highlight some information, such as the capital letters in example (15b).

(15) a. ¿Que pensai vosotros sobre este tema?

‘What do you [vosotros] think [2PL] about this topic?’

b. ¿VOSOTROS K OPINAIS?

‘WHAT DO YOU [vosotros] THINK [2PL]?’

In example (15a) above, the writer is not using any online special features and is only asking for the interlocutors’ opinion. In (15b), the writer is using capital letters.

Using capital letters for an entire message is normally understood as yelling, whereas using capital letters for just one word typically has the goal of merely highlighting it.

However, in (15b), the writer is only asking about the interlocutors’ opinions, as in

(15a). Therefore, capital letters here are used in a neutral context and not for yelling in an anger context. The writer might merely be trying to make his contribution and his question for the other forum participants more visible, i.e. just highlighting it with the use of capital letters.

The variable that helps me answer the sub-question within my first and second research question, regarding different patterns of the variable use of plural address forms in Central and Western Andalusia is region. The factor region included a level for each one of the Andalusian provinces where the phenomenon occurs, i.e. Huelva,

Cádiz, and Seville (Western Andalusia), and Málaga and Córdoba (Central

Andalusia), following the geographical distribution proposed by Alvar (1973).

Eventually, however, to achieve a more balanced distribution and because it has been

73 argued that the phenomenon originated in Western Andalusia (Lara Bermejo 2015),

Huelva, Cádiz, and Seville (Western Andalusia) were collapsed into one category;

Málaga and Córdoba (Central Andalusia) were collapsed into another category.

Additionally, for the verbal alternation, the factor semantic classes of verb was also included, since I found that it was significant for the verbal alternation in a previous study (Jaime Jiménez 2014). This variable helps answer how the verbal variation is different from the subject pronominal variation in general too. Following

Erker & Guy’s division (2012), this factor includes three levels: (i) external, dynamic activity verbs, which refer to external actions, e.g. saltar ‘to jump;’ (ii) mental verbs, which include cognition verbs such as pensar ‘to think,’ saber ‘to know,’ creer ‘to believe,’ emotion and estimative verbs such as amar ‘to love,’ gustar ‘to like,’ and verbs of opinion, such as juzgar ‘to judge;’ (iii) stative verbs, which refer to nondynamic processes, like vivir ‘to live,’ tener ‘to have.’

For the online and PRESEEA Málaga dataset, mixed effects logistic regression models were created with R analyzing the pronominal alternation and the verbal alternation separately (R core team 2015). The independent variables described above were included as fixed effects. Speaker was included as a random effect to avoid the effect of individual speakers, since particular individuals in a study do not represent all the possible individuals of the population (Tagliamonte and Baayen 2012). The step function and the random forest in R were used to determine the ordering of factors in terms of their statistical significance. Then, logistic regression models were developed using the glmer function, and Anova was used to compare the models.

Conditional inference trees were used to determine predictor interactions

(Tagliamonte and Baayen 2012).

3.2. Oral Elicitation Task and Online Forced-Choice Survey

I constructed, additionally, two guided elicitation tasks and an online forced- choice survey. I constructed these tasks due to the impossibility of including social factors in the analysis of the online and PRESEEA Málaga dataset, because such information is not available in online sources. In order to obtain that social information, questions about participants’ background were included both in the guided elicitation tasks and the online forced-choice survey. Furthermore, the guided elicitation tasks and the online forced-choice survey include different pragmatic contexts, which allows controlling for pragmatic contexts and thus further exploring discourse-pragmatic effects on the variation. One of the goals of this dissertation is to determine what the pragmatic differences between the variants are, going beyond

(in)formality or (a)symmetry of interpersonal relationships. Therefore, one formal context was included in each task, but all the other constructed scenarios are informal contexts, with presence of family and friends with whom the participant has to imagine he/she is interacting.

3.2.1. Online Forced-Choice Survey

3.2.1.1. Task, Participants and Tokens

I constructed an online forced-choice survey with Google Docs, which was sent to a large number of participants using social media and e-mail (cf. Johnson 2016;

Morgan and Schwenter 2016). The online forced-choice survey includes sixteen written daily-life scenarios with different pragmatic contexts. The scenarios are each

75 followed by possible target utterances given to the participants in writing too; the participants had to choose the utterance that they thought they would more naturally say given the context. The options include either ustedes or vosotros and either the

2PL or the 3PL form of the verb with ustedes. An example context along with its target utterances is shown in (16). Participants had to choose one option only, i.e. a), b). or c).

(16) La casa de tus padres está muy mal, muy dejada y ellos están muy mayores para arreglarla o limpiarla ya. Entonces, os toca a ti y a tus hermanos, pero, al tiempo de haber quedado en arreglarla, tus hermanos te dicen que no limpiaste ni arreglaste nada al final, que sigue todo muy mal. Sabes que ellos tampoco hicieron mucho. Les dices:

a) Claro, como que vosotros limpiasteis mucho

b) Claro, como que ustedes limpiaron mucho

c) Claro, como que ustedes limpiasteis mucho

‘Your parents’ house is in a very bad state, but your parents are too old to fix it or

clean it. Therefore, your siblings and you have to do it. However, sometime after

having cleaned and organized the house, your siblings tell you that you didn’t

clean much, that the house is still dirty and messy. You know that they didn’t do

much either. You tell them: “Right, as if you guys are very clean”.’

Since infinitives, with enclisis of se, e.g. sentarse ‘you guys sit down’, is the preferred form for a command in Andalusia, both for ustedes and vosotros (Fernández

Martín 2012), for commands, infinitive forms were included in the options of the online survey, along with the finite verb forms, i.e. the 2PL or the 3PL verb form in the imperative mood.

Eight of the sixteen scenarios are distractors, where the variants under scrutiny are not included but instead other nonstandard features are included. Those features include phenomena such as the use of the definite article in front of proper names, e.g. la Elena ‘the Elena,’, or the non-normative ordering of clitics in front of the verb, with the indirect pronoun in front of the reflexive pronoun, e.g. me se cayó el vaso instead of the normative se me cayó el vaso ‘I dropped the glass’ (see Appendix A).

However, these other dialectal features are not exclusive to Andalusian Spanish, whose differential morphosyntactic features are scarce (Bustos Tovar 2001; Jiménez

Fernández 1999). The complete survey as presented to participants is in Appendix A.

A total of 474 participants completed the survey, including speakers from all the provinces where the phenomenon studied in this dissertation occurs, i.e., Málaga,

Córdoba, Huelva, Cádiz and Seville. Out of 474 participants, 58% are women, and

42% are men; 4.6% have a low level of education, 20.8% a medium level of education, and 74.5% of participants have a high level of education, as illustrated in

Figure 5.

100%

90%

80% 74.5% 70%

60%

50%

40%

30% 20.8% 20%

10% 4.6% 0% Low education Medium education High education

Figure 5. Distribution of online forced-choice survey participants by education group.

Out of the 474 participants, there are 67% young participants (from 18 to 34 years old), 28% adult participants (35 to 54 years old), and 5% old participants (55 or more years old), as illustrated in Figure 6.

100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

0% Young Adult Old

Figure 6. Distribution of online forced-choice survey participants by age group.

Out of the 474 participants, there are 31% of participants from Central Andalusia, and 69% from Western Andalusia; 82% of urban origin, and 18% of rural origin.

From the online forced-choice survey, a total of 3,681 tokens were collected.

These include a total of 3,243 tokens for the pronominal alternation, with 55.4% of vosotros and 44.6% of ustedes; the data collected also include a total of 1,675 tokens for the verbal alternation, with 74% of 2PL with ustedes, and 26% of 3 PL with ustedes. Figure 7 shows the distribution of the pronominal and verbal variants. The difference between ustedes and vosotros in the online forced-choice survey is lesser than in the online and PRESEEA Málaga data. On the other hand, the difference

78 between 2PL and 3PL verb forms is the same, with almost triple the rate of 2PL verb forms with ustedes in comparison with the rate of 3PL with ustedes.

Pronominal and verbal alternation 100% 90% 80% 74% 70% 60% 55.4% 50% 44.6% 40% 30% 26% 20% 10% 0% Ustedes Vosotros 3PL 2PL

Figure 7. Rates of pronominal and verbal variants. Online forced-choice survey.

3.2.1.2. Data Analysis

Each token was coded for five social factors and one pragmatic factor. Social factors include education: low, medium, and high; sex: female versus male; age: young, adult, and old; region: Western versus Central Andalusia; and urban versus rural origin. Following the categorization used in the oral corpus PRESEEA Málaga, I define a low level of education as elementary education or less; a medium level of education is defined as having secondary education; a high level of education is defined as university education. Young speakers are those from 18 to 34 years old; middle age or adult speakers are those between 35-54 years old; and old speakers are those 55 years or older (Vida Castro 2007). For the factor urban/rural origin, I

79 established that speakers had an urban origin if they were from the capital city of their province, e.g. speakers from the capital city of Seville in the province of Seville were considered as having an urban origin. On the other hand, speakers had a rural origin if they were from any town or village from their province, out of the capital city, e.g. speakers from the town Nerja, in the province of Málaga, were considered to have a rural origin.

Regarding the pragmatic factor, this is context type, which is the result from operationalizing the contexts I constructed for the tasks, including non-neutral and neutral contexts. Neutral contexts are those where no particular feelings are expressed toward the interlocutors, as the context included in example (17). Non-neutral contexts in the online forced-choice survey include anger (18), apology (19), criticism/irony (20), formal (21), praise contexts (22), and commands (23).

(17) Quieres ir al cine con tus amigos, a ver la película Batman vs. Superman,

pero no sabes si tus amigos la han visto. Les preguntas:

a) ¿Ustedes la han visto?

b) ¿Vosotros la habéis visto?

c) ¿Ustedes la habéis visto?

‘You want to go to the theater with your friends, to watch the movie Batman

vs. Superman, but you are not sure whether your friends already saw it or not.

You ask them: “Have you guys seen it?”’

(18) Los vecinos de la puerta de al lado llevan toda la noche con la música a todo

volumen porque tienen una fiesta en casa. No es la primera vez que lo hacen, ya lo

han hecho muchas veces. No puedes dormir por la música y al día siguiente tienes

que trabajar. Además, tienes a tu madre enferma, que tampoco puede dormir. Vas

80 y les pides amablemente que le bajen el volumen a la música, por favor, que tu madre está enferma. Te dicen que no, que no molestes, que te vayas a dormir, que eres un pesado/a. Les dices:

a) ¡¡Y vosotros lo que sois es unos hijos de puta!!

b) ¡¡Y ustedes lo que son es unos hijos de puta!!

c) ¡¡Y ustedes lo que sois es unos hijos de puta!!

‘The neighbors next door have been playing music very loud all night because

they have a party at home. They have done it before, many times. You can’t

sleep because of the music and you have to work the next day. Furthermore,

your mother, who is sick, can’t sleep either. You knock at their door and

nicely ask them to lower the volume, since your mother is sick. They refuse to

do it and tell you not to bother them, that you are nuts. You tell them: “And

you guys are motherfuckers, that’s what you are!!”’

(19) Vas conduciendo y ves que una pareja joven, en otro coche, de manera

peligrosa, te adelanta y se pone delante de ti. En un semáforo en rojo te pones

en el carril de al lado de ese coche y aprovechas para gritarle al conductor por

su irresponsabilidad, pero él te dice que van al hospital, que su mujer,

embarazada, ha roto aguas. Ves que es verdad y te sientes avergonzado. De

todas formas, ellos te piden perdón por adelantarte así. Les pides que te

perdonen ellos; les dices:

a) No, perdonarme ustedes

b) No, perdonadme vosotros

c) No, perdónenme ustedes

d) No, perdonadme ustedes

e) No, perdonarme vosotros

‘You are driving and see a young couple, in another car, pass you in a dangerous way. They come in front of you with their car. Later, when stopped because of a red light, you get to the lane next to the car that just passed you.

You use the opportunity to yell at the driver because of his irresponsibility. He tells you that they are going to the hospital, that his pregnant wife just broke water. You see that it is true and you feel ashamed. In any case, they apologize to you for having passed you like that. You apologize to them; you tell them:

“No, forgive me, you guys”.’

(20) Tus cuñado/a y tu hermana/o se quejan todo el tiempo de que se cansan

mucho al subir las escaleras de tu casa. Les dices, para que no se quejen tanto:

a) Ustedes lo que tienen que hacer es un poquito de ejercicio.

b) Ustedes lo que tenéis que hacer es un poquito de ejercicio.

c) Vosotros lo que tenéis que hacer es un poquito de ejercicio.

‘Your brother/sister in law and your sister/brother complain all the time

because they get very tired when they have to go up the stairs of your

building. In order to make them stop complaining so much, you tell them:

“What you should do is some exercise”.’

(21) Tu hijo es muy buen estudiante. Estás reunido con dos de sus profesores, que son catedráticos. Les dices que piensas que tu hijo podría llegar lejos también, como profesor, y les preguntas por su opinión. Les dices:

a) ¿qué pensáis ustedes?

b) ¿qué pensáis vosotros?

c) ¿qué piensan ustedes?

‘Your son is a good student. You are in a meeting with two of his professors,

who are full professors. You tell them that you think that your son has a great

future ahead too, as a professor. You ask them about their opinion on the

matter. You tell them: “What do you guys think?”’

(22) Tus primos te preparan una paella riquísima. Les dices:

a) ¡¡Qué bien cocináis ustedes!!

b) ¡¡Qué bien cocinan ustedes!!

c) ¡¡ Qué bien cocináis vosotros!!

‘Your cousins prepare a delicious paella for you. You tell them: “You guys cook

so well!!”’

(23) Tus hijos quieren que llames al electricista para reparar un problema que

hay en la casa. Quieres que llamen ellos. Les dices:

a) Llamadlo vosotros

b) Llamarlo ustedes

c) Llámenlo ustedes

d) Llamarlo vosotros

e) Llamadlo ustedes

‘Your children want you to call the electrician to fix some problem that you have

at home. You want them to call instead, so you tell them: “You guys call him.”’

Additionally, for the pronominal alternation, each factor was coded for a linguistic factor which I called verb. This includes the levels 2PL verb form, and infinitive verb form, since commands do frequently occur as an infinitive. Infinitives were actually included among the options in the online forced-choice survey. This factor was not included for the verbal alternation, since the dependent variable in that alternation

83 refers to the verb form, i.e. either 2PL or 3PL, with ustedes as a subject; therefore, with an infinitive, none of them is used. On the other hand, the infinitives could be included in the analysis of the pronominal alternation because in this case the variable refers to the pronoun being used, i.e. vosotros or ustedes.

Like for the data from the online Fora and Twitter and PRESEEA Málaga, for the data from the online forced-choice survey, mixed effects logistic regression models were created with R, analyzing the pronominal alternation and the verbal alternation separately (R core team 2015). The social and pragmatic independent variables described above were included as fixed effects. Speaker was included as a random effect. The step function and the random forest in R were used to determine the ordering of factors in terms of their statistical significance. Then, logistic regression models were developed using the glmer function, and Anova was used to compare the models. Conditional inference trees were used to determine predictor interactions

(Tagliamonte and Baayen 2012).

3.2.2. Oral Elicitation Task

3.2.2.1. Task, Participants and Tokens

I constructed two elicitation tasks. The first elicitation task contains 28 daily life scenarios, and the second one 30 daily life scenarios. I explained the scenarios from the elicitation tasks orally to the participants. Like Lara Bermejo (2015), I used a guiding contextualizing sentence after I explained the scenario, in order to elicit the production of ustedes/vosotros plus 2PL/3PL form of the verb, e.g. les dices que se callen ‘you tell them to shut up.’ Then, the participants had to speak to an imaginary

84 group of people described in the scenario; therefore, ustedes/vosotros plus 2PL/3PL form of the verb were orally produced by the participants. Their answers were audio- recorded. Following Lara Bermejo (2015), several scenes taken from the TV sitcom

Aquí no hay quien viva ‘No one could live here,’ and the YouTube series Malviviendo

‘Bad Living’ were included in the oral elicitation tasks. However, these scenes were also explained orally to the participant, and only some images from the video were shown to participants, in order to provide more contextualization. An example of a context including a sitcom is illustrated in example (24) below. In the daily life scenarios explained orally to the participants, I used contrastive contexts, with switch in reference, i.e. two conjoined sentences with different subjects, to elicit the production of an overt pronoun (Amaral and Schwenter 2005; Mayol 2010; Silva-

Corvalán 2001). Otherwise, whether the token was a case of use of ustedes or vosotros could not have been determined. One of the main functions of overt pronouns is to mark contrast, and the contrast between subjects in contrastive contexts in Spanish cannot be expressed by the inflectional verbal morphology only.

Therefore, overt pronouns, or, in other cases, other devices for explicitly marking a switch in reference such as contrastive adverbials, convey the contrast information

(Amaral and Schwenter 2005). A contrastive context with switch in reference used in this dissertation for eliciting the production of an overt pronoun is (24). In (24), the expected answer is ella/mi novia no será muy guapa, pero vosotros/ustedes tampoco sois/son Kevin Costner/la gran cosa/muy guapos ‘She/My girlfriend might not be very beautiful, but you guys aren’t Kevin Costner/very handsome either.’ In (24), the pronoun ustedes or vosotros is expected to be produced explicitly, contrasting with ella/mi novia ‘she’/’my girlfriend.’

(24) Dos amigos critican la fealdad de la novia de otro. Él les dice que sabe que

ella no es muy guapa pero que ellos, sus amigos, tampoco es que sean Kevin

Costner o la gran cosa (Temporada 1, capítulo 8, Malviviendo).

https://www.youtube.com/watch?v=i--XduiLsI0 (Minute 32:14; mute: 32:20)

‘Two friends are saying that the other’s girlfriend is ugly. He acknowledges that

she is not very beautiful but tells them that they are not Kevin Costner either, or

very handsome either (Season 1, episode 8, Bad Living).’

Like for the online forced-choice survey, half of the scenarios in the two oral elicitation tasks are distractors, so that the elicitation of the variants under scrutiny is not sought in those, but instead other nonstandard features are included, like those mentioned in section 3.2.1.1. The complete elicitation tasks are in Appendix B.

For the elicitation tasks, I interviewed a total of 69 speakers from Central

Andalusia, including Málaga, and from Western Andalusia, including Huelva, Cádiz and Seville. Out of the 69 speakers, there are 52% women and 48% men. Table 6 shows the participants from the province of Málaga, Central Andalusia. There is a total of 46 speakers from Málaga, Central Andalusia, including 21 males and 25 females, all of them with rural origins.

Younger generation Middle generation Older generation

Male Female Male Female Male Female Low edu 1 1 2 1 4 7 Medium edu. 5 3 1 6 1 0 High edu. 3 5 4 2 0 0

Table 6. Distribution of oral elicitation task participants from Málaga, Central

Andalusia.

There is a total of 23 speakers from Western Andalusia, including 9 males and 14 females, from Cádiz, Seville, and Huelva. 7 participants have rural origins, and the rest, i.e. 16 speakers, have an urban origin. Figure 8 shows the distribution of participants by age. 51.2% of participants are young, 17.4% are adults, and 30.4% are old.

100%

90%

80%

70%

60% 52.2% 50%

40% 30.4% 30% 17.4% 20%

10%

0% Young Adult Old

Figure 8. Distribution of participants by age. Oral elicitation task. Western Andalusia.

100%

90%

80%

70%

60% 56.5%

50%

40%

30% 21.7% 21.7% 20%

10%

0% Low education Medium education High education

Figure 9. Distribution of participants by education. Oral elicitation task. Western

Andalusia.

From the guided elicitation tasks, a total of 1,459 tokens were collected. For the ustedes/vosotros alternation, for the data from the oral elicitation task, there is a total of 1,398 tokens, with 59.6% of vosotros and 40.4% of ustedes. For the verbal alternation, there is a total of 569 tokens from the oral elicitation task, with 89.3% of

2PL with ustedes versus 10.7% of 3PL with ustedes. This means than in the data from the oral task, the difference between 3PL and 2PL is by far larger than in the other datasets, i.e. those from online sources and from the online forced-choice survey; more precisely, in the data from the oral elicitation task, the 2PL with ustedes comprises most of the data. The difference of ustedes and vosotros rates is also larger in the oral data than in the data from the online forced-choice survey; more precisely, the difference between the variants in the pronominal alternation is similar to that in the online and PRESEEA Málaga data. This is illustrated in Figure 10.

Pronominal and verbal alternation 100% 89.3% 90% 80% 70% 59.6% 60% 50% 40.4% 40% 30% 20% 10.7% 10% 0% Ustedes Vosotros 3PL 2PL

Figure 10. Rates of pronominal and verbal variants. Oral elicitation task.

3.2.2.2. Data Analysis

Each token was coded for five social factors and one pragmatic factor. Social factors include the same ones as the online forced-choice survey, i.e. education (low, medium, high), gender (female, male), age (young, adult, old), region (Western versus

Central Andalusia), and urban versus rural origin. Regarding the pragmatic factor, it refers to context type, which includes neutral contexts and different types of non- neutral contexts. A neutral context from the elicitation task is included in example

(25a), which illustrates what the interviewer explained orally to the participants. (25b) includes an answer given by a participant, after being explained the scenario in (25a).

(25) a. Les estás contando a unos vecinos tuyos que vas a ir a un concierto, de un

cantante que te gusta en tu pueblo, el fin de semana que viene, les dices que tú

vas a ir y les preguntas que si ellos van a ir también.

‘You are telling your neighbors that you are going to a concert next weekend,

to see the singer/band you like. You tell them that you are going to the concert

and ask them whether they are going too.’

b. Mira, Victoria, Loli, Carmina, hay un concierto, ¿vosotros vais a ir? Porque

yo pienso de ir, así que, si queréis ir, me avisáis.

‘Look, Victoria, Loli, Carmina, there is a concert, are you guys going?

Because I plan to go, so that, if you guys want to go, let me know.’

Non-neutral contexts include anger contexts, criticism/irony contexts, formal contexts, praise contexts, and apology contexts, like in the online forced-choice survey, and, additionally, petition contexts. Due to lack of sufficient tokens, petitions and apologies were collapsed into one category, since they both have a similar pragmatic meaning, where there is a face threatening situation that can be alleviated with the use of one variant or the other. An anger context for the elicitation task is included in example (26a) which, again, includes what the interviewer explained orally to participants, and an answer given by a participant is shown in (26b).

(26) a. Vas conduciendo por una rotonda y estás para salir de la rotonda, pero

te rebasan dos taxistas en un coche por la izquierda, aunque tú tenías

preferencia, con riesgo de accidente. Les pitas a los taxistas y encima te sacan

el dedo los dos y te llaman “hijo de puta” y te dicen que mires por dónde vas y

que aprendas a conducir, que dónde te sacaste el carnet. Te enfadas mucho y

les dices que los hijos de puta son ellos, que son ellos los que tienen que mirar

por dónde van, que son unos sinvergüenzas y que ellos son los que tienen que

aprender a conducir, no tú.

‘You are driving through a roundabout, but two taxi drivers pass you on your

left, even though you had priority. They pass you dangerously, so you honk,

but they flip you off, call you “motherfucker”, and tell you that you should

look where you are going, that you should learn to drive, and where did you

get your license. You get very mad, and tell them that they are the

motherfuckers, that they are the ones who have to look where they are going,

that they are assholes, that they are the ones who need to learn to drive, not

you.’

b. ¡¡Ustedes sí que sois unos hijos de vuestra gran madre!! ¡Así no se va por

la carretera!

A criticism/irony context is included in (27a) and the answer given by a speaker is included in (27b).

(27) a. Tus amigos te dicen que has estado todo el verano sin trabajar ni hacer nada.

Tú sabes que ellos tampoco han hecho nada. Entonces, les dices que, claro, que

como que ellos no han descansado nada este verano, que como que ellos no han

parado de trabajar ni nada este verano.

‘Your friends are telling you that you didn’t work at all this summer, that you

didn’t do anything. You know they didn’t do anything either this summer. You

tell them “yeah, right, as if you didn’t rest at all this summer, as if you didn’t stop

working at all this summer”.’

b. Claro que vosotros habéis estado todo el verano trabajando. Lo mismo que yo,

sin hacer nada todo el día.

‘Sure, as if you guys have been working all summer. The same as me, you didn’t

do anything at all.’

A formal context is illustrated in (28a), where the speaker has to address two doctors, and the answer given by a speaker is included in (28b).

(28) a. Estás en el hospital llevando a tu anciana madre, que está enferma. Los

doctores te piden permiso para hacerle un tratamiento que necesita, porque es

peligroso. Les dices que lo que ellos decidan, que tú no sabes, que ellos son los

expertos.

‘You are at the hospital, with your old mother, who is sick. The doctors ask you

for permission to use a dangerous treatment with her, a treatment she needs. You

tell them that whatever they decide, that you don’t know, and that they are the

experts.’

b. Pues lo que ustedes veáis, yo no sé de eso, pues lo que ustedes veáis será lo

correcto.

‘Whatever you guys say, I don’t know anything about this, so whatever you guys

think, will be the right thing to do.’

An apology context is included in (29a), and the answer given by a speaker is in

(29b).

(29) a. Tus hermanos te piden perdón por no haberte llamado ni visitado mucho

este año, que no os habéis visto casi nada, pero tú sabes que tú tampoco has estado

muy pendiente de ellos, porque has estado muy ocupado, entonces les dice que no,

que te perdonen ellos, que tampoco has llamado mucho ni nada, que tú también

has estado muy ocupado.

‘Your siblings apologize because this year they haven’t paid much attention to

you, they haven’t called you or visited you, because they have been very busy.

You tell them that they are the ones who should forgive you because you have

also been busy and haven’t called or visited either.’

b. No, que no pasa nada, que perdonad vosotros también por mi parte.

‘Not, it’s ok, forgive me you guys, I am also to blame.’

A petition context is included in (30a), and the answer given by a speaker is included in (30b).

(30) a. Tus vecinos son muy simpáticos y te llevas muy bien con ellos, pero no

son muy limpios y siempre te toca a ti y tu familia limpiar las escaleras y el portal

de tu bloque. Un día te encuentras a varios de ellos al bajar las escaleras. Hablas

con ellos y les dices amablemente que sería bueno que no sólo tú y tu familia, sino

que también ellos limpiaran la escalera y el portal, para mantenerlo todo limpio.

‘Your neighbors are very nice and you get along with them very well, but they are

not very clean. You and your family have always to clean the stairs and the

hallway. One day you run into several of you neighbors when going down the

stairs. You talk to them and nicely and tell them that it would be good that not

only you and your family but also, they cleaned the stairs and the hallway, to keep

everything clean.’

b. Mira que nosotros hemos limpiado ya varias veces el portal, y que creo que, si

no os importa que lo deberíamos partir entre todos, si no os importa a vosotros

también limpiar de vez en cuando, nos vamos turnando.

‘Look, we have cleaned several times the hallway and I was thinking, if you don’t

mind, that we all should clean, if you guys don’t mind cleaning as well

sometimes, we can take turns.’

A praise context is included in (31a), and the answer given by a speaker is included in (31b).

(31) a. Publicas una foto en Facebook y la gente de tu Facebook te dice que eres guapísimo. Te sientes halagado y les dices que ellos sí que son guapos.

‘You post a picture of yourself on Facebook. Your Facebook friends tell you that

you are so beautiful. You feel flattered and you tell them that they are the

beautiful ones.’

b. ¡Ay que me estáis sacando los colores! ¡Dejarme, que me estáis sacando los

colores! ¡¡Ustedes sí que sois guapísimos!!

‘Aw, you guys are embarrassing me! Stop, you guys are embarrassing me! You

guys are super beautiful!!’

Additionally, like for the online forced-choice survey, for the pronominal alternation, each token was coded for one linguistic factor, which, in this case, I called person (called verb in the online forced-choice survey analysis). This factor includes infinitive verb, 2PL, or object of preposition in order to determine whether ustedes is more likely to occur as a subject or as object of preposition. As mentioned for the online forced-choice survey, this factor was not included for the verbal alternation, since in this case, the dependent variable is the verb form, 2PL or 3PL, so that with objects of prepositions or infinitives the verb form cannot be determined.

Like for the online and PRESEEA Málaga and the online forced-choice survey, for the data from the online forced-choice survey, separate mixed effects logistic regression models were created with R, for the pronominal alternation and for the verbal alternation (R core team 2015). The social and pragmatic independent variables described above, and the linguistic factor, called verb or person, were included as fixed effects. Speaker was included as a random effect. The step function and the random forest in R were used to determine the ordering of factors. Then, logistic

94 regression models were developed using the glmer function, and Anova was used to compare the models. Conditional inference trees were used to determine predictor interactions (Tagliamonte and Baayen 2012).

Chapter 4: Results

4.1. Online and PRESEEA Data Results

4.1.1. Ustedes/Vosotros Alternation

The tokens for the pronominal alternation were coded for nine different independent variables, as described in section 3.1.2. These are: region, priming, intervening material between subject and verb, overt/null pronoun, frequency of the lexeme, clause type, placement, tense-mood-aspect (TMA), and special online features.

A conditional inference tree is provided first to determine the main splits in these data and to justify later collapsing of the data; this procedure allows for a more balanced distribution or when there is a lack of sufficient tokens for some categories.

Furthermore, there are factors that are shown to be very important in the conditional inference trees and the random forests but which are not significant in the regression model; this is due to the fact that they have an effect only in interaction with other factors, as shown in the inference tree. Figure 11 shows this conditional inference tree for the pronominal alternation. The dark area in the inference tree corresponds to ustedes rates and the rest is vosotros rates. As can be observed in the conditional inference tree, the most important split in the data is that of region, more precisely, between Central and Western Andalusia. In Western Andalusia, the most important predictor is online features, and there is a split between neutral online special features or absence of them on one hand, and, on the other hand, online special features showing positive or negative emotions toward the interlocutors, i.e. non-neutral online features. Within neutral or absence of special online features, there is a split between a pre-verbal placement of the subject pronoun on one hand and post-verbal placement

96 or null subject pronoun on the other hand. There is a higher rate of ustedes over vosotros in Western Andalusia, with non-neutral online features, and with pre-verbal placement as opposed to post-verbal and null subject when there are neutral online features or there aren’t online features. In Central Andalusia, there are just fewer occurrences of ustedes than in Western Andalusia.

Neutral or absence

Non-neutral Pre-verbal Post-verbal or null

Figure 11. Conditional inference tree. Ustedes/vosotros alternation. Online and

PRESEEA data.

Table 7 contains the best-fit model for the pronominal alternation before collapsing the categories. This analysis is also included to show the proper collapsing of categories conducted next to achieve a more balanced distribution. The reference value of the dependent variable is ustedes; a positive estimate means a higher probability of ustedes over vosotros, and a negative estimate means the opposite. As

97 shown in Table 7, region and online features are the most important predictors included in the best-fit model, as selected by the step function. There is not a significant difference between online special features showing positive feelings toward the interlocutors and online special features showing negative feelings toward the interlocutors. On the other hand, neutral online features and absence of online features are significantly different from online features showing negative emotions toward the interlocutors, more precisely, absence of online features or neutral online features have an inhibiting effect on ustedes over vosotros as opposed to online features showing negative emotions toward the interlocutors. After releveling the category online features, I found that neutral or absence of online features are also significantly different from online features showing positive feelings toward the interlocutors; more precisely, neutral or absence of online features have an inhibiting effect on ustedes over vosotros as opposed to presence of positive online features.

Although there is a significant split between Central and Western Andalusia, as shown in the inference tree in Figure 11, the difference between Western and

Central Andalusia was not shown to be significant in the best-fit model. This is due to the presence of interactions between factors as shown in Figure 11; more precisely it is due to the fact that there is a higher rate of ustedes in Western Andalusia than in

Central Andalusia but only in interaction with non-neutral online features and pre- verbal placement, which are only relevant for Western Andalusia. The analysis was done including the factor placement and not overt pronoun, intervening material, or priming, because all these include a category coded for null subjects. Analyses including each of these other factors separately, i.e. overt pronoun, intervening material, and priming, didn’t yield significant best-fit models.

Estimate SE z-value p-value (Intercept) -9.30748 0.51148 0.51148 < 2e-16 Region (reference level is Central) Western 0.70172 0.54570 1.286 0.198476 Online features (reference is negative) Positive online features -0.04993 0.55680 -0.090 0.928546 Online features (reference is negative) Neutral online features -4.47608 1.61855 -2.765 0.005684 Online features (reference is negative) Absence of online features -1.13883 0.33878 -3.362 0.000775 Table 7. Best-fit model. Ustedes/vosotros alternation. Online and PRESEEA data.

Due to the splits observed in the data as shown in the conditional inference tree in

Figure 11, namely, a higher rate of ustedes over vosotros with with pre-verbal placement as opposed to post-verbal and null subject, the factor placement was collapsed into a two-way distinction, i.e. pre-verbal as opposed to post-verbal or null.

In Figure 11, it is also shown that there is a split between neutral online special features or absence of these on one hand, and, on the other hand, online special features showing positive or negative emotions toward the interlocutors. Furthermore, the results shown by the best-fit model in Table 7 indicate that there is not a significant difference between online special features showing positive feelings toward the interlocutors and online special features showing negative feelings toward the interlocutors, and that that neutral or absence of online features are significantly different from online features showing positive feelings toward the interlocutors and from online special features showing negative feelings toward the interlocutors.

Therefore, online features were collapsed also into a two-way distinction, i.e. neutral

99 or absence of online features as opposed to positive or negative online features. This achieves a more balanced distribution, since there was poor data distribution through the categories for online features, with lack of sufficient tokens for some levels. More precisely, before collapsing, absence of online features comprises 58.6% of all the data and online features showing negative emotions toward the interlocutors comprise

31.2% of the data, whereas online features showing positive emotions toward the interlocutors comprise less than 10% of all the data, and neutral online features also comprise less than 10% of all the data. Because of lack of sufficient tokens for coordinate clauses, the factor clause type was also collapsed into a two-way distinction, including main and coordinate clauses on one hand, and subordinate clauses on the other. Main and coordinates have common characteristics, more precisely, they both can appear as independent sentences, whereas subordinates cannot. Whereas coordinate clauses comprised 11.5% of the data, subordinates comprised 38.5%, and main clauses 50.1%. Due to poor data distribution, the factor tense-mood-aspect was also collapsed into a two-way distinction, including indicatives as opposed to non-indicatives. Whereas indicatives comprised 81.6% of all the data, each of the rest of categories, i.e. conditionals, imperatives, and subjunctives, didn’t reach 10% of the data.

After collapsing, as can be observed in the conditional inference tree in Figure 12, the most important split is again that of region, between Central and Western

Andalusia, and similar tendencies are observed, with a higher rate of ustedes as opposed to vosotros with non-neutral online features than when there are not online features or these are neutral in Western Andalusia, and a higher rate of ustedes when

100 there is a pre-verbal placement of the subject pronoun than when there is a post-verbal placement of the subject pronoun or the subject is null, in Western Andalusia.

Neutral or absence of online features

Post-verbal Pre-verbal or null Non-neutral

Figure 12. Conditional inference tree. Ustedes/vosotros alternation after collapsing.

Online and PRESEEA data.

The random forest in Figure 13 shows the most important predictors in hierarchical order in terms of their statistical significance, after collapsing. As can be observed in Figure 13, the most important predictor for the pronominal alternation is region, followed by online special features, placement, and clause type, i.e. main or coordinate and subordinates. Frequency and TMA are not significant predictors, as indicated by their placement to the left of the zero in the random forest. The analysis was done including the factor placement and not overt pronoun, intervening material, or priming, because all these include a category coded for null subjects.

101

Figure 13. Random forest. Ustedes/vosotros alternation. Online and PRESEEA data.

The best-fit regression model for the ustedes/vosotros alternation including the factor placement and not priming, overt pronoun or intervening material is shown in

Table 8. This regression was conducted after collapsing the categories placement, online features, clause type and TMA. Analyses including the factors overt pronoun, intervening material, and priming, separately, didn’t yield a significant best-fit model.

102

Estimate SE z-value p-value (Intercept) -9.5315 0.5101 -18.684 <2e-16 Region (reference level is Central) Western 0.7912 0.5426 1.458 0.144754 Online features (reference is non-neutral) Absence or neutral -1.2249 0.3172 -3.862 0.000113 Placement (reference is post-verbal or null) Pre-verbal placement 0.5998 0.2844 2.109 0.034927

Table 8. Best-fit regression model. Ustedes/vosotros alternation. Online and

PRESEEA data.

Absence of online special features or neutral online features have an inhibiting effect of ustedes over vosotros as opposed to presence of non-neutral online features, expressing positive or negative feelings toward the interlocutors. On the other hand, a pre-verbal placement of the subject pronoun leads to significantly more use of ustedes over vosotros as opposed to a post-verbal placement of the subject pronoun or a null subject pronoun. The difference between Central and Western Andalusia does not appear as significant in the best-fit model, due to the interactions shown in the conditional inference tree in Figure 12, where the effect of the factors placement and online features and only relevant for Western Andalusia; more precisely, there is a higher rate of ustedes in Western Andalusia than in Central Andalusia but only in interaction with non-neutral online features and pre-verbal placement.

Figure 14 illustrates how, whereas the rate of vosotros increases with post-verbal or null placement, the rate of ustedes decreases, so that the difference between both variants becomes greater with post-verbal placements or null subjects. More precisely,

103 tokens of the variant ustedes comprise 35.2% of all the post-verbal or null tokens, whereas tokens of the variant vosotros comprise 64.8% of all the post-verbal or null subjects, which means that the rate of post-verbal or null vosotros is about double the rate of post-verbal or null ustedes. The differences are significant (χ2 = 8.86; degrees of freedom = 1; probability = 0.003).

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Pre-verbal placement Post-verbal placement or null subject

Ustedes Vosotros

Figure 14. Rates of ustedes and vosotros with pre-verbal placement and post-verbal or null placement. Online and PRESEEA data.

Figure 15 illustrates how, whereas the rate of vosotros increases with neutral or absence of online features, the rate of ustedes decreases, so that the difference between both variants becomes greater with neutral or absence of online features.

More precisely, tokens of the variant ustedes comprise 34.4% of all the tokens co- occurring with neutral or absence or online features, whereas tokens of the variant vosotros comprise 65.6% of all the tokens co-occurring with neutral or absence or online features tokens, which means that the rate of vosotros co-occuring with neutral

104 or absence or online features is about double the rate of ustedes co-occuring with neutral or absence or online features. The differences are significant (χ2 = 15.6; degrees of freedom = 1; probability = 0.000).

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Non-neutral online features Neutral or absence of online features

Ustedes Vosotros

Figure 15. Rates of ustedes and vosotros with non-neutral and neutral or absence of online features. Online and PRESEEA data.

4.1.1.1. Ustedes/Vosotros Alternation in Western Andalusia vs. Central Andalusia

The predictor region, although it was not significant in the best-fit models (see tables 6 and 7, section 4.1.1.), it was shown as the most important factor by the random forest in Figure 13, and different tendencies are observed for Western

Andalusia as opposed to Central Andalusia on the inference trees in Figures 11 and

12. More precisely, the effects of online features and placement are found only for

Western Andalusia, whereas Central Andalusia just shows fewer occurrences of ustedes than Western Andalusia. Therefore, a separate analysis, more precisely, mixed 105 effects logistic regression, of the data from Cádiz, Seville, and Huelva, i.e. Western

Andalusia, and another of the data from Málaga and Córdoba, i.e. Central Andalusia, were conducted. Figure 16 illustrates how the difference between the rates of ustedes and vosotros is greater in Central Andalusia, with a higher rate of vosotros than ustedes in Central Andalusia. The difference between ustedes and vosotros in Western

Andalusia area is lesser and the rate of ustedes is higher in Western Andalusia than in

Central Andalusia. More precisely, out of 1,292 tokens for Western Andalusia, there is a rate of 57.6% of vosotros and 42.4% of ustedes. Out of 1,116 tokens for Central

Andalusia, there is a rate of 68% of vosotros versus 32% of ustedes, which means that the rate of vosotros is about double the rate of ustedes in Central Andalusia.

100% 90% 80% 68% 70% 57.6% 60%

50% 42.4% 40% 32% 30% 20% 10% 0% Western Andalusia Central Andalusia

Ustedes Vosotros

Figure 16. Ustedes/vosotros rates in Western Andalusia and in Central Andalusia.

Online and PRESEEA data.

The same factors were included for both analyses, i.e. for Western and Central

Andalusia. Except for the factor region, these are the factors included for the whole

106 dataset, collapsed, as explained in section 4.1.1. The conditional inference tree for

Western Andalusia in Figure 17 shows that, as seen before for the whole dataset, in

Western Andalusia, there is a significant split between neutral or absence of online features and negative or positive online features; also, when there is absence or neutral online features, there is a higher rate of ustedes over vosotros when the subject pronoun is placed pre-verbally as opposed to post-verbally or is null.

Neutral or absence

Pre- verbal Non-neutral Post-verbal or null

Figure 17. Conditional inference tree. Ustedes/vosotros alternation. Western

Andalusia. Online and PRESEEA data.

The random forest in Figure 18 shows the strongest predictors for the ustedes/vosotros alternation in Western Andalusia. As depicted in Figure 18, online features, placement and clause type are the most important predictors. Like for the

107 whole dataset, frequency and TMA are not significant predictors, as indicated by their placement, in the random forest, to the left of the zero.

Figure 18. Random forest. Ustedes/vosotros alternation. Western Andalusia. Online and PRESEEA data.

The best-fit regression model for the Western Andalusia data, including the factor placement, and not overt pronoun, priming or intervening material, is shown in

Table 9. In the best-fit model, online features and placement are the significant predictors. Like for the whole dataset, absence of online special features, or presence of neutral online features have an inhibiting effect on ustedes over vosotros as opposed to presence of online features showing positive or negative emotions toward the interlocutors in Western Andalusia. Pre-verbal position of the subject pronoun leads to significantly more use of ustedes over vosotros, as opposed to post-verbal position or null subject in Western Andalusia as well. Analyses including the factors overt pronoun, intervening material, and priming, separately, didn’t yield a significant best-fit model. No significant model was found for Central Andalusia either, because

108 the factors that determine the pronominal variation, i.e. placement and online features, or the other factors, do not have an effect in that area.

Estimate SE z-value p-value (Intercept) -7.6464 0.6739 -11.346 <2e-16 Online features (reference is presence) Absence or neutral -1.5316 0.3579 -4.280 1.87e-05 Placement (reference is post-verbal or null) Pre-verbal 0.8862 0.3284 2.699 0.00696

Table 9. Best-fit regression model for the ustedes/vosotros alternation in Western

Andalusia. Online and PRESEEA data.

4.1.1.2. Ustedes/Vosotros Alternation. Frequent vs. Infrequent Verbs

A separate analysis of frequent versus infrequent verbs was also conducted, following Erker and Guy (2012), who found that frequency has no independent effect but interacts with other factors for variable subject pronoun expression in Spanish.

More precisely, Erker and Guy (2012) argue that frequency magnifies the effect of certain factors which, without frequent forms, don’t emerge as significant. In the current study, there is a total of 1,353 tokens with frequent verb lemmas, and the rate of vosotros is 62.2% and the rate of ustedes is 37.8% with frequent verb lemmas.

An analysis with frequent forms was conducted including the factor placement and not overt pronoun, intervening material, or priming, because all these include a category coded for null subjects. Similar results like for the whole dataset, which included frequent and infrequent verbs, were obtained with frequent verbs. More precisely, absence of online special features or neutral online features have an 109 inhibiting effect of ustedes over vosotros as opposed to presence of non-neutral online features; pre-verbal placement of the subject pronoun leads to significantly more use of ustedes over vosotros. Another analysis with frequent forms was done including the factor priming, and not intervening material, placement or overt pronoun. Table

10 contains the best-fit model for the ustedes/vosotros alternation with frequent verb lemmas including the factor priming. As found before, absence of online special features or neutral online features have an inhibiting effect of ustedes over vosotros as opposed to presence of non-neutral online features. On the other hand, unlike found before, main or coordinate clauses as opposed to subordinate clauses result in a significantly higher probability of employing ustedes over vosotros with the 2PL verb form, when including the factor priming.

Estimate SE z-value p-value (Intercept) -9.8697 0.7023 -14.053 < 2e-16 Region (reference level is Central) Western 1.0596 0.6944 1.526 0.127022 Online feature (reference is non-neutral) Neutral or absence -1.6943 0.4548 -3.725 0.000195 Clause type (reference is subordinate) Main or coordinate 0.7775 0.3950 1.968 0.049025

Table 10. Best-fit model for the ustedes/vosotros alternation. Frequent verb lemmas.

Priming. Online and PRESEEA data.

With the inclusion of the factor intervening material in the analysis instead of placement, overt pronoun, or priming, and also with the inclusion of the factor overt pronoun instead of placement, priming or intervening material, the same results are

110 obtained as with the factor priming, i.e. non-neutral online features and main or coordinate clauses leading to significantly more use of ustedes over vosotros. This is because placement, which is the third most important predictor in the pronominal alternation, after region and online features, obliterates effects of other factors which don’t come into play when placement is included.

With infrequent verb lemmas, there is a total of 1,055 tokens, with 62.7% of vosotros and 37.3% of ustedes. However, like for Central Andalusia, no significant model was found for infrequent verb lemmas. Therefore, the effect of the factors placement, online features, and clause type is deactivated with infrequent forms.

4.1.1.3. Conclusion

Summarizing, pre-verbal position of the subject pronoun, co-occurrence of non-neutral special online features in Western Andalusia, and main and coordinate clauses with frequent verb lemmas lead to significantly more use of ustedes over vosotros with the 2PL verb form; the effect of these factors is not present with infrequent verbs only. On the other hand, in Central Andalusia there are less occurrences of ustedes than in Western Andalusia. In (32a), ustedes is in pre-verbal position, in a main clause, and with co-occurrence of an online feature showing anger,

. In (32b) vosotros is in post-verbal position with no co-occurring online features, in a neutral context, and in a subordinate clause.

(32) a. Ustedes dos sobre favoritismos sois los que más tenéis

callar .

‘Especially you [ustedes] two are [2PL]. the ones who can’t have a word

111

about favoritisms.’

b. la fecha y el lugar que le pongáis vosotros.

‘the date and place that you guys [vosotros] choose [2PL].’

4.1.2. 2PL/3PL Verb Form Alternation

For the verbal alternation, ten independent variables were included, as described in section 3.1.2. These are: region, priming, intervening material, placement, overt/null pronoun, frequency, clause type, tense-mood-aspect (TMA), special online features, and semantic classes of verbs. TMA was collapsed into a two- way distinction, including imperatives versus non-imperatives, since I found that it was the most important split in the data in a previous study, with imperatives leading to more use of the 3PL (Jaime Jiménez 2014). There was also a poor distribution of the data throughout the different categories of TMA, with indicatives comprising most of the data, and a lack of sufficient tokens for non-indicatives, i.e. conditionals and subjunctives, not reaching 10% of all the data, and imperatives comprising 14.2% of the data. A conditional inference tree is included first, in Figure 19, to determine the main splits in the data, and, thus, further collapsing to be made in order to achieve a more balanced distribution or when there is a lack of sufficient tokens for some categories. Furthermore, there are factors that are shown as very important in the conditional inference trees and the random forests but which are not significant in the regression model; this is due to the fact that they have an effect only in interaction with other factors, which is shown in the inference trees. In the inference tree, the dark area corresponds to 3PL rates and the rest to 2PL rates. As can be observed in the

112 conditional inference tree in Figure 19, the rate of the 3PL verb forms is higher with post-verbal placement as opposed to pre-verbal or null placement. Therefore, placement was collapsed into a two-way distinction, including post-verbal placement of ustedes as opposed to pre-verbal placement of ustedes or null subject.

Imperative Non-imperatives Pre-verbal or Pre-verbal null Dynamic or null Post-verbal

Mental Dynamic & Mental & stative Mental stative

Figure 19. Conditional inference tree. 2PL/3PL verb form alternation. Online and

PRESEEA data.

Due to poor data distribution throughout the categories of the factor online features, this was also collapsed into a two-way distinction, including, like in the pronominal alternation, neutral online features or absence of online features versus non-neutral online features. Online features showing negative emotions toward the interlocutors didn’t reach 10% of the data and neutral online features comprised didn’t reach 10% of the data either; online features showing positive emotions toward the interlocutors comprised 17.4%, whereas absence of online features comprised 72.3%.

Verb semantics was also collapsed due to poor data distribution throughout the

113 categories. Cognitive or mental verbs comprised a 15.1% of the data, whereas dymanic, external verbs comprised 43.7% of the data and stative verbs comprised

41.2% of the data. There are different splits of semantic classes of verbs, as observed in the conditional inference tree in Figure 19; however, mental verbs were collapsed with stative verbs as opposed to dynamic external activity verbs, because both, mental and stative, are non-external, and because there is a higher rate of 3PL verb forms with mental and stative verbs in the imperative mood as seen on the right side of the conditional inference tree in Figure 19. Furthermore, in a previous research I found that mental and stative verbs lead to significantly more use of the 3PL (Jaime Jimenez

2014). Due to lack of sufficient tokens for coordinate clauses, the factor clause type was also collapsed into a two-way distinction, including main and coordinate clauses as opposed to subordinate clauses. Coordinate clauses comprised 11.2% of the data, whereas subordinates comprised 32.2%, and main clauses 56.6%.

After collapsing, as illustrated in the conditional inference tree in Figure 20, with non-imperative verb forms, there is a higher rate of 3PL forms of the verb with post-verbal placement of ustedes than with pre-verbal placement of ustedes or null ustedes. With imperatives, there is a higher rate of 3PL verb form with post-verbal placement of ustedes and with stative or mental verbs as opposed to dynamic verbs, and to pre-verbal placement of ustedes or null ustedes. Therefore, there is a split between stative or mental verbs on one hand, and dynamic verbs on the other. More generally, the rate of 3PL verb form with ustedes is higher with the imperative mood than with the no-imperative moods, and the effect of placement is the same for imperatives and non-imperatives.

114

Imperatives Non-imperatives

Pre- Pre- verbal Post-verbal verbal or null or null Stative or mental Dynamic

Figure 20. Conditional inference tree for the 2PL/3PL alternation. Factor placement included. Online and PRESEEA data.

The random forest in Figure 21 shows the most important predictors for the verbal alternation including the factor placement. Placement, overt pronoun, and intervening material were analyzed separately since all these categories include a level for null subjects. The results of these separate analyses are included below. As shown in

Figure 21, the most important predictors are TMA, placement, and online features, and then priming, clause type and frequency, in that order in terms of their statistical significance. Region and verb semantics were not selected as significant predictors, as indicated by their placement, in the random forest, to the left of the zero

115

Figure 21. Random forest. 2PL/3PL alternation. Placement. Online and PRESEEA data.

The best-fit regression model for the verbal alternation including the factor placement is shown below. The reference value of the dependent variable is the 3PL verb from; a positive estimate means a higher probability of the 3PL verb form over the 2PL verb form, and a negative estimate means the opposite.

Table 11. Best-fit model. 2PL/3PL alternation. Placement. Online and PRESEEA data.

Estimate SE z-value p-value (Intercept) 3.5643 1.4275 2.497 0.012532 TMA (reference level is imperative) Non-imperatives -10.3726 1.4387 -7.210 5.61e-13 Placement (reference is post-verbal) Pre-verbal or null subject -3.0168 0.5871 -5.139 2.77e-07 Online features (reference is non-neutral)

116

Neutral or absence 2.3443 0.6933 3.382 0.000721 Priming (reference is same prior verb form) Different prior verb form -0.9116 0.5664 -1.609 0.107525 Clause type (reference is subordinate) Main or coordinate 0.2399 0.4684 0.512 0.608496 Frequency (reference is frequent) Infrequent verb lemma 0.9425 0.4627 2.037 0.041670

Regarding the most important predictor in the verbal alternation, i.e. TMA, non- imperatives as opposed to imperatives result in a significantly lower probability of employing the 3PL verb form with ustedes over the 2PL; pre-verbal placement of ustedes or a null ustedes also result in a significantly lower probability of employing the 3PL verb form with ustedes over the 2PL. On the other hand, absence or neutral online features as opposed to non-neutral online features and infrequent verb lemmas as opposed to frequent verb lemmas lead to significantly more use of the 3PL form of the verb with ustedes over the 2PL verb form with ustedes.

The most important predictors in the verbal alternation are first tense-mood-aspect and then placement. Figure 22 illustrates how, with non-imperatives, the 3PL decreases significantly whereas the 2PL increases significantly. More precisely, tokens of the variant 3PL verb form comprise 69.3% of all the imperatives but only

18.8% of all non-imperatives, whereas tokens of the variant 2PL comprise 30.7% of all the imperatives versus 81.2% of all non-imperatives. The differences are significant (χ2 = 189; degrees of freedom = 1; probability = 0.000).

117

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Imperative Non-imperative

3PL 2PL

Figure 22. Rates of 3PL and 2PL with imperatives and non-imperatives. Online and

PRESEEA data.

The second most important predictor in the verbal alternation is placement. Figure

23 illustrates how, with post-verbal placement of ustedes, the 2PL decreases significantly whereas the 3PL increases significantly. More precisely, tokens of the variant 3PL verb form comprise 16.6% of all pre-verbal placement or null subject tokens but 48.1% of all the post-verbal placement tokens. On the other hand, tokens of the variant 2PL comprise 83.4% of all pre-verbal placement or null subject tokens versus 51.9% of all the post-verbal placement tokens. The differences are significant

(χ2 = 126; degrees of freedom = 1; probability = 0.000).

118

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Pre-verbal placement or null subject Post-verbal placement

3PL 2PL

Figure 23. Rates of 3PL and 2PL with pre-verbal or null placement and post-verbal placement of the subject. Online and PRESEEA data.

The conditional inference tree in Figure 24 includes the factor intervening material in the analysis instead of placement or overt pronoun. As can be observed in the conditional inference tree in Figure 24, with non-imperatives, there is a higher rate of the 3PL verb form when there is no intervening material between the subject ustedes and the verb than when there is intervening material between verb and subject or when the subject, ustedes, is null. With the imperative, there is a higher rate of the

3PL verb form with stative or mental verbs when there is no intervening material

119 between the subject ustedes and the verb, than when there is intervening material or the subject, ustedes, is null, and as opposed to dynamic, external verbs.

Non- Imperative imperatives

Stative or mental

Intervening Intervening No intervening or null or null Dynamic

Figure 24. Conditional inference tree. 2PL/3PL alternation. Intervening material.

Online and PRESEEA data.

The random forest in Figure 25 shows the most important predictors for the

2PL/3PL verb form alternation, including the factor intervening material. As shown in the random forest, the ordering of the factors in terms of their statistical significance is tense-aspect-mood, verb semantics, intervening material, and online features, and

120 then, clause type, region, priming, and frequency; all were found significant, as none are placed to the left of the zero in the random forest.

Figure 25. Random forest. 2PL/3PL alternation. Intervening material. Online and

PRESEEA data.

Table 12. Best-fit model for the 2PL/3PL alternation. Intervening material. Online and

PRESEEA data.

Estimate SE z-value p-value (Intercept) 4.6246 1.3997 3.304 0.000953 TMA (reference level is imperative) Non-imperatives -9.1309 1.2798 -7.135 9.71e-13 Verb semantics (reference is dynamic) Non -dynamic -0.0775 0.4381 -0.177 0.859586 Intervening material (reference is presence) Null subject -1.2663 0.6410 -1.976 0.048202 Intervening material (reference is presence) No intervening material 1.6720 0.5099 3.279 0.001042 Online features (reference is non-neutral) Neutral or absence 1.0508 0.5171 2.032 0.042144 Clause type (reference is subordinate) 121

Main or coordinate -0.1187 0.4286 -0.277 0.781798 Region (reference is Central) Western -2.3929 0.6418 -3.729 0.000193 Priming (same prior verb form) Different prior verb form -1.9430 0.5953 -3.264 0.001098 Frequency (frequent verb lemma) Infrequent verb lemma 1.0616 0.4390 2.418 0.015613

Same results as before were obtained regarding TMA, online features and frequency (see best-fit model in Table 11). More precisely non-imperatives have an inhibiting effect on the 3PL verb form with ustedes over the 2PL and neutral online features and infrequent verb lemmas lead to significantly more use of the 3PL verb form over the 2PL. Furthermore, there is an effect of intervening material between ustedes and the verb form; more precisely, when there is a null ustedes, the odds of encountering the 3PL over the 2PL are significantly lower than when there is intervening material between the subject and verb, but when there is no intervening material between subject and verb, the odds of encountering the 3PL over the 2PL are significantly higher than when there is intervening material. On the other hand, after releveling, I found that lack of intervening material between subject and verb leads to significantly more use of the 3PL over the 2PL as opposed to null subject too. There is also an effect of region, more precisely, when the region is Western Andalusia, the odds of encountering the 3PL verb form over the 2PL are significantly lower than when it is Central Andalusia. There is also an effect of priming so that when the prior verb form is different, i.e. when it is a 2PL instead of a 3PL, which means that there is no priming, the odds of encountering the 3PL over the 2PL are significantly lower

122 than when there is priming, i.e. when there is a prior 3PL verb form. In other words, priming leads to a significantly higher probability of 3PL over 2PL. The factor verb semantics is not found significant in the best-fit model in Table 11 even though it was the second most important in the random forest in Figure 25. This is because the factor verb semantics has an effect only in interaction with other factors, more precisely, in interaction with TMA and intervening material, as shown in the inference tree in Figure 24.

The conditional inference tree in Figure 26 includes the factor overt pronoun instead of placement or intervening material. As illustrated in Figure 26, the rate of

3PL verb form is higher with non-imperative when the subject pronoun ustedes is overt as opposed to null. With imperative verb-forms, the rate of 3PL verb form is higher when the subject pronoun ustedes is overt and the verb is non-dynamic, non- external, i.e. it is a mental or stative verb, as opposed to null subjects and as opposed to a dynamic, external verb.

Imperative Non-imperative

Stative or mental

Dynamic Null Overt Overt Null

Figure 26. Conditional inference tree. 2PL/3PL alternation. Overt pronoun. Online and PRESEEA data.

123

The random forest in Figure 27 shows the most important predictors including overt pronoun in the analysis, instead of placement or intervening material. As can be observed in the random forest, the most important predictors are TMA, verb semantics, and overt pronoun, and then clause type, region, online special features, frequency and priming, in that order, and all were found significant since none of them are placed to the left of the zero in the random forest.

Figure 27. Random forest for the 2PL/3PL alternation. Overt pronoun. Online and

PRESEEA data.

Table 13. Best-fit model. 2PL/3PL alternation. Overt pronoun. Online and PRESEEA data.

Estimate SE z-value p-value

(Intercept) 2.4265 1.2118 2.002 <0.05 TMA (reference level is imperative) Non-imperatives -8.3080 1.2256 -6779 <0.0001 Verb semantics (reference is dynamic) Non -dynamic 0.4585 0.3902 1.175 0.23996

124

Overt/null pronoun (reference is null) Overt pronoun 2.4996 0.5453 4.584 <0.0001 Clause type (reference is subordinate) Main and coordinates -0.8593 0.4016 -2.140 <0.05 Region (Central) Western -0.7835 0.5567 -1.405 0.15990 Online features (reference is non-neutral) Neutral or absence 1.1433 0.4672 2.447 <0.01 Frequency (reference is frequent) Infrequent verb lemmas 0.5768 0.3833 1.505 0.13234 Priming (reference is prior same verb

form) Different prior verb form -1.8439 0.5613 -3.285 <0.01

Same results as before were obtained for TMA, online features, and priming (see best-fit models in Tables 10 and 11), with non-imperatives and different prior verb form, i.e. lack of priming, having an inhibiting effect on the 3PL verb form with ustedes over the 2PL, and absence of online features leading to significantly more use of the 3PL verb form over the 2PL. On the other hand, the odds of encountering the

3PL verb form over the 2PL verb form are significantly higher when the subject pronoun ustedes is overt than when it is null. Main and coordinate clauses as opposed to subordinates have an inhibiting effect on the 3PL verb form over the 2PL verb form. The factor verb semantics is not found significant in the best-fit model in Table

12 even though it was the second most important in the random forest in Figure 27 because it has an effect only in interaction with other factors, more precisely, in interaction with TMA and overt/null pronoun as shown in the inference tree in Figure

26.

125

Furthermore, factors which didn’t have an effect when including placement, are significant when not including it, more precisely, priming and region when including intervening material, and priming and clause type when including overt/null pronoun.

This is because placement, which is the second most important predictor in the verbal alternation, after TMA, obliterates effects of other factors which don’t come into play when placement is included.

4.1.2.1. 2PL/3PL Verb Form Alternation in Western Andalusia

Different tendencies were found before for Western and Central Andalusia for the pronominal alternation (see Section 4.1.1.), and Western Andalusia has an inhibiting effect on the 3PL over the 2PL as opposed to Central Andalusia in the verbal alternation (see best-fit model in Table 11, Section 4.1.2.). Therefore, a separate analysis of Western and Central Andalusia was conducted as well for the verbal alternation. There is a total of 690 tokens for the verbal alternation in Western

Andalusia, with a rate of 75.2% of 2PL verb forms and 24.8% of 3PL verb forms.

The results of logistic regression yielded similar best-fit models to the whole dataset (see Section 4.1.2.) for Western Andalusia, with the imperative mood, absence of non-neutral online features, post-verbal placement, no intervening material between subject ustedes and verb, same prior verb form, i.e. priming, overt ustedes, and infrequent verb lemmas leading to significantly more use of the 3PL verb form over the 2PL verb form with ustedes. For the most important predictor, i.e. TMA, tokens of the variant 3PL form of the verb with ustedes comprise 65.2% of all imperatives but only 17% of all non-imperatives in Western Andalusia. Conditional

126 inference trees including the different variables that had to be analyzed separately, i.e. placement, intervening material, and overt/null pronoun, are included below, to show the differences with respect to the whole dataset.

Figure 28 shows the conditional inference tree for the 2PL/3PL alternation in

Western Andalusia including the factor placement. As can be observed in Figure 28, there is a higher rate of 3PL verb form with non-imperatives when the subject ustedes is placed post-verbally and there are not online special features or these are neutral as opposed to presence of non-neutral online features and as opposed to a pre-verbal placement or null subject. This interaction, i.e. of non-imperatives and placement with online features, was not shown before, for the whole dataset (see conditional inference tree in Figure 20, Section 4.1.2.). With the imperative mood, like for the whole dataset, there is a higher rate of 3PL verb form with post-verbal placement and stative or mental verbs as opposed to dynamic, external verbs, and as opposed to pre-verbal placement or null subject.

Imperatives Non-imperatives

Post-verbal Neutral or Pre- Post-verbal absence verbal or Mental null Non- or stative Dynamic neutral

Figure 28. Conditional inference tree for the 2PL/3PL alternation. Western Andalusia.

Placement. Online and PRESEEA data.

127

The conditional inference tree in Figure 29 shows that including the factor intervening material instead of placement or overt pronoun, in Western Andalusia, there is also a split between imperatives and non-imperatives. However, unlike for the whole dataset (see conditional inference tree in Figure 20, Section 4.1.2.), in Western

Andalusia there is an interaction of mood and online features; more precisely, with non-imperatives, there is a higher rate of 3PL verb form when there are no online features or these are neutral, as opposed to presence of non-neutral online features. On the other hand, like for the whole dataset, there is an interaction of the imperatives with verb semantics; more precisely, with the imperative mood, there is a higher rate of 3PL verb form with stative or mental verbs as opposed to dynamic, external verbs, when there is no intervening material between the subject ustedes and the verb form, as opposed to when there is intervening material or the subject, ustedes, is null.

Imperative Non-imperatives

Stative or mental Non- Neutral or neutral absence Intervening or null No Dynamic

Figure 29. Conditional inference tree. 2PL/3PL alternation. Western Andalusia.

Intervening material. Online and PRESEEA data.

128

As shown in the conditional inference tree in Figure 30, with the factor overt pronoun, there is a split between imperatives and non-imperatives.With non- imperatives, there is again a higher rate of 3PL verb form with presence of non- neutral online features. Like for the whole dataset, with the imperative mood, there is a higher rate of 3PL verb form with stative or mental verbs when ustedes is overt.

More generally, the rate of 3PL verb forms is higher with the imperative mood than with the non-imperative mood, like for the whole dataset (see conditional inference tree in Figure 20 on Section 4.12.).

Imperatives Non-imperatives

Stative or mental Neutral or absence

Dynamic Null Non-neutral Overt

Figure 30. Conditional inference tree. 2PL/3PL alternation. Western Andalusia. Overt pronoun. Online and PRESEEA data.

129

4.1.2.2. 2PL/3PL Verb Form Alternation in Central Andalusia

From Central Andalusia, there is a total of 480 tokens for the verbal alternation, with a rate of 72.3% of 2PL verb forms and 27.7% of 3PL verb forms. For

Central Andalusia, the results from logistic regression, from the best-fit model, indicate that only the factor TMA is significant, with the imperative mood resulting in a significantly higher probability of employing the 3PL over the 2PL verb form than the non-imperative mood. More precisely, tokens of the variant 3PL comprise 77.8% of all imperatives but only 21.4% of all non-imperatives in Central Andalusia.

Conditional inference trees including the different variables that had to be analyzed separately, i.e. placement, intervening material, and overt/null pronoun, are included below, to show the differences with respect to the whole dataset and with respect to

Western Andalusia.

The conditional inference tree in Figure 31 shows the interactions for Central

Andalusia, including the factor placement. Unlike for Western Andalusia (see Section

4.1.2.1.), the effect of verb semantics and online features is not present in Central

Andalusia. The tree shows that, with the imperative mood, there is a higher rate of the

3PL verb form with post-verbal placement as opposed to pre-verbal placement of ustedes or null subject. With non-imperative moods, there is also a higher rate of 3PL verb form with post-verbal placement of ustedes. The rate of 3PL verb form is higher with imperative moods more generally.

130

Imperative Non-imperatives

Post-verbal Pre-verbal or Post-verbal Pre-verbal or null null

Figure 31. Conditional inference tree. 2PL/3PL alternation. Central Andalusia.

Placement. Online and PRESEEA data.

As shown in Figure 32, including the factor intervening material, with non- imperative moods, there is a higher rate of the 3PL verb form with non-intervening material as opposed to intervening or null subject ustedes in Central Andalusia, like for the whole dataset (see conditional inference tree in Figure 24 on Section 4.1.2.).

On the other hand, unlike for Western Andalusia (see conditional inference tree in

Figure 28, Section 4.1.2.1.), there is no interaction with online features in Central

Andalusia. Furthermore, unlike for the whole dataset, there is no interaction between verb semantics and intervening material, but instead, with the imperative mood, there is a higher rate of 3PL verb form when the verb is stative or mental as opposed to dynamic, external.

131

Imperative Non-imperatives

Stative or Intervening Dynamic mental or null No

Figure 32. Conditional inference tree. 2PL/3PL alternation. Central Andav lusia.

Intervening material. Online and PRESEEA data.

In Central Andalusia, including the factor overt pronoun, as shown in Figure

33, there are no interactions of overt pronoun with verb semantics, unlike for the whole dataset (see conditional inference tree in Figure 26, Section 4.1.2.), and there is no effect of online features, unlike for Western Andalusia (see conditional inference tree in Figure 30, Section 4.1.2.1). The conditional inference tree in Figure 33 illustrates how there is a higher rate of the 3PL verb form with overt pronouns as opposed to null pronouns in Central Andalusia, with non-imperatives, like for the whole dataset (see conditional inference tree in Figure 26, Section 4.1.2.). There is also a higher rate of 3PL verb form with imperatives when the verb is stative or mental as opposed to dynamic, external, but stative or mental verbs are not in

132 interaction with overt pronoun, unlike for the whole dataset (see conditional inference tree in Figure 26, Section 4.1.2.).

Non-imperatives Imperatives

Stative or Dynamic mental Null Overt

Figure 33. Conditional inference tree. 2PL/3PL alternation. Central Andalusia. Overt pronoun. Online and PRESEEA data.

4.1.2.3. 2PL/3PL Verb Form Alternation with Frequent Verbs

Like for the pronominal alternation, separate analyses of frequent versus infrequent verb forms were conducted for the verbal alternation. This was done following Erker and Guy (2012), who claim that frequency has no independent effect, and based on the effect of frequency found in the present dissertation for the whole dataset and for Western Andalusia (see best-fit models in Tables 10 and 11 on Section

4.1.2., and Section 4.1.2.1, respectively), in which infrequent verb lemmas lead to significantly more use of the 3PL verb form over the 2PL verb form with ustedes.

There is a total of 539 tokens with infrequent verb lemmas, with a rate of 69.6% of

2PL verb forms and 30.4% of 3PL verb forms. With frequent verb lemmas, there is a 133 total of 631 tokens for the verbal alternation, with a rate of 77.8% of 2PL verb forms, and a rate of 22.2% with the 3PL verb form. The difference in rates between 3PL and

2PL rates are greater with frequent verb lemmas than with infrequent verb lemmas, as illustrated in Figure 34. In Figure 34, it is shown how the rate of 3PL increases with infrequent verbs, whereas the rate of 2PL decreases. The differences are significant

(X2 = 10.3; degrees of freedom = 1; probability = 0.001).

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Frequent Infrequent

2PL 3PL

Figure 34. 2PL/3PL rates with frequent and infrequent verbs. Online and PRESEEA data.

With infrequent verb lemmas, same interactions and splits like for the whole dataset were found (see conditional inference trees in Figures 20, 24, and 26, Section

4.1.2.). On the other hand, with frequent verb lemmas, including the factor placement, the effect of verb semantics is not present in the conditional inference tree, unlike for the whole dataset (see conditional inference tree in Figure 20, Section 4.1.2.); the effect of online features is not present either, unlike for Western Andalusia (see conditional inference tree in Figure 28, Section 4.1.2.1). As shown in the conditional 134

Non-imperative inference tree in Figure 35, there is a higher rate of the 3PL with the imperative mood, in post-verbal position.

Figure 35. Conditional inference tree. 2PL/3PL alternation. Frequent verbs.

Placement. Online and PRESEEA data.

As shown in Figure 36, including the factor intervening material, with frequent verb lemmas, there is a higher rate of 3PL verb form with the imperative mood, with stative or mental verbs, when there is no intervening material between ustedes and the verb, as opposed to presence of intervening material, null subject, dynamic, external verbs, and non-imperatives. Unlike for the whole dataset or for infrequent lemmas

(see conditional inference tree in Figure 24, Section 4.1.2.), with frequent lemmas, with non-imperatives, there are no interactions with intervening material.

135

Imperative

Stative or mental Stative or mental Dynamic Non- Dynamic Intervening or imperatives null No

Figure 36. Conditional inference tree for the 2PL/3PL alternation. Frequent verbs.

Intervening material. Online and PRESEEA data.

As shown in Figure 37, including the factor overt pronoun, with frequent verb lemmas, there is a higher rate of 3PL verb form with the imperative mood with stative and mental verbs, but interactions with overt pronoun are not present, unlike for the whole dataset (see conditional inference tree in Figure 26, Section 4.1.2.); interactions with online features are not present either, unlike for Western Andalusia (see conditional inference tree in Figure 30, Section 4.1.2.1).

136

Imperative Imperative No-imperatives Non- Dynamic Stative or mental imperatives

Figure 37. Conditional inference tree for the 2PL/3PL alternation. Frequent verbs.

Overt pronoun. Online and PRESEEA data.

Best-fit models with infrequent verb lemmas showed similar results to those for the whole dataset, with post-verbal placement, imperative mood, lack of intervening material, absence of non-neutral online features, and overt subject ustedes leading to significantly more use of the 3PL verb form over the 2PL verb form. On the other hand, priming is activated with frequent verb lemmas and not with infrequent verb lemmas. With frequent verb lemmas, the significant factors are TMA and priming, with the imperative mood and same prior verb form leading to significantly more use of the 3PL verb form over the 2PL.

137

Estimate SE z-value p-value (Intercept) -7.1697 -4.293 2.002 -1.76e-05 TMA (reference level is imperative) Non-imperatives -2.6046 1.1883 -2.192 0.0284 Verb semantics (reference is dynamic) Non -dynamic 0.4421 0.6717 0.658 0.5104 Overt/null pronoun (reference is null) Overt pronoun 1.0641 0.8558 1.243 0.2137 Priming (reference is prior same verb form) Different prior verb form -1.5614 0.7912 -1.973 0.0484 Table 14. Best-fit model. 2PL/3PL alternation. Frequent verbs. Overt pronoun. Online and PRESEEA data.

Estimate SE z-value p-value (Intercept) -7.1239 1.6539 -4.307 1.65e-05 TMA (reference level is imperative) Non-imperatives -2.5867 1.1770 -2.198 0.0280 Verb semantics (reference is dynamic) Non -dynamic 0.3520 0.6817 0.516 0.6056 Intervening material (reference is null) No Intervening material 1.2433 0.8852 1.404 0.1602 Intervening material (reference is null) Intervening material 0.670 1.0448 0.581 0.5613 Priming (reference is prior same verb form) Different prior verb form -1.5925 0.7892 -2.018 0.0436 Table 15. Best-fit model. 2PL/3PL alternation. Frequent verbs. Intervening material.

Online and PRESEEA data.

138

4.1.2.4. Conclusion

Summarizing, Central Andalusia, presence of neutral online features or absence of these in Western Andalusia, the imperative mood, post-verbal placement, no intervening material between subject ustedes and verb, priming with frequent verbs, subordinate clauses, overt ustedes, and infrequent mental or stative verbs lead to significantly more use of the 3PL verb form over the 2PL verb form with ustedes.

Ustedes plus the 3PL verb form is actually limited to specific lexical items which include a few infrequent stative or mental verb lemmas. More precisely, the infrequent lemma most frequently occurring in the 3PL verb form is juzgar ‘to judge,’ with twenty occurrences with the 3PL, and only one with the 2PL; other infrequent lemmas also occurring with the 3PL, in five or more instances, are; disfrutar ‘to enjoy,’ with eight occurrences with the 3PL, perdonar ‘to forgive’ and mirar ‘to look,’ ‘to consider,’ with six occurrences, and opinar ‘to give an opinion,’ ‘to consider,’ with five occurences. Juzgar occurs in examples such as (34), extracted from the data collected for this dissertation. The use of disfrutar is illustrated in (34), perdonar is illustrated in (35); mirar is used in the sense of ‘to consider’ in example

(36); opinar use is illustrated in (37):

(33) Estaba bueno? Juzgen ustedes mismos

‘Was it good? Judge [3PL] it you guys [ustedes] for yourselves.’

(34) Que maravilla. Disfruten ustedes.

‘How wonderful. Enjoy [3PL] it you guys [ustedes].’

(35) PERDONEN USTEDES ;-; ES QUE QUEZO ES MALO

‘FORGIVE [3PL] ME YOU GUYS [USTEDES]; BUT QUEZO IS BAD.’

139

(36) Pues no, miren ustedes, esto es una mierda como un fauno de grande.

‘No, you [ustedes] know [3PL], this is a shit as big as a faun.’

(37) Por lo visto se han vendido en 3 sitios las entradas, las mismas

localizaciones tres veces... Opinen ustedes.

‘It looks like tickets have been sold in three different places, the same tickets three

times. Give [3PL] your opinion, you guys [ustedes].’

Example 38 illustrates the results for the verbal alternation. In (38a), the 3PL verb form occurs with ustedes in post-verbal placement, in the imperative mood, with no intervening material between verb and subject, with a mental infrequent verb, and absence of online features. In (38b), the 2PL verb form occurs with ustedes in pre- verbal placement, in the indicative mood, with intervening material (the clitic lo,

‘him’) between subject and verb, with a dynamic, external activity frequent verb, and presence of special online features, i.e. repetition of question marks.

(38) a. Noticia del AS, juzguen ustedes mismos...

‘It is news from the AS [Spanish sports newspaper], you guys [ustedes]

judge [3PL] it.’

b. Ustedes lo habéis visto jugar con Boca, cuando el Valencia lo fichó por

18-20 millones??

‘Have [2PL] you guys [ustedes] seen him playing with Boca

[Argentinian soccer team], when Valencia [Spanish soccer team]

signed him for 18-20 million??’

140

4.2. Online Forced-Choice Survey Results

4.2.1. Ustedes/Vosotros Alternation

The tokens for the pronominal alternation for the online forced-choice data were coded for seven independent variables, as described in section 3.2.1.2. These are education (low, medium, high), sex (female or male), age (young, adult old), urban/rural origin, region (Western or Central Andalusia), verb (infinitive or 2PL verb form), and pragmatic context, including neutral contexts, anger contexts, apology contexts, criticism/irony contexts, formal contexts, and praise contexts. A conditional inference tree for the online forced-choice survey data is presented first to determine the main splits and interactions in these data and to justify later collapsing of the data conducted for a more balanced distribution, or when there is a lack of sufficient tokens for some categories. Furthermore, there are factors that are shown as very important in the conditional inference trees and the random forests but which are not significant in the regression model; this is due to the fact that they have an effect only in interaction with other factors, which is shown in the inference tree. The conditional inference tree in Figure 38 shows that the main split in is between Western and

Central Andalusia. In Western Andalusia, there is higher rate of ustedes with infinitives than with 2PL, and there is a higher rate of ustedes with 2PL with male speakers as opposed to females. In Central Andalusia, there is a higher rate of ustedes with old speakers. With adult and young speakers, there is a higher rate of ustedes with the 2PL with young speakers of medium level of education as opposed adults and as opposed to high or low education.

141

2PL Adult, young 2PL

High Medium or low Old Infinitive Infinitive

Figure 38. Conditional inference tree. Ustedes/vosotros alternation. Online forced- choice survey.

Due to the splits observed in Figure 38, and due to a lack of sufficient tokens from old participants, the factor age was collapsed into a two-way distinction, including old participants as opposed to non-old participants. Before collapsing, young participants comprised 67.7% of all the data, adult participants comprised

27.9% of all data, and old participants didn’t reach 10% of the data. Education was also collapsed into a two-way distinction, including, as indicated by the splits shown in the conditional inference tree in Figure 18, medium education as opposed to non- medium education. Before collapsing, high education comprised 75.5% of the data, medium education comprised 20% of the data, and low education comprised less than

10% of the data. The split made, where high and medium education are collapsed into one category, as opposed to medium education, is due to the fact that low education participants are only 4.6% of all participants in the online forced-choice survey.

Therefore, the opposition in education is actually between high and non-nigh. 142

The random forest in Figure 39, obtained after collapsing, shows region as the most important factor, and then verb type, i.e. 2PL or infinitive, sex, context, education, and age, in that order. The origin of the speakers as urban or rural was not selected as significant predictor, since it appears to the left of the zero in the random forest.

Figure 39. Random forest. Ustedes/vosotros alternation. Online forced-choice survey.

Table 16. Best-fit model after collapsing. Ustedes/vosotros alternation. Online forced- choice survey.

Estimate SE z-value p-value (Intercept) -3.54628 0.33661 -10.535 <2e-16 Region (reference level is Central) Western 3.44115 0.32201 10.686 <2e-16 Verb (reference is 2PL) Infinitive 0.61555 0.23168 2.657 0.00789 Sex (reference is female) Male 0.60064 0.26843 2.238 0.02525 Context (reference is neutral) Anger 0.45367 0.19790 2.292 0.02188 Context (reference is neutral)

143

Apology 0.67981 0.22744 2.989 0.00280 Context (reference is neutral) Command -0.04710 0.21388 -0.220 0.82571 Context (reference is neutral) Criticism 0.07922 0.17132 0.462 0.64378 Context (reference is neutral) Formal 0.17818 0.25590 0.696 0.48626 Context (reference is neutral) Praise 0.84044 0.19999 4.202 2.64e-05

The odds of encountering ustedes over vosotros are significantly higher in

Western Andalusia than in Central Andalusia, when the verb is infinitive than when is

2PL, and when the speaker is male than when is female. Furthermore, anger, apology, and praise contexts result in a higher probability of employing ustedes over vosotros than neutral contexts. After releveling, I found that anger, apology, and praise contexts result in a higher probability of employing ustedes over vosotros than command contexts, and than criticism contexts too. Praise contexts lead to significantly more use of ustedes over vosotros as opposed to formal and anger contexts too.

4.2.1.1. Ustedes/Vosotros Alternation in Western Andalusia

Due to the different tendencies presented in Central versus Western Andalusia, where there is a higher rate of ustedes in Western Andalusia (see inference trees in

Figures 11 and 12, Section 4.1.1, Figure 16, Section 4.1.1.1, and Figure 38 and best- fit model in Table 15, Section 4.2.1.), Central and Western Andalusia data from the

144 online survey were analyzed separately, like the data from online fora, Twitter, and

PRESEEA; the separate analyses of Central and Western Andalusia were conducted after the collapsing of the categories, as described in section 4.2.1. In Western

Andalusia, there is a total of 2,266 tokens from the online forced-choice survey data, with 43.6% of vosotros and 56.4% of ustedes. The conditional inference tree in Figure

40 for Western Andalusia shows that there is a higher rate of ustedes than vosotros with infinitives, and, with 2PL verb forms, with males as opposed to females.

2PL

Infinitive

Figure 40. Conditional inference tree. Ustedes/vosotros alternation. Western

Andalusia. Online forced-choice survey.

145

The random forest in Figure 41 shows verb, i.e. 2PL or infinitive, as the main predictor, followed by sex, context, and education. Age and urban/rural origin were not selected as significant predictors, as indicated by the random forest, where these appear to the left of the zero.

Figure 41. Random forest. Ustedes/vosotros alternation. Western Andalusia. Online forced-choice survey.

Table 17. Best-fit model. Ustedes/vosotros alternation. Western Andalusia. Online forced-choice survey.

Estimate SE z-value p-value

(Intercept) -0.12742 0.23837 -0.535 0.5930 Verb (reference level is 2PL) Infinitive 0.54049 0.24975 2.164 0.0305 Sex (reference level is female) Male 0.66981 0.31122 2.152 0.0314 Context (reference is neutral) Anger 0.54506 0.21341 2.554 0.0106 Context (reference is neutral) Apology 0.48502 0.24625 1.970 0.0489

146

Context (reference is neutral) Command 0.06394 0.22912 0.279 0.7802 Context (reference is neutral) Criticism 0.11333 0.18233 0.622 0.5342 Context (reference is neutral) Formal -0.05129 0.26416 -0.194 0.8461 Context (reference is neutral) Praise 0.86687 0.21662 4.002 6.29e-05

The best-fit model for Western Andalusia is similar to the model for the whole dataset, with male speakers as opposed to females and infinitives as opposed to 2PL verb forms leading to significantly more use of ustedes over vosotros. Like for the whole dataset, anger, apology and praise contexts as opposed to neutral contexts result in a significantly higher probability of employing ustedes over vosotros in Western

Andalusia. After releveling, I found that praise and anger contexts lead to significantly more use of ustedes over vosotros in Western Andalusia as opposed to command, criticism, and formal contexts as well.

4.2.1.2. Ustedes/Vosotros Alternation in Central Andalusia

In Central Andalusia, there is a total of 977 tokens from the online forced- choice survey, with 83% of vosotros versus 17% of ustedes. Therefore, like for the data from online fora, Twitter, and PRESEEA, the difference between rates of the variants is larger in Central Andalusia than in Western Andalusia, where there is more use of ustedes overall. The conditional inference tree in Figure 42 shows the interactions for the data from Central Andalusia. The inference tree shows that the

147 main split is that of age, with old speakers as opposed to adult and young speakers

(‘adult’ includes adult and young). With old speakers, there is a higher rate of ustedes than vosotros with female speakers as opposed to male speakers. With adult and young speakers, there is a higher rate of ustedes than vosotros with infinitives as opposed to 2PL verb forms. With 2PL verb forms, there is a higher rate of ustedes than vosotros when the speaker has a medium level of education as opposed to high or low.

Adult & young

Infinitive Medium

Figure 42. Conditional inference tree. Ustedes/vosotros alternation. Central

Andalusia. Online forced-choice survey.

148

The random forest in Figure 43 shows age as the main predictor, followed by sex, context, and education. Urban/rural origin was not selected as significant predictor, as indicated by its placement in the random forest to the left of the zero.

Verb, i.e. infinitive or 2PL, was not selected as significant predictor either, whereas in

Western Andalusia it was the main predictor.

Figure 43. Random forest. Ustedes/vosotros alternation. Central Andalusia. Online forced-choice survey.

Table 18. Best-fit model. Ustedes/vosotros. Central Andalusia. Online forced-choice survey.

Estimate SE z-value p-value

(Intercept) -3.59773 0.51949 -6.925 4.35e-12 Age (reference level is adult & young) Old 1.94568 0.89846 2.166 0.030344 Sex (reference is female) Male 0.46537 0.47683 0.976 0.329085 Context (reference is neutral) Anger 0.23305 0.42816 0.544 0.586224 Context (reference is neutral)

149

Apology 1.59373 0.42003 3.794 0.000148 Context (reference is neutral) Command -0.07748 0.43965 -0.176 0.860103 Context (reference is neutral) Criticism 0.06182 0.37740 0.164 0.869885 Context (reference is neutral) Formal 1.33417 0.57141 2.335 0.019550 Context (reference is neutral) Praise 0.85250 0.41778 2.041 0.041295

In Central Andalusia, the odds of encountering ustedes over vosotros are significantly higher when the speaker is old as opposed to adult or young. Unlike for

Western Andalusia, the effect of sex, with more use of ustedes in Western Andalusia among male speakers, is not present in Central Andalusia. Unlike for Western

Andalusia, where anger contexts as opposed to neutral contexts lead to more use of ustedes over vosotros, anger contexts are not significantly different from neutral contexts in Central Andalusia. On the other hand, formal contexts lead to significantly more use of ustedes over vosotros as opposed to neutral contexts in Central

Andalusia. Like for Western Andalusia, apologies and praise result in a significantly higher probability of employing ustedes over vosotros as opposed to neutral contexts.

Furthermore, after releveling, I found that apology, formal, and praise contexts lead to significantly more use of ustedes over vosotros as opposed to command and criticism contexts as well, in Central Andalusia. Apology and formal contexts lead to significantly more use of ustedes over vosotros as opposed to anger contexts too.

150

4.2.1.3. Conclusion

Summarizing, there is more use of ustedes in Western Andalusia than in

Central Andalusia. Furthermore, in Western Andalusia, ustedes is more likely to be produced by male speakers than by female speakers. Infinitive verbs, anger, apology, and praise contexts lead to more use of ustedes over vosotros in Western Andalusia.

Infinitives also lead to more use of ustedes over vosotros in Central Andalusia when the speakers are adult or young. In Central Andalusia, when the speaker is an old female, the odds of encountering ustedes over vosotros are significantly higher.

Apology, praise, and formal contexts lead to more use of ustedes over vosotros in

Central Andalusia. Example (39a) illustrates the use of ustedes with an infinitive form in an apology. (39b), (39c), and (39d) illustrate the use of ustedes in anger, praise, and formal contexts, respectively. In (39d), the participant must imagine himself/herself addressing two full professors, inquiring about their opinion about his child’s future in academia.

(39) a. No, perdonarme ustedes

‘No, you [ustedes] forgive [3PL] me, you guys’

b. ¡¡Y ustedes lo que sois es unos hijos de puta!!

‘And you guys [ustedes] are [3PL] mutherfuckers!!’

c. ¡¡Qué bien cocináis ustedes!!

‘You guys [ustedes] cook [3PL] so well!’

d. ¿qué pensáis ustedes?

‘What do you guys [ustedes] think [3PL]?’

151

4.2.2. 2PL/3PL Alternation

The tokens for the verbal alternation for the online forced-choice data were coded for six independent variables, as described in section 3.2.1.2. These are education (low, medium, high), sex (female or male), age (young, adult old), urban/rural origin, region (Western or Central Andalusia), and pragmatic context, which includes neutral contexts, anger contexts, apology contexts, criticism/irony contexts, formal contexts, and praise contexts. A conditional inference tree for the verbal alternation for the online forced-choice survey data is presented first to determine the main splits and interactions in these data and to justify later collapsing of the data conducted for a more balanced distribution, or when there is a lack of tokens for some categories. Furthermore, there are factors that are shown as very important in the conditional inference trees and the random forests but then, they are not significant in the regression model; this is due to the fact that they have an effect only in interaction with other factors, which is shown in the inference tree. The conditional inference tree in Figure 442 shows that the main split in the data from the online forced-choice survey for the verbal alternation is that of context. With apologies and formal contexts, there is a higher rate of 3PL in Central Andalusia, as can be observed on the left side of the tree. As can be observed on the right side of the tree, with the other contexts, i.e. anger contexts, commands, criticism, neutral and praise contexts, there is a higher rate of 3PL with male old speakers with low or high education, as opposed to females and adult speakers. With young speakers, there is a

2 The text is overwritten in the conditional inference tree, as it was shown by the program, R, due to the quantity and complexity of interactions. This tree is, however used only to see main splits in the data. 152 higher rate of 3PL in rural areas in Central Andalusia, as opposed to urban areas, and as opposed to Western Andalusia.

153

High or low

Medium

Figure 44. Conditional inference tree. 2PL/3PL verb form alternation. Online forced-choice survey.

154

Due to the splits observed in Figure 44, and due to poor distribution of the data

throughout the different levels of the factors age and education, both factors were collapsed

into a two-way distinction. Age was collapsed into a two-way distinction including young

and non-young participants. Before collapsing, young speakers comprised 69.6% of the data

and adult speakers comprised 24.8% of the data, whereas old speakers didn’t reach 10% of

the data. Education was collapsed into medium as opposed to non-medium education, i.e.

high and low education. Before collapsing, high education comprised 74.2% of the data and

medium education comprised 21.6% of the data, whereas low education comprised less than

10% of the data, more precisely, 4.2% of the data.

The random forest in Figure 45, obtained after collapsing, shows context as the most

important predictor, followed by region, sex, age, urban/rural origin, and education, in that

order.

Figure 45. Random forest. 2PL/3PL alternation. Online forced-choice survey.

155

The step function selected context, age, region, and urban/rural origin as significant predictors, and that is the order that yielded a significant model. The best-fit model for the

2PL/3PL alternation including the factors collapsed as explained in section 4.2.2 is shown below.

Estimate SE z-value p-value (Intercept) -1.4178 0.4576 -3.098 0.00195 Context (reference level is neutral) Anger 0.7246 0.4573 1.584 0.11313 Context (reference level is neutral) Apology 2.8458 0.4331 6.571 5.01e-11 Context (reference is neutral) Command -0.1030 0.6360 -0.162 0.87137 Context (reference is neutral) Criticism 0.5936 0.4246 1.398 0.16211 Context (reference is neutral) Formal 4.7391 0.4352 10.890 < 2e-16 Context (reference is neutral) Praise 1.0809 0.4306 2.510 0.01207 Age (reference is adult and old) Young -0.9629 0.2417 -3.984 6.78e-05 Region (reference is Central) Western -1.3125 0.2677 -4.903 9.44e-07 Urban/rural (reference is rural) Urban -0.4837 0.2399 -2.016 0.04377 Table 19. Best -fit model for the 2PL/3PL alternation. Online forced-choice survey.

Apology, formal and praise contexts as opposed to neutral contexts lead to significantly more use of the 3PL over the 2PL verb form with ustedes. After releveling, I found that apology contexts lead to significantly more use of the 3PL over the 2PL verb form with ustedes as opposed to command, criticism, anger, and praise contexts as well, and that 156 formal contexts lead to significantly more use of the 3PL over the 2PL verb form with ustedes as opposed to all the other contexts; lastly, praise contexts result in a significantly higher probability of employing the 3PL over the 2PL than command contexts too. Western

Andalusia as opposed to Central Andalusia, young speakers as opposed to adults and old speakers, and urban origin as opposed to a rural origin result in a significantly lower probability of employing the 3PL over the 2PL verb form.

4.2.2.1. 2PL/3PL Alternation in Western Andalusia

Due to the different tendencies observed, where there is more use of the 3PL over the

2PL verb form with ustedes in Central Andalusia than in Western Andalusia, specifically in formal contexts and apologies, and in rural areas (see best-fit model in Table 18, and conditional inference tree in Figure 44, Section 4.2.2.), separate analyses were conducted of

Central and Western Andalusia, with the factors collapsed as explained in section 4.2.2. In

Western Andalusia, there is a total of 1,373 tokens for the verbal alternation, with 80% of

2PL with ustedes versus 20% of 3PL verb forms with ustedes. The conditional inference tree in Figure 46 shows that in Western Andalusia there is a higher rate of 3PL verb forms in formal contexts, as seen on the right side of the tree, and in apologies, as seen on the left side of the tree. With the other contexts, i.e. neutral, praise, anger, commands, criticism, there is a higher rate of 3PL verb forms with old and adult speakers with medium level of education as opposed to young speakers. There is also a higher rate of 3PL with young speakers from rural areas than with young speakers from urban areas. There is also a higher rate of 3PL in apology contexts than in anger, command, criticism, praise and neutral contexts, as can be observed on the left side of the tree.

157

Old & adult

Medium

Figure 46. Conditional inference tree. 2PL/3PL verb form alternation, Western Andalusia.

Online forced-choice survey.

The random forest in Figure 47 shows that context is the most important predictor, like for the whole dataset, followed by age, and then, urban/rural origin, sex, and education.

Figure 47. Random forest. 2PL/3PL alternation. Western Andalusia. Online forced-choice survey. 158

Estimate SE z-value p-value (Intercept) -2.3416 0.4782 -4.897 9.75e-07 Context (reference is neutral) Anger 0.7952 0.4923 1.615 0.106237 Context (reference is neutral) Apology 2.6575 0.4767 5.574 2.49e-08 Context (reference is neutral) Command -0.4631 0.7578 -0.611 0.541120 Context (reference is neutral) Criticism 0.6000 0.4629 1.296 0.194905 Context (reference is neutral) Formal 4.4701 0.4667 9.577 < 2e-16 Context (reference is neutral) Praise 0.7930 0.4834 1.640 0.100940 Age (reference is adult or old) Young -1.0514 0.2776 -3.788 0.000152 Urban/rural (reference is rural) Urban -0.6722 0.2679 -2.509 0.012092 Table 20. Best-fit model for the 2PL/3PL alternation. Western Andalusia. Online forced- choice survey.

The odds of encountering the 3PL over the 2PL verb form are significantly higher in apology and formal contexts in Western Andalusia, as opposed to neutral contexts. After releveling, I also found that apology contexts lead to significantly more use of the 3PL over the 2PL as opposed to all other context except for formal contexts, which lead to significantly more use of the 3PL over the 2PL in Western Andalusia as opposed to all the other contexts including apologies. Lastly, after releveling, I found that praise contexts result in a significantly higher probability of employing the 3PL over the 2PL form of the verb than neutral and command contexts. Young speakers and an urban origin of the speaker as 159 opposed to adult or old speakers, and to rural origin, result in a significantly lower probability of employing the 3PL over the 2PL verb form in Western Andalusia.

4.2.2.2. 2PL/3PL Alternation in Central Andalusia

In Central Andalusia, there is a total of 302 tokens for the verbal alternation, with

47% of 2PL verb forms and 53% of 3PL verb forms. Therefore, unlike in Western Andalusia, and unlike in previous data, i.e. the online fora, Twitter data, and PRESEEA Málaga data, the rate of 3PL verb forms in Central Andalusia is not lower than 2PL verbs. As can be observed in the conditional inference tree in Figure 48, there is a higher rate of 3PL verb forms in apologies and formal contexts in Central Andalusia, as opposed to the other contexts, i.e. neutral, praise, command, criticism, anger, like in Western Andalusia (see conditional inference tree in Figure 46, Section 4.2.2.1). Unlike in Western Andalusia, with the other contexts, i.e. non-formal and non-apology contexts, there is an interaction with sex, with a higher rate of 3PL with male speakers in general, and with females with medium level of education, as opposed to females with high or low education. In Western Andalusia, the significant interaction was with age, with a higher rate of 3PL with old or adult speakers as opposed to young speakers (see conditional inference tree in Figure 46, Section 4.2.2.1).

160

Medium Medium

Figure 48. Conditional inference tree. 2PL/3PL alternation. Central Andalusia. Online forced- choice survey.

The random forest in Figure 49 shows that context is the most important predictor also for Central Andalusia, followed by age, sex, and education. Urban/rural origin is not selected as significant, since it is placed to the left of the zero in the random forest.

Figure 49. Random forest for the 2PL/3PL alternation. Central Andalusia. Online forced- choice survey.

161

Estimate SE z-value p-value (Intercept) -2.56340 0.41090 -6.239 4.42e-10 Context (reference is neutral) Anger 0.70303 0.45841 1.534 0.12512 Context (reference is neutral) Apology 3.00438 0.43391 6.924 4.39e-12 Context (reference is neutral) Command -0.02227 0.63272 -0.035 0.97192 Context (reference is neutral) Criticism 0.61333 0.42470 1.444 0.14869 Context (reference is neutral) Formal 4.88058 0.43821 11.138 < 2e-16 Context (reference is neutral) Praise 1.13067 0.43112 2.623 0.00873 Age (reference is adult or old) Young -1.41147 0.23382 -6.037 1.57e-09 Table 21. Best-fit model for the 2PL/3PL alternation. Central Andalusia. Online forced- choice survey.

Like in Western Andalusia, apologies and formal contexts result in a significantly higher probability of employing the 3PL over the 2PL as opposed to neutral contexts in

Central Andalusia. As found after releveling, apology contexts result in a significantly higher probability of employing the 3PL over the 2PL verb form than all the other contexts too except for formal contexts, which lead to significantly more use of the 3PL verb form over the 2PL verb form as opposed to all the other contexts including apologies. In Central

Andalusia, praise contexts also lead to significantly more use of the 3PL over the 2PL, as opposed to neutral, and, as found after releveling, as opposed to anger, criticism, and command contexts as well. Like in Western Andalusia, when the speaker is young, the odds

162 of encountering the 3PL over the 2PL are significantly lower than when the speaker is adult or old in Central Andalusia.

4.2.2.3. Conclusion

Summarizing, rural areas, apologies, formal, and praise contexts lead to more use of the 3PL verb form with ustedes over the 2PL, as well as the 3PL verb form is less likely to be used by young speakers than by adult and old speakers. For the pronominal alternation, as expected, there is more use of ustedes over vosotros in Western Andalusia; however, for the verbal alternation, there is more use of the 3PL verb form in Central Andalusia than in

Western Andalusia, contrary to expectations. Furthermore, in Central Andalusia, there is more use of the 3PL by females with medium level of education, and by male speakers.

(40a), (40b), and (40c) illustrate the use of the 3PL verb form with ustedes in formal, apology, and praise contexts, respectively. In (40a), participants had to imagine they were addressing two full professors, inquiring about their opinion on their child’s future in academia.

(40) a. ¿qué piensan ustedes?

‘What do you guys [ustedes] think [2PL]?’

b. No, perdónenme ustedes

‘No, you [ustedes] forgive [2PL] me, you guys.’

c. ¡¡Qué bien cocinan ustedes!!

‘You guys [ustedes] cook [2PL]so well!!’

163

4.3. Oral Elicitation Task Data Results

4.3.1. Ustedes/Vosotros alternation

The tokens for the pronominal alternation for the oral elicitation task data were coded for seven independent variables, as described in section 3.2.2.2. These are education (low, medium, high), gender (female or male), age (young, adult old), urban/rural origin, region

(Western or Central Andalusia), person (infinitive verb, 2PL verb form, or object of preposition) and pragmatic context, including neutral contexts, anger contexts, criticism/irony contexts, formal contexts, praise contexts, and apology/petition contexts

A conditional inference tree for the oral elicitation task data is presented first to determine the main splits in these data, and to justify collapsing of the data conducted for a more balanced distribution, or when there is a lack of enough tokens for some categories.

Furthermore, there are factors that are shown as very important in the conditional inference trees and the random forests but then, they are not significant in the regression model; this is due to the fact that they have an effect only in interaction with other factors, which is shown in the inference trees. The conditional inference tree in Figure 503 shows that the main split in the data from the elicitation task, for the pronominal alternation, is that of age, with young speakers (Y) as opposed to both old (O) and adults (A) showing different tendencies.

3 The text is overwritten in the conditional inference tree, as was shown by the program, R, due to the quantity and complexity of interactions. This tree is presented only to see main splits in the data. 164

Figure 50. Conditional inference tree. Ustedes/vosotros alternation. Oral elicitation task.

165

Due to the splits shown in the conditional inference tree in Figure 50, age was collapsed into a two-way distinction, including young speakers on one hand, and old and adult speakers on the other. Before collapsing, adult speakers comprised 27.9% of the data, old speakers comprised 29.9% of the data, whereas young speakers comprised 42.2% of the data. Regarding education, in the inference tree, the split is most of the time between high (H) and medium (D) education as opposed to low education (L). Therefore, following that split in the data, education was collapsed into a two-way distinction, including low education as opposed to medium and high.

The random forest in Figure 51 shows the most important predictors for the ustedes/vosotros alternation from the oral elicitation task data, after collapsing. The most important factor is age, followed by region, education, gender, rural/urban origin, context, and person, i.e. 2PL, infinitive or object of preposition, in that order in terms of their statistical significance.

Figure 51. Random forest for the ustedes/vosotros alternation. Oral elicitation task data

166

The best-fit model for the oral elicitation task data for the ustedes/vosotros alternation after collapsing is shown below.

Estimate SE z-value p-value

(Intercept) -5.6404 1.5877 -3.553 0.000382 Age (reference is adult and old) Young -3.5131 1.1563 -3.038 0.002380 Region (reference is Central) Western 6.8108 1.8378 3.706 0.000211 Education (reference is high & medium) Low 5.4487 1.3328 4.088 4.35e-05 Gender (reference is female) Male 0.2015 1.0867 0.185 0.852921 Rural/Urban (reference is rural) Urban -0.8498 1.8823 -0.451 0.651640 Context (reference is neutral) Anger -0.0492 0.3856 -0.128 0.898472 Context (reference is neutral) Criticism -0.2773 0.3567 -0.777 0.436908 Context (reference is neutral) Formal 1.6266 0.5330 3.052 0.002275 Context (reference is neutral) Apology/Petition -0.1282 0.4306 -0.298 0.766012 Context (reference is neutral) Praise -0.8779 0.4647 -1.889 0.058878 Person (reference is object) 2PL 0.8036 0.3566 2.254 0.024215 Person (reference is object) Infinitive -1.2028 1.4039 -0.857 0.391561 Table 22. Best-fit model for the ustedes/vosotros alternation. Oral elicitation task.

167

The odds of encountering ustedes with the 2PL verb form over vosotros with the 2PL verb form are significantly lower when the speaker is young as opposed to old or adult. The odds of encountering ustedes over vosotros are significantly higher in Western Andalusia than in Central Andalusia, like found before for the online forced-choice survey and the online data from fora and Twitter, and PRESEEA Málaga. Low education as opposed to high or medium education results in a significantly higher probability of employing ustedes over vosotros. Formal contexts also result in a significantly higher probability of employing ustedes with the 2PL form of the verb over vosotros with the 2PL verb form, as opposed to neutral contexts, and, as found after releveling, as opposed to all the other contexts as well.

The pronoun used as a subject of a 2PL verb form as opposed to object of preposition also leads to significantly more use of ustedes over vosotros. After releveling the factor person and placing the 2PL in the reference level, I found that there is not significant difference between the pronoun used as a subject of an infinitive and the pronoun used as a subject of a

2PL verb form.

4.3.1.1. Ustedes/Vosotros Alternation in Western Andalusia

Like for previous datasets, from online sources, and from the online forced-choice survey, separate analyses of Western and Central Andalusia were conducted, since the two areas show different tendencies; more precisely, Western Andalusia leads to significantly more use of ustedes over vosotros (see best-fit model in Table 21 in section 4.3.1.). The two analyses, of Western and Central Andalusia, were conducted with the categories collapsed as described in section 4.3.1. In Western Andalusia, there is a total of 431 tokens for the pronominal alternation, with 41% of vosotros and 58.9% of ustedes. The conditional

168 inference tree in Figure 52 for Western Andalusia shows that there is a higher rate of ustedes with old or adult speakers as opposed to young speakers; with young speakers there is a higher rate of ustedes with low education as opposed to high or medium. With young speakers with high education, there is a higher rate of ustedes with male speakers than with female speakers, in formal and anger contexts. With female young speakers with high or medium education, there is a higher rate of ustedes in rural areas than in urban areas.

Young

High or medium

Low

Old & adult Low

Figure 52. Conditional inference tree. Ustedes/vosotros alternation. Western Andalusia. Oral elicitation task.

The random forest in Figure 53 shows age as the most important predictor for

Western Andalusia (see random forest in Figure 51, Section 4.3.1.), like for whole dataset,

169 followed by education, rural/urban origin, gender, context, and person, i.e. infinitive or 2PL verb form.

Figure 53. Random forest for the ustedes/vosotros alternation. Western Andalusia. Oral elicitation task.

Table 23. Best-fit model for the ustedes/vosotros alternation. Western Andalusia. Oral elicitation task.

Estimate SE z-value p-value

(Intercept) 0.4549590 2.0578748 0.221 0.8250 Age (reference is old and adult) Young -2.8892336 1.9404132 -1.489 0.1365 Education (reference is high and medium) Low 5.6347568 2.2069901 2.553 0.0107 * Urban/rural (reference is rural) Urban -0.6056270 1.7027703 -0.356 0.7221 Gender (reference is female) Male 0.7212751 1.8537416 0.389 0.6972 Context (reference is neutral) Anger 0.5173308 0.6064786 0.853 0.3937

170

Context (reference is neutral) Criticism -0.0594783 -0.0594783 -0.112 0.9111 Context (reference is neutral) Formal 2.8445206 1.1303652 2.517 0.0119 * Context (reference is neutral) Apology/Petition -0.0003272 0.6862213 0.000 0.9996 Context (reference is neutral) Praise 0.1527839 0.6711317 0.228 0.228

In Western Andalusia, the odds of encountering ustedes over vosotros are significantly higher when the speaker has low education as opposed to high or medium education. Formal contexts, as opposed to neutral, and, as found after releveling, also as opposed to all the other contexts, also lead to significantly more use of ustedes over vosotros in Western Andalusia.

4.3.1.2. Ustedes/Vosotros Alternation in Central Andalusia

For Central Andalusia, there is a total of 967 tokens from the oral elicitation task, with

67.8% of vosotros versus 32.2% of ustedes. The conditional inference tree in Figure 54 shows that there is a higher rate of ustedes in Central Andalusia with old or adult speakers with low education as opposed to high or medium, when ustedes occurs as subject of 2PL verb forms or as object of preposition, as opposed to when it occurs with the verb in the infinitive form; this can be observed on the right side of the tree. With young speakers, there is a higher rate of ustedes with females with low education as opposed to male speakers and high or medium education. With male young speakers, there is a higher rate of ustedes when

171 used as object of preposition as opposed to subject of 2PL verb form, as can be observed on the left side of the tree.

Young Old or adult

Low

High or High or Object Object medium Low medium 2PL Infinitive or 2PL

Figure 54. Conditional inference tree for the ustedes/vosotros alternation. Central Andalusia.

Oral elicitation task.

The random forest in Figure 55 shows that for Central Andalusia, like for Western

Andalusia, age is the most important predictor, then followed by education, gender, context, and person. Rural/urban origin was not included in the analysis of Central Andalusia data since all these data come from speakers with a rural origin.

172

Figure 55. Random forest for the ustedes/vosotros alternation. Central Andalusia. Oral elicitation task.

Table 24. Best-fit model for the ustedes/vosotros alternation. Central Andalusia. Oral elicitation task.

Estimate SE z-value p-value (Intercept) -4.51411 1.78861 -2.524 0.01161 Age (reference is old and adult) Young -3.17071 1.45491 -2.179 0.02931 Education (reference is high and medium) Low 5.20181 1.70234 3.056 0.00225 Gender (reference is female) Male 0.08548 1.36818 0.062 0.95018 Context (reference is neutral) Anger -0.46636 0.49067 -0.950 0.34188 Context (reference is neutral) Criticism -0.47297 0.46125 -1.025 0.30517 Context (reference is neutral) Formal 0.97798 0.62950 1.554 0.12028

173

Context (reference is neutral) Apology/Petition -0.40952 0.53524 -0.765 0.44421 Context (reference is neutral) Praise -1.86868 0.66475 -2.811 0.00494

In Central Andalusia, like for the whole dataset (see best-fit model in Table 21 in section 4.3.1), young speakers as opposed to old and adults, have an inhibiting effect on ustedes over vosotros, whereas low education leads to significantly more use of ustedes over vosotros as opposed to high and medium education. Praise contexts as opposed to neutral contexts results in a significantly lower probability of employing ustedes over vosotros in

Central Andalusia, and, as found after releveling, as opposed to all the other contexts too.

Formal contexts lead to significantly more use of ustedes over vosotros as opposed to all the other contexts, as found after releveling. After releveling, I also found that apology/petition contexts also lead to significantly more use of ustedes over vosotros, as opposed to anger, criticism, and praise contexts in Central Andalusia.

4.3.1.3. Conclusion

Summarizing, there is more use of ustedes over vosotros by non-young speakers, and speakers with low education. There is also more use of ustedes over vosotros in Western than in Central Andalusia, in formal contexts, and as subject of the 2PL than as an object of preposition. Furthermore, in Western Andalusia, there is more use of ustedes by male speakers than by females, and in rural areas than in urban areas with female speakers. There is more use of ustedes when the context is anger as well as when it is formal in Western

Andalusia. In Central Andalusia, there is more use of ustedes in formal contexts, and also in 174 apologies/petitions. Examples (41a), (41b), and (41c) illustrate the use of ustedes in formal, anger, and apology/petition contexts, respectively, in the oral elicitation task. In (41a), the speaker had to imagine he is addressing two doctors.

(41) a. Ustedes sois los expertos. Si ustedes decís que este tratamiento es el mejor

para curar a mi tío, esto es lo que hay. Ustedes decidís.

‘You guys [ustedes] are [2PL] the experts. If you guys [ustedes] say [2PL] that

this treatment is the best for healing my uncle, that’s is what has to be done. You

guys [ustedes] decide [2PL].’

b. ¡Y los tontos sois ustedes!

‘And you [ustedes] are [2PL] the stupid ones, you guys!’

c. pues si ustedes os pusierais de acuerdo y concedieseis el permiso para realizar

esta obra que tanto necesito.

‘So if you guys [ustedes] would reach [2PL] an agreement and granted [2P] the

permission to proceed with this construction that I need so much.’

4.3.2. 2PL/3PL Alternation

The tokens for the verbal alternation for the oral elicitation task data were coded for six independent variables, as described in section 3.2.2.2. These are education (low, medium, high), gender (female or male), age (young, adult, old), urban/rural origin, region (Western or

Central Andalusia, and pragmatic context, including neutral contexts, anger contexts, criticism/irony contexts, formal contexts, praise contexts, and apology/petition contexts

A conditional inference tree for the oral elicitation task data is presented first to determine the main splits in these data. The conditional inference tree in Figure 56 shows that

175 the main split is that of context, with a higher rate of 3PL verb forms in formal contexts. With the rest of the contexts, i.e. neutral, anger, criticism/irony, petition/apology (‘petition’ refers to both petitions and apologies), praise, there is a higher rate of 3PL verb forms with adult or old speakers in petition/apology contexts, as opposed to young speakers.

Adult or old

Young

Figure 56. Conditional inference tree. 2PL/3PL verb form alternation. Oral elicitation task.

Due to the splits observed in the conditional inference tree in Figure 56, age was collapsed into a two-way distinction, more precisely, young as opposed to adult or old, as was done for the pronominal alternation. Before collapsing, adult speakers comprised 37.8% of the data, old speakers also comprised 37.8% of the data, and young speakers comprised

24.4% of the data. The random forest in Figure 57 shows the most important predictors in hierarchical ordering in terms of their statistical significance, after collapsing. More precisely, the random forest in Figure 57 shows that context is the most important predictor for the

176 verbal alternation, followed by gender, age, rural/urban origin, region, and education, in that order.

Figure 57. Random forest. 2PL/3PL verb form alternation. Oral elicitation task.

The best-fit model for the data from the oral elicitation task data for the 2PL/3PL verb form alternation after collapsing is shown below.

Estimate SE z-value p-value (Intercept) -4.5992 0.8112 -5.670 <0.001 Context (reference is neutral) Formal 2.7843 0.5989 4.649 <0.001 Context (reference is neutral) Apology/ Petition 1.1829 0.5498 2.151 <0.05 Gender (reference is female) Male 0.8968 0.8785 1.021 0.3074 Age (reference is old and adult) Young 1.9698 0.9514 2.070 <0.05 Table 25. Best-fit model for the 2PL/3PL verb form alternation. Oral elicitation task.

177

The odds of encountering the 3PL verb form with ustedes over the 2PL verb form are significantly higher in formal and petition and apology contexts than in neutral contexts, and in formal contexts as opposed to all other contexts as well, as found after releveling. The odds of encountering the 3PL verb form with ustedes over the 2PL verb form are significantly higher with young speakers than with old or adult speakers.

4.3.2.1. Conclusion

Example (42) illustrates the results for the verbal alternation for the oral elicitation task. In example (42a), the use of the 3PL verb form with ustedes in a petition context is illustrated. Example (42b) shows the use of the 3PL verb form with ustedes in a formal context. In (42b) the speakers have to imagine they are addressing two doctors.

42) a. Por favor, ¿pueden ustedes cerrar los ojos que voy a explicaros un tema?

‘Please, can you guys [ustedes] close [3PL] your eyes, because I am going to

explain you something?’

b. Ustedes son los que tienen que decidir, yo estoy de acuerdo con ustedes y yo

confío en vosotros.

‘You [ustedes] are [3PL] the ones to decide, I agree with you guys [ustedes], and

I trust in you guys [vosotros].’

178

Chapter 5: Discussion and Conclusions

A summarize of the main results for the pronominal and verbal alternation is shown below, including what linguistic, social, and pragmatic factors lead to more use of ustedes over vosotros and to the 3PL verb form over the 2PL verb form with ustedes.

Ustedes over vosotros 3PL over 2PL verb form -Imperative mood -Post-verbal placement -Pre-verbal placement -Stative and mental verbs -Infinitive verbs -Infrequent verbs Linguistic factors -Subject and not object -No intervening material -Main and coordinate clauses -Overt subject -Priming -Subordinate clauses

-Western Andalusia -Central Andalusia -Males; old females in rural areas -Rural areas of Central Andalusia also Social factors -Females with non-high -Non-young speakers education & males -Low education -Non-young speakers

-Non-neutral online features -Non-neutral contexts: -Non-neutral contexts: -Formal

Pragmatic factors -Formal -Praise -Praise -Apology/petition -Apology -Anger

Table 26. Summary of results.

179

As summarized in Table 26, ustedes plus the 2PL or the 3PL verb form, as opposed to vosotros, is more likely to be produced in rural areas and by male speakers, but is less likely to be produced by young speakers or high-education speakers. Furthermore, ustedes plus the 2PL is more likely to occur in Western

Andalusia, as opposed to vosotros, and ustedes plus the 3PL is more likely to occur in

Central Andalusia, as opposed to ustedes plus the 2PL. In addition to males, females of an old age and from rural areas are also more likely to produce ustedes plus the

2PL over vosotros; females with non-high education are more likely to produce ustedes plus the 3PL over the 2PL with ustedes.

5.1. Change in Progress

The first research question of the present study referred to the status of plural address forms in Central and Western Andalusia as a change in progress, and regarding both the pronominal and the verbal alternations.

5.1.1. Pronominal Alternation

I predicted that the 2PL verb form with ustedes is the obsolete variant as opposed to vosotros plus the 2PL in Central Andalusia, which shows more leveling toward the standard Castilian norm. Results confirm that ustedes plus the 2PL verb form is more likely to occur in Western Andalusia than in Central Andalusia. Moreover, the effect of main factors affecting the variation, such as pre-verbal placement and presence of non-neutral online features is not present in Central Andalusia but only in Western

180

Andalusia. In Central Andalusia, the use of ustedes plus the 2PL verb form is just less common, it is a less productive form overall in Central Andalusia than in Western

Andalusia.

It was expected that frequent verbs would lead to more use of ustedes plus the

2PL verb form if this variant is the obsolete linguistic construction, leaving the plural address form system. Frequency did not play an independent role in the variation, but the effect of the factors placement and online features is deactivated with infrequent forms. Moreover, the effect of main or coordinate clauses leading to more use of ustedes as opposed to subordinates is activated only with frequent verbs, and not with the whole dataset. This means that factors with no effect with infrequent forms have a predictive effect with frequent forms; more precisely, when considering frequent and infrequent forms separately, the effect of placement, online features, and clause type is significant only with frequent forms. This confirms that lexical frequency does not have an independent effect on Spanish subject pronoun expression, but interacts with other constraints, as argued by Erker and Guy (2012). In the present study, lexical token frequency interacts with placement, online features, and clause type, showing that the factors determining the pronominal variation between ustedes and vosotros do not operate across the entire lexicon, but rather, they are significant for frequent verbs and do not have an effect with infrequent forms. High token frequency forms, unlike infrequent forms, are those with which the speakers have more experience. This evidence allows speakers to learn and reproduce certain patterns, since high token frequency forms provide more evidence about their contexts of occurrence of and their probable collocations (Erker and Guy 2012). Therefore, high token frequency forms such as ser, hacer ‘to be, to do’ provide sufficient information about the

181 collocation of the subject pronoun and the context of its occurrence; the collocation and context of occurrence is, in this case, that of ustedes in a preverbal placement, in main clauses, with co-occurrence of non-neutral online features, e.g. ustedes hacéis cosas que hacen daño ☹ ‘you guys do things that do harm ☹.’ The fact that the factors leading to more use of ustedes plus the 2PL verb form are activated with frequent forms suggests that ustedes plus the 2PL is the obsolescing linguistic element, kept in a certain collocation, i.e. pre-verbal placement, by the entrenchment effect of frequent forms. Similarly, in Canadian French, where the subjunctive has been lost, residues of the subjunctive remain with high token frequency verb forms

(Poplack 1992).

Furthermore, I predicted that overt pronoun, priming and no intervening material between subject and verb would lead to more use of ustedes plus the 2PL verb form.

However, these factors didn’t have predictive power in the pronominal alternation.

Nevertheless, the lack of effect of priming and intervening material is due to the fact that priming or intervening material were not considered when the subject was null, and there are more overt than null subjects after discarding null subjects (see methodology in section 3). The lack of effect of the overt/null pronoun factor is explained by the fact that numerous tokens with null subjects were not included in the dataset, when it couldn’t be established whether vosotros or ustedes was the subject

(see methodology in section 3), so that there are more overt subject tokens than null subject tokens. More precisely, out of a total of 2,408 tokens for the data from online sources and PRESEEA Málaga, there is a total of 943 (39%) null subjects and 1,465

(61%) overt subjects. On the other hand, the tendency in Spanish in general is just the opposite. Erker and Guy (2012) report that, out of 4,916 verbs included in their study,

182

1,709 (34.8%) occurred with overt pronouns while 3,207 (65.2%) were null, and that

Otheguy and colleagues (2007) also reported a 33.4% of overt pronoun occurrence as opposed to about 60% of null pronouns.

On the other hand, as hypothesized, a pre-verbal placement of the subject pronoun leads to more use of ustedes with the 2PL verb form. Therefore, ustedes plus the 2PL, like its precursor vuestras mercedes with 2PL verb form, appears more frequently in pre-verbal position, in Western Andalusia. Therefore, ustedes with the 2PL is following an archaic mixed pattern, which was present in the past, both with vuestras mercedes and ustedes, when the system was in a process of change. More precisely, the old pattern maintained is that of ustedes in the pre-verbal position. Furthermore, as predicted, ustedes is more likely to occur as subject than as object of preposition. This indicates a higher productivity of vosotros as opposed to ustedes, which is then restricted to occur within a collocation in which it is entrenched, i.e. as a pre-verbal subject of the verb, keeping the old pattern of vuestras mercedes. That ustedes plus the 2PL is an obsolescing construction is further supported by the fact that the effect of placement is activated by the entrenchment effect of frequent verbs and disappears with infrequent forms. It was also predicted that subordinate clauses would lead to more use of ustedes plus the 2PL verb form if it is in fact an obsolescing linguistic element. However, the opposite was found, more precisely, ustedes plus the 2PL verb form is more likely to occur in main clauses. Both main clauses and pre-verbal placement are positions better controlled by the speaker, since the position semantically prominent is better controlled by the speaker (Bybee 2001; Fernández

Martín 2012), suggesting that ustedes plus the 2PL verb form, which is more likely to occur in both pre-verbal placement and main clauses as opposed to vosotros, is less

183 productive than vosotros. On the other hand, although ustedes plus the 2PL verb form is less productive than vosotros, conserved in a pre-verbal collocation, and although it is the obsolescing linguistic element, ustedes plus the 2PL verb form is, however, not a wholly automated, routinized chunk of language. This is because, it is in main clauses where more complex content and pragmatic relations are expressed.

Furthermore, the fact that ustedes plus the 2PL verb form is more likely to co-occur with non-neutral online features in main clauses supports that it is not a pragmatically flat, non-productive chunk.

Furthermore, the effect of social factors also suggests that the plural address form system in both Central and Western Andalusia is in a process of change and that ustedes plus the 2PL verb form is an obsolescing linguistic element, as hypothesized.

More precisely, male speakers in Western Andalusia, and old females in Central

Andalusia and females in rural areas in Western Andalusia are more likely to use ustedes plus the 2PL as opposed to vosotros. While there is an effect of female speakers, this effect leading to more use of ustedes occurs always in interaction with other factors, more precisely, rural areas and old speakers; on the other hand, the effect of male speakers leading to more use of ustedes over vosotros is independent of other factors. Low education also leads to more use of ustedes plus the 2PL as opposed to vosotros, in both Central and Western Andalusia. It was predicted that the change in progress would be toward vosotros in Central Andalusia and toward the ustedes plus the 3PL verb form in Western Andalusia. However, results indicate that the change is toward vosotros in both regions. Ustedes plus the 2PL verb form as opposed to vosotros is associated with old speakers, male speakers, rural areas, and with low education. That, along with the fact that ustedes plus the 2PL verb form is

184 the less common variant in both regions, suggests a change in progress towards the

Castilian norm of vosotros plus the 2PL both in Western and Central Andalusia.

However, since ustedes plus the 2PL is less common and less productive in Central

Andalusia overall, this change in progress toward the standard Castilian variant of vosotros seems to be more advanced in Central Andalusia. This is further supported by the fact that the effect of old speakers leading to more use of ustedes over vosotros occurs specifically in Central Andalusia.

5.1.2. Verbal Alternation

I hypothesized that ustedes with the 3PL is more common in Western Andalusia.

However, for the verbal variation, i.e. the 2PL/3PL verb form alternation, contrary to expectations, I found that there is more use of ustedes plus the 3PL verb form in

Central than in Western Andalusia. This is confirmed by results obtained from analyzing different types of data, namely, the online and PRESEEA data, and the data from the online forced-choice survey. This is due to the fact that the differences between 2PL and 3PL with ustedes are larger in Western Andalusia than in Central

Andalusia. More precisely, in the data from online sources and PRESEEA Málaga, the difference between 2PL and 3PL in Western Andalusia are more than triple, i.e.

75.2% of 2PL versus 24.8% of 3PL; in Central Andalusia, the difference is less than triple, i.e. 72.3% of 2PL and 27.7% of 3PL. For the online forced-choice survey data, the difference in Western Andalusia between 2PL and 3PL is four times more 2PL than 3PL, i.e. 80% of 2PL versus 20% of 3PL, whereas in Central Andalusia there is

6% more of 3PL than 2PL, i.e. 47% of 2PL and 53% of 3PL. This is because in

185

Western Andalusia ustedes plus 2PL is more productive overall than in Central

Andalusia, where both 2PL and 3PL with ustedes are more similar regarding productivity, with low productivity compared to vosotros. Figure 58 shows these differences in Western and Central Andalusia for the online forced-choice survey. As

Figure 58 shows, the difference between rates of use of ustedes/vosotros is larger in

Central than Western Andalusia. On the other hand, the difference between rates of use of 2PL/3PL with ustedes is larger in Western than in Central Andalusia.

Pronominal alternation Verbal alternation 100% 100% 90% 83% 90% 80% 80% 80% 70% 70% 56.4% 60% 60% 53% 47% 50% 43.6% 50% 40% 40% 30% 30% 17% 20% 20% 20% 10% 10% 0% 0% Western Central Western Central

Vosotros Ustedes 2PL 3PL

Figure 58. Pronominal and verbal alternation in Western and Central Andalusia.

Online forced-choice survey.

I hypothesized that the combination of the 3PL verb form with ustedes is the innovative form. However, results contradict this hypothesis. More precisely, results show that the 3PL as opposed to the 2PL is the less common variant, more uncommon than ustedes plus the 2PL. This is especially true in speech, as shown by the data collected from the oral elicitation task, where the occurrences of the 3PL verb form

186 with ustedes are overwhelmingly lesser than those of 2PL verb forms with ustedes.

There are but a few instances of 3PL verb forms with ustedes in speech, more specifically, 61 out of 569, i.e. 10.7% of 3PL, versus 508 tokens of 2PL (89.3%). This suggests that ustedes plus the 3PL verb form is not a productive variant. Furthermore, it was expected that priming, frequent verbs, overt pronoun, and lack of intervening material between subject and verb would lead to more use of ustedes plus the 2PL as opposed to the 3PL, if the 2PL with ustedes is the obsolescing linguistic construction and the 3PL verb form with ustedes is the innovation. However, not only wasn’t that the case, but overt pronoun, priming, and lack of intervening material between subject and verb lead instead to more use of the 3PL with ustedes over the 2PL with ustedes.

Therefore, ustedes plus the 2PL verb form is not the unproductive sequence, but these results further indicate that ustedes plus the 3PL is the non-productive variant, the obsolescing linguistic element and not the innovation. More precisely, I found that infrequent stative or mental verbs lead to more use of the 3PL over the 2PL with ustedes; priming activated by frequent verbs also leads to more use of the 3PL over the 2PL with ustedes. The fact that ustedes plus the 3PL verb form is more likely to occur with infrequent stative or mental verbs is due to the association of this variant with a few verbs, in formulaic expressions of the type juzguen ustedes ‘you guys be the judges,’ disfruten ustedes ‘enjoy, you guys,’ perdonen ustedes ‘forgive me, you guys,’ which use infrequent stative or mental verbs. A formulaic expression is a specific routinized construction, a sequence that has become entrenched in its structure and is not likely to follow the most productive patterns of the language; it has separate lexical entries, i.e. it is stored in the lexicon as a whole and independent unit, a non-segmented and autonomous unit. This is further supported by the fact that

187 priming, activated by frequent verbs, and lack of intervening material between overt subject and verb lead to more use of the 3PL verb form with ustedes, which indicates that ustedes plus the 3PL is accessed and used like a chunk, like a non-productive single unit, as opposed to the 2PL verb form with ustedes, in both Central and

Western Andalusia, and is thus an obsolescing linguistic construction. It could argued that there is astable situation for the verbal alternation and not a change in progress, with the non-productive, lexicalized chunk ustedes plus 3PL being part of the vocabulary. However, this is not the case, since ustedes plus the 3PL is very infrequent, and even rare in speech, and the effects of social factors, as I argue below, show that ustedes plus the 3PL is an obsolescing linguistic element. Unlike for the pronominal alternation, tokens including null subjects didn’t have to be discarded for the 3PL verb forms because there is no variation between ustedes and vosotros with

3PL verb forms, which only occur with ustedes. Priming could then be checked more appropriately than for the pronominal alternation. Rosemeyer and Scwhenter (2017) found for the imperfective subjunctive that priming effects appear to be restricted to low-frequency -se forms because high-frequency forms are already entrenched by the frequency effect. My results appear to be the opposite, where frequent verb forms, not infrequent forms, rely on priming for activation of the 3PL with ustedes. However, this is because of the specific nature of the verbal data for plural address forms in

Andalusian Spanish, where the 3PL with ustedes is associated, by default, with a few specific infrequent lexical items with which it co-occurs in routinized constructions.

Therefore, for infrequent forms, priming is not needed, but priming is restricted to high-frequency verbs. Furthermore, with frequent forms, priming increases the productivity of ustedes plus the 3PL, not limited to a few specific lexical items as is

188 the case among infrequent forms, e.g. juzguen ustedes (cf. Rosemeyer and Schwenter

2017).

I also hypothesized that ustedes plus the 3PL verb form, if in fact the innovative variant, would occur more frequently in main clauses. Contrary to expectations, ustedes plus the 3PL is more likely to occur in subordinate clauses, which further suggests that ustedes plus the 3PL is the obsolete variant, since older linguistic forms, with less complex content or pragmatic relations, are retained in subordinate clauses over long periods of times (Bybee 2001). Ustedes plus the 3PL verb form in Central and Western Andalusia is thus following a similar pattern to that of Canadian French subjunctive, in the sense that it is limited to specific routinized constructions with specific lexical items, automated sequences that are accessed in production. Like the

Canadian French subjunctive, the 3PL verb form with ustedes is therefore preserved in entrenched sequences which behave more like lexical items than hierarchical syntactic structures (Bybee and Thompson 1997). Ustedes with the 3PL, as a formulaic sequence thus fulfils a specific function in discourse (cf. Posio 2015). More precisely, it is used as an attention-seeking marker, as in miren ustedes ‘look, you guys.’ Ustedes with the 3PL as a formulaic sequence is also used to make interlocutors take an epistemic stance, frequently the same one as the speaker. For example, in example (36) in section 4.1.2.4, reproduced below as (42) for convenience, the writer, in an online soccer forum, is suggesting that a scam has been perpetrated in the sale of soccer tickets. With opinen ustedes, the author is trying to get his interlocutors to confirm what a scam that was.

(42) Por lo visto se han vendido en 3 sitios las entradas, las mismas

localizaciones tres veces... Opinen ustedes.

189

‘It looks like tickets have been sold in three different places, the same tickets three

times. Give [3PL] your opinion, you guys [ustedes].’

In (43), the writer, in the online Carnival forum from Cádiz, is putting into question rankings and prizes given to Carnival groups. He is explaining evidence that proves that the rankings are fixed to the interlocutors. Then, he asks his interlocutors to give their opinion, but says that the proof is there, juzguen ustedes pero… ‘you guys be the judges, but…’, i.e. he wants the interlocutors to agree with him that the rankings are rigged.

(43) Juzguen ustedes mismos, pero aqui están las pruebas

‘You guys be the judges, but here is the proof.’

Furthermore, the effect of social factors also suggests that ustedes plus the 3PL verb form is an obsolescing linguistic element. More precisely, ustedes plus the 3PL is more likely to occur in rural areas, it is more likely to be produced by male speakers in Central Andalusia, and less likely to be produced by young speakers in both

Western and Central Andalusia. Female speakers with medium education are also more likely to produce ustedes plus the 3PL as opposed to females with low or high education; more precisely, the effect of medium education among females refers is mainly as opposed to high education, since there are only 4.6% of participants with low education in the online-forced choice survey, and it was in the online-forced choice survey data where this effect was found. Therefore, this means, that high education has an inhibiting effect on ustedes plus the 3PL. On the other hand, like for the pronominal alternation, while the effect of females leading to more use of the 3PL occurs in interaction with other factors, more precisely, education, the effect of male speakers leading to more use of the 3PL over the 2PL with ustedes is independent of

190 other factors. Results also indicate that young speakers have more use of the 3PL verb form in the data from the oral elicitation task, contradicting the results from the online forced-choice survey for both Central and Western Andalusia analyzed separately and together. However, for the verbal alternation, there are 569 tokens from the oral elicitation task, with a rate of only 10.7% of 3PL verb forms with ustedes, i.e. 61 tokens, as opposed to 89.3% of 2PL verb forms with ustedes, i.e. 508 tokens. On the other hand, from the data from the online forced-choice survey, there are 1,675 tokens for the verbal alternation, and the rates are 26% of 3PL verb forms with ustedes versus 74% of 2PL verb forms with ustedes. Therefore, the effect of young speakers on the verbal alternation in the oral elicitation task could be due to the fact that the few instances of 3PL in the oral data have been produced by mostly young speakers in that specific task; however, the better distribution of verb forms and the larger amount of data in the online-forced choice survey data provide more reliable results for the verbal alternation. Furthermore, that results from the oral elicitation task indicate that young speakers lead to more use of the 3PL verb form in the data is due to the fact that there is an interaction. As shown in Figure 56, there is actually a higher rate of ustedes plus the 3PL with old and adult speakers, as opposed to young speakers, in interaction with petition/apology contexts, whereas there are no interactions for young speakers. Therefore, it is old and adult speakers who are more likely to produce ustedes plus the 3PL but specifically in petition/apology contexts.

191

5.2. Discourse-Pragmatic Differences

The second research question of the present study relates to discourse-pragmatic differences between ustedes and vosotros and between the 2PL and the 3PL verb forms, and to how discourse-pragmatic factors affect the variation in Central

Andalusia differently from Western Andalusia, as well as the subject pronominal variation versus the verbal variation.

5.2.1. Pronominal Alternation

I hypothesized that ustedes with 2PL verb form is pragmatically non-neutral as opposed to vosotros. Results confirm this hypothesis. In the data from online sources, i.e. fora and Twitter, non-neutral online special features, i.e. emojis showing positive or negative emotions toward the interlocutors, lead to significantly more use of ustedes over vosotros, indicating that ustedes is the non-neutral variant, although in

Western Andalusia only, because Central Andalusia just show fewer occurrences of ustedes than Western Andalusia. Results from analyzing the data from the online forced-choice survey and the oral elicitation task also indicate that ustedes is the non- neutral form as opposed to vosotros, but both in Central and Western Andalusia.

Furthermore, results from the online forced-choice survey and the oral elicitation task indicate which non-neutral contexts specifically lead to more use of ustedes plus the

2PL verb form. More precisely, both in Western and Central Andalusia, contexts indicating positive feelings toward the interlocutors, where closeness toward the interlocutors is sought, e.g. in praise contexts, and, on the other hand, contexts

192 indicating negative feelings toward the interlocutors, where temporary distance from the interlocutors is sought, e.g. in anger contexts, lead to more use of ustedes plus the

2PL verb form. Furthermore, ustedes is used for distancing in formal situations, in

Central and Western Andalusia and specially in Central Andalusia; ustedes is also used for alleviating a face-threatening situation, more precisely, in apologies.

Furthermore, ustedes over vosotros was found to be more likely to occur with infinitive verbs, in the online forced-choice survey. This is because ustedes over vosotros is more likely to occur in apologies, and apologies in the online forced- choice survey were expressed as commands. More precisely, the options for the apology context in the online forced-choice survey were: perdonarme ustedes, perdonénme ustedes, perdonadme ustedes, perdonadme vosotros or perdonarme vosotros ‘forgive me, you guys;’ commands are frequently expressed using an infinitive in Andalusian Spanish, so the preferred option was ustedes with an infinitive, i.e. perdonarme ustedes. The tokens from the oral elicitation task were also coded for infinitive verb forms in the pronominal alternation, but infinitives didn’t have an effect in the data from the oral elicitation task. This is due to the fact that the number of coded infinitives is only 20 out of 1,398 tokens, i.e. there is a total of 1.4% of infinitives. This is because whether an infinitive was used in a command could not be determined in many cases, since final consonants are frequently dropped in speech in Andalusian Spanish.

The effect of anger, apology, and formal contexts leading to more use of ustedes over vosotros was found in both the online forced-choice survey and the oral elicitation task. On the other hand, the effect of praise leading to more use of ustedes over vosotros was found in the online forced-choice survey but not in the oral

193 elicitation task data. Therefore, the most important contexts leading to more use of ustedes plus the 2PL over vosotros are anger, formal, and apology contexts.

Since ustedes over vosotros is more likely to occur in non-neutral contexts, both in

Western and Central Andalusia, ustedes plus the 2PL over vosotros, is the pragmatically charged, pragmatically restricted, marked, non-neutral variant. This is supported by the greater likeliness of ustedes plus the 2PL verb form to occur in main clauses. More precisely, more complex pragmatic relations, greater richness, occur in main clauses as opposed to subordinate clauses. The markedness and non-neutrality of ustedes is further supported by the fact that it is more likely to occur in a pre-verbal placement. Syntactic fronting of the subject pronoun to a pre-verbal placement makes it, along with its occurrence in main clauses, topicalized, highlighted, given prominence, and thus marked (Bolinger 1972; Miller 1996; Raymond 2016). That ustedes plus the 2PL is more pragmatically rich is especially true in Western

Andalusia, where non-neutral online features also lead to more use of ustedes plus the

2PL; in Central Andalusia, ustedes plus the 2PL is less productive overall; online features had little effect in Central Andalusia but formal contexts are the most important ones leading to more use of ustedes plus 2PL.

On the other hand, non-apology commands and criticism contexts do not lead to more used of ustedes as opposed to neutral contexts. This is because these contexts are not as marked as the previous ones. More precisely, they don’t show strong emotions toward the interlocutors, as anger in anger contexts, or, on the other hand, praise. Furthermore, commands and criticism contexts are not as face-threatening as apologies, or do not require as much distancing as formal situations. In other words, criticism and commands are less marked and are thus closer to neutral contexts. In

194 fact, anger, praise, formal and apology contexts lead to significantly more use of ustedes over vosotros as opposed to commands and criticisms, as well as opposed to neutral contexts. Commands and criticisms are illustrated in (44), which shows a command context from the online forced-choice survey, and (45), which shows a criticism/irony context from the oral elicitation task. Compare (44) and (45) with an anger context, where there are even insults, in example (46).

(44) Tus hijos quieren que llames al electricista para reparar un problema que hay

en la casa. Quieres que llamen ellos. Les dices:

f) Llamadlo vosotros

g) Llamarlo ustedes

h) Llámenlo ustedes

i) Llamarlo vosotros

j) Llamadlo ustedes

‘Your children want you to call the electrician to fix some problem that you

have at home. You want them to call instead, so you tell them to call him.’

(45) Tus cuñados/as te quieren llevar de parranda; sabes que son muy fiesteros

para ti, beben mucho, y no te fías mucho de ir con ellos, y les dices que ellos

es que son muy peligrosos.

‘Your siblings- in- law want to take you to party; you know that they are party

animals and that they drink too much so you don’t trust going out with them.

You tell them that you don’t trust them and that they are very dangerous.’

(46) Estás hablando con tus primos sobre las pasadas elecciones generales.

Ellos han votado a un partido diferente al tuyo y la discusión se vuelve

acalorada, no estáis de acuerdo para nada. Te insultan, te dicen: “¿¿cómo has

195

podido votar a esa mierda de político y de partido, es que no piensas con la

cabeza, eres tonto o qué te pasa??” Les dices que los tontos son ellos y su

partido es el malo, y que igual que ellos pueden votar a quien ellos quieran tú

puedes votar a quien tú quieras.

‘You are talking with your cousins about the past general elections. They

voted for a different party than you. The argument becomes heated, you don’t

agree at all. They insult you, they tell you: “How could you vote for that shitty

politician and party, do you even use your head? Are you stupid or what is

wrong with you?” You tell them that they are the stupid ones, and the same as

they can vote for whoever they want to, you can vote for whoever you want

too as well.’

In conclusion, ustedes plus the 2PL verb form is the marked form conventionally implicating non-neutrality, the pragmatically restricted form as opposed to vosotros. It is used for both aggravating the force of speech acts and creating temporary distancing from the interlocutor, like in anger contexts, to save face, in apologies or petitions, to soften the force of a message, like when expressing disagreement, or to show affection, in praise contexts. Ustedes plus the 2PL verb form is then used for expressing both positive or negative feelings toward the interlocutors. The specific attitude or feeling toward the interlocutors, either positive or negative, is context- dependent, which means that the specific attitude is a conversational implicature, i.e. anger or praise are not encoded in the meaning of ustedes. On the other hand, vosotros is the non-marked form, used when the speaker desires to express neutrality. Sinnott argues that usted ‘you’ entails a conventional implicature of distance which is underspecified enough to include the different possibilities that may cause distance,

196 such as age, formality, respect; thus “exactly which is meant is conveyed via CVI”, so that “distance is entailed by the V form; the motivation for acknowledging or imposing distance can vary by the context” (Sinnott 2010: 212-213). The implicature of non-neutrality entailed by ustedes plus the 2PL verb form is underspecified for the different attitudes, positive or negative, that may be expressed toward one’s interlocutors. These attitudes are instead specified via conversational implicatures.

Sinnott also argues that “the V form entails specific information whereas the T form does not” (2010: 210). In the present dissertation, I have shown that the ustedes form entails specific information, that of non-neutrality, there has to be a reason to use of ustedes, whereas that is not the case of vosotros. Vosotros is the default unmarked form, which is the most frequent occurring in least specified, neutral, unmarked contexts (cf. Scwhenter and Cacoullos 2008). Therefore, given the fact that results of the present dissertation suggest that the change in progress is toward vosotros plus the

2PL verb form, in the plural address form system in Andalusian Spanish there could be a grammaticalization process in the sense that vosotros is acquiring the functional domain of ustedes, i.e. expressing marked contexts. This means that vosotros could be spreading even to those contexts previously reserved for ustedes, i.e. marked contexts, as found for Central and Northern Peninsular Spanish by Morgan and Schwenter

(2016). According to Andersen (2008), innovations extend to unmarked contexts first; that would be the case of vosotros, occurring first more frequently in neutral contexts; then, the use of the unmarked variant, i.e. vosotros, becomes more extended and the use of the marked form, i.e. ustedes, becomes more curtailed, until the unmarked variant becomes generalized, which would be the case of vosotros becoming generalized as the second person plural address form. This is supported by the effects

197 of social factors previously mentioned, where non-young speakers, males, low education speakers, and rural areas lead to more use of ustedes as opposed to vosotros; it is also supported by the effects of linguistic factors indicating that ustedes is the obsolescing linguistic element (see Section 5.1.1.).

5.2.2. Verbal Alternation

For the verbal alternation, I hypothesized that ustedes plus the 3PL verb form would also implicate non-neutrality, like ustedes plus the 2PL verb form. As hypothesized, the 3PL plus ustedes in both Western and Central Andalusia is more likely to occur in the imperative mood. However, the contexts from the online forced- choice survey and the oral elicitation task where ustedes plus the 3PL is more likely to occur are not commands. Therefore, the association of ustedes plus the 3PL verb form with the imperative mood is not so much to create temporary distancing from the interlocutors, to mitigate or to aggravate a command, but it is due to the fact that ustedes plus the 3PL is more likely to occur in formulaic expressions of the type juzguen ustedes ‘you guys be the judges.’ That ustedes plus the 3PL is more likely to occur in formulaic expressions is shown by the effect of the other linguistic factors as well, i.e. non-intervening material between overt subject and verb, priming, infrequent mental or stative verbs.

Results from the analysis of the data from online sources showed that ustedes plus the 3PL is not more likely to co-occur with online features, but ustedes plus the 2PL is. However, results from the online forced-choice survey and the oral elicitation task indicate that the main factor affecting the verbal variation is pragmatic context. More

198 precisely, in the online forced-choice survey data, ustedes plus the 3PL is more likely to occur in formal contexts, as expected, like in the standard use in Castilian Spanish, but also in apology contexts, to alleviate the face threatening situation caused by these contexts, and in praise contexts. In the oral elicitation task, ustedes plus the 3PL is more likely to occur in formal and apology/petition contexts. Therefore, these two, i.e. formal and apology/petition, are the most important contexts leading to more use of

3PL verb form with ustedes; this is also shown by the higher rate of 3PL verb form with ustedes than 2PL in formal and apology/petition contexts as opposed to the other contexts, as indicated in the conditional inference trees in Figures 44, 46, and 48 (see

Section 4.2.2.). On the other hand, the most important contexts leading to more use of ustedes plus the 2PL are anger, formal, and apology contexts. The range of contexts in which the 3PL more frequently occurs is then smaller than the range of contexts in which the 2PL occurs. Therefore, ustedes plus the 3PL is more pragmatically restricted than ustedes plus the 2PL verb form. Moreover, the contexts where ustedes plus the 2PL verb form occurs include contexts that are more marked, i.e. anger contexts. Furthermore, ustedes plus the 2PL verb form over vosotros is more likely to occur in apologies in the online forced-choice survey, but, for the oral task, ustedes plus the 2PL verb form is more likely to occur in apology/petitions only in Central

Andalusia. On the other hand, the 3PL verb form is more likely to occur in apologies, in both the online forced-choice survey and the oral task, and in the oral task, in petitions too, in Western and Central Andalusia. Petitions, like commands or criticism, are less marked than other contexts, such as anger contexts. Example (46) includes a petition context. Compare it with the anger context in (45) above.

199

(46) Tus vecinos son muy simpáticos y te llevas muy bien con ellos, pero no son

muy limpios y siempre te toca a ti y tu familia limpiar las escaleras y el portal de

tu bloque. Un día te encuentras a varios de ellos al bajar las escaleras. Hablas con

ellos y les dices amablemente que sería bueno que no sólo tú y tu familia, sino que

también ellos limpiaran la escalera y el portal, para mantenerlo todo limpio.

‘Your neighbors are very nice and you get along with them very well, but they are

not very clean. You and your family have always to clean the stairs and the

hallway. One day you run into several of you neighbors when going down the

stairs. You talk to them and nicely and tell them that it would be good that not

only you and your family but also, they cleaned the stairs and the hallway, to keep

everything clean.’

Furthermore, non-neutral online features lead to more use of the 2PL verb form with ustedes over the 3PL verb form with ustedes, showing that the 3PL with ustedes is less pragmatically charged than the 2PL with ustedes. This is also supported by the fact that the 3PL verb form over the 2PL is more likely to occur in subordinate clauses, which are more pragmatically flat than main clauses, where the 2PL with ustedes is more likely to occur. It was hypothesized that the factors online features and type of context shouldn’t lead to a significant difference between 2PL and 3PL with ustedes as both variants were hypothesized to similarly indicate non-neutrality.

The fact that both factors show significant differences between the variants indicate a pragmatic difference, more precisely, that the 3PL with ustedes is less pragmatically rich than the 2PL with ustedes. Notwithstanding, ustedes plus the 3PL verb form is still a variant conventionally implicating non-neutrality, whereas the specific attitude, e.g. alleviating a face-threatening situation in apologies or petitions, distancing for

200 formality, or praise, is not encoded in its meaning but it is context dependent, i.e. it is a conversational implicature.

5.3. Final Conclusions and Contributions

The pronominal and verbal alternation are similar regarding pragmatic factors, where both the 3PL and 2PL with ustedes are variants used to express non-neutrality.

Both alternations are also similar in terms of social factors affecting the variation, with old speakers, males, low education, and rural areas leading to more use of ustedes plus the 2PL or the 3PL. Both linguistic and social factors indicate that the pronominal and the verbal plural address form systems in Central and Western

Andalusia are in a process of change, where ustedes plus the 2PL or the 3PL are obsolescing linguistic elements as opposed to vosotros. Moreover, ustedes plus the

2PL or the 3PL are the marked forms in different levels indicating, in a probabilistic way, the same thing, i.e. that ustedes plus the 2PL or the 3PL are obsolescing linguistic elements leaving the system as opposed to vosotros. Those levels include frequency markedness, where ustedes plus the 2PL or the 3PL are the less frequent variants; productivity markedness, where ustedes plus the 2PL or the 3PL are less productive than vosotros; social markedness, where these variants are more likely to be produced by speakers belonging to marked social categories. Finally, ustedes plus the 2PL or the 3PL are also marked at a pragmatic level, where both variants are non- neutral, as opposed to vosotros. Therefore, the plural address form system seems to be leveling toward the standard Castilian variant vosotros plus the 2PL verb form, and, especially in Central Andalusia, which is closer to the Castilian norm; this is the main

201 difference between the regions, Central and Western Andalusia. Regarding differences between the verbal and the pronominal alternation, ustedes plus he 2PL seems to be entrenched in the pattern of the old form vuestras mercedes, in a pre- verbal placement. On the other hand, ustedes plus the 2PL verb form is not an unproductive sequence following a construction-specific pattern as ustedes plus the

3PL verb form does, which is the main difference between the verbal and the pronominal variations in terms of linguistic factors. What’s more, ustedes plus the

2PL presents more pragmatic richness than ustedes plus the 3PL verb form, which is also a difference between the verbal and pronominal variations, in terms of pragmatic factors. More precisely, as shown in Figure 59, there is a continuum where vosotros is the default, least specified and most productive variant, and the innovation which is becoming more widespread as the main plural address form in Central and Western

Andalusia. By contrast, ustedes with the 2PL is more specified than vosotros, entails specific information of non-neutrality, and is a marked variant, an obsolescing linguistic construction; ustedes with the 3PL is the most pragmatically restricted variant, but occurs in less marked contexts than ustedes plus the 2PL; it is also the least productive variant, following a construction-specific pattern.

+ Productive------Productive - Specified ------+ Pragmatically restricted -Marked ------+Marked

Vosotros+2PL ------Ustedes+2PL ------Ustedes+3PL Figure 59. Productivity, markedness and pragmatic restriction continuum.

202

Whereas previous work on plural address forms in Andalusian Spanish argues that there are no pragmatic differences between ustedes/vosotros or between the 2PL/3PL verb forms (Lara Bermejo 2015; Narbona et al. 1998), the present dissertation shows that, as found by prior research for singular address form variation in Spanish (Millán

2011; Sinnott 2010), plural address form variation in Andalusian Spanish is likewise closely tied to pragmatic differences going beyond contextual (in)formality and

(a)symmetric interpersonal relationships, in both Central and Western Andalusia, which present similar tendencies regarding pragmatic differences between the variants. This study therefore sheds new light on address form shifts in Spanish, which are typically assigned to static social categories (cf. Raymond 2016). It shows how address form switching is used by the same speaker in the same interaction to convey different pragmatic meanings.

Furthermore, the present dissertation includes an innovative approach to study pragmatic differences, by analyzing the effect of emojis, emoticons, and other online features, as independent variables in a variationist study. This is innovative both with respect to the subject pronoun research tradition and with respect to variationist linguistics research more generally. The fact that different pragmatic contexts are operationalized, based on either the presence of different online features or on the type of context that was constructed, and included as factors within a variationist sociolinguistic study is also innovative.

Moreover, the approach to subject pronoun variation undertaken in the present study is innovative with respect to the subject pronoun variation research tradition also in the sense that this research typically addresses overt versus null singular pronouns in Spanish (Carvalho et al. 2015). Furthermore, the present dissertation

203 analyzes priming effects on subject pronoun variation differing from the subject pronoun variation research tradition. Subject pronoun variation studies have looked at subject continuity, related to switch reference, and/or at priming in relation to subject continuity, as well as in relation to overt/null subject expression; more precisely, previous studies on subject pronoun variation analyze the effects of priming on types of subjects, i.e. overt subject priming overt subject or null subject priming null subject

(Carvalho et al. 2015; Erker and Guy 2012; Torres Cacoullos and Travis 2013). On the other hand, the present dissertation analyzes the effects of priming on specific subject pronouns and verb forms that alternate but have the same referent, i.e. 2PL.

This dissertation also contributes to the study of language change, corroborating the role of clause type, formulaic expressions, priming and frequency in changes in progress (cf. Erker and Guy 2012; cf. Rosemeyer and Schwenter 2017). More precisely, I have shown that priming has a conserving effect; however, I have demonstrated that obsolescing linguistic constructions are more likely to occur when there is a prior same form only when the lexical item involved needs to rely on priming for activation; this is true for lexical items which fill the slot of the obsolescing formulaic construction less frequently, and not the specific lexical items which typically occur within the formulaic expression. Furthermore, this dissertation corroborates that the conserving effect of frequency doesn’t work independently but works in interaction with other factors constraining the variation which are activated by frequency. On the other hand, in this dissertation I have shown that formulaic sequences, such as that of ustedes plus the 3PL, does not have to be necessarily linked to frequent forms, as has been argued (Erker and Guy 2012). Moreover, the results of this dissertation also show that the imperative mood is not necessarily associated with

204 innovations, as has been argued (Lara Bermejo 2015). On the contrary, the results of the present study suggest that obsolescent forms may be conserved in the imperative mood. Similarly, Millán (2011) found that voseo forms, which are rare and stigmatized in Bogotá, Colombia, where tú and usted are used instead as singular address forms, are restricted to the imperative mood, to commands.

Lastly, linguistic studies of address form, including those addressing pragmatic differences, have focused overwhelmingly on singular forms, whereas plural address forms have been ignored (Morgan and Schwenter 2016). Therefore, the present dissertation contributes to the analysis of plural address forms more generally, which are highly understudied from a cross-linguistic perspective.

Regarding future steps, more spoken data are needed, to have a better representation of the different social categories. Then, with more data from speech, linguistic factors will be included in the analysis of oral data, which could not be done in the present dissertation. Therefore, the interactions between social, pragmatic, and linguistic factors will be analyzed. More precisely, it will be explored whether the effects of frequency, intervening material between subject and verb, priming, clause type and placement are stronger among speakers in those extra-linguistic categories where ustedes plus the 2PL form of the verb is less likely to occur. These are women and young speakers with urban origin and high education, in Central Andalusia. It will also be explored whether, on the other hand, male, old, rural, low education speakers show more use of ustedes plus the 2PL verb form in general, without the need for priming or frequent forms to trigger its use. I will also be able to explore whether the pragmatic specialization of ustedes plus the 2PL or the 3PL verb form of the verb, as non-neutral variants, will have a stronger effect for participants for which

205 ustedes plus the 2PL or 3PL verb form is less common, i.e. speakers in the social categories which have an inhibiting effect on the use of ustedes plus the 2PL or 3PL.

This means that non-neutral online features or non-neutral contexts will lead to more use of ustedes plus the 2PL or the 3PL verb form of the verb especially among women, young and higher education speakers with urban origin, and in Central

Andalusia, since the phenomenon is less common among those speakers, and in

Central Andalusia. Furthermore, in future studies, interactions will be included in the models.

Furthermore, attitudes of Andalusian native speakers toward the different variants under study need to be addressed, by conducting a perception study, using a five-point

Likert scale to measure the acceptability of the different variants, in different contexts, as perceived by Central and Western Andalusian native speakers. Finally, a comparison of the use of ustedes/vosotros in Central and Western Andalusia and in the rest of Peninsular Spanish needs to be conducted. More precisely, vosotros seems to be spreading even to formal contexts in Peninsular Spanish, thereby replacing ustedes (Fernández Martín 2012; Morgan and Schwenter 2016). Therefore, a variationist analysis of ustedes/vosotros including linguistic factors, in Peninsular

Spanish, not including Central and Western Andalusia, needs to be conducted to determine whether ustedes plus the 3PL is also more likely to occur in the imperative mood in formulaic expressions, like in Central and Western Andalusia.

206

References

Alvar López, Manuel. 1973. Atlas lingüístico y etnográfico de Andalucía (vols. 1-6).

Granada: Universidad de Granada- CSIC.

Amaral, Patrícia, and Scott A. Schwenter. Contrast and the (Non)-Occurrence of

Subject Pronouns. In David Eddington (ed.), Selected Proceedings of the 7th

Hispanic Linguistics Symposium, 116-127. Somerville, MA: Cascadilla

Proceedings Project.

Andersen, Henning. 2008. Grammaticalization in a speaker-oriented theory of change.

In Thoårhallur Eythoårsson (ed.), Grammatical Change and Linguistic

Theory. The Rosendal papers (Linguistik Aktuell/Linguistics Today, 113),

11-44. Amsterdan & Philadelphia: John Benjamins.

Bentivoglio, Paola. 2003. Spanish forms of address in the sixteenth century. In Irma

Taavitsainen & Andreas H. Jucker (eds.), Diachronic Perspectives on Address

Term Systems (Pragmatics & Beyond New Series 107), 177-191. Amsterdam

& Philadelphia: John Benjamins.

Birner. Betty J. 2013. Introduction to Pragmatics. Malden: Wiley-Blackwell.

Blake, Barry J. 1994. Case. Cambridge: Cambridge University Press.

Bolinger, Dwight. 1972. A look at equations and cleft sentences. Studies for Einar

Haugen:Presented by friends and colleagues, ed. by Evelyn Scherabon

Firchow, Kaaren Grimstad, Nils Hasselmo, andWayne A. O’Neil, 96–114. The

Hague: Mouton.

Bustos Tovar. 2001. Lo común y lo diverso en el andaluz. In Carmen Lucía Reina

207

Reina (coord.), Actas de las Jornadas “El habla andaluza: Historia, normas,

usos”, 255-276. Estepa: Ayuntamiento de Estepa.

Bybee, Joan. 2001. Main clauses are innovative, subordinate clauses are conservative:

consequences for the nature of constructions. In Joan Bybee & Michael

Noonan (eds.), Complex sentences in grammar and discourse: Essays in honor

of Sandra A., 1-17. Amsterdam: John Benjamins.

Bybee, Joan. 2002. Word frequency and context of use in the lexical diffusion of

phonetically conditioned sound change. Language Variation and Change 14.

261-290.

Bybee, Joan. 2006. From usage to grammar: the mind’s response to repetition.

Language 82 (4). 711-733-

Bybee, Joan & Mary Alexandra Brewer. 1980. Explanations in morphophonemics:

changes in provençal and Spanish preterite forms. Lingua 52. 201-242.

Bybee, Joan & Sandra Thompson. 1997. Three frequency effects in syntax. In

Mathew L. Juge & Jery L. Moxley (eds.), Proceedings of the Twenty-Third

Annual Meeting of the Berkeley Linguistics Society: General Session and

Parasession on Pragmatics and Grammatical Structure (BLS), 378-388.

Berkeley: Berkeley Linguistics Society.

Carbonero, Pedro & Asunción Ortiz (eds.). 2005. Sociolingüística andaluza 14:

Léxico del habla culta de Sevilla. Sevilla: Universidad de Sevilla.

Carvalho, Ana M, Rafael Orozco & Naomi Lapidus Shin (eds). 2015. Subject

Pronoun Expression in Spanish: A Cross-Dialectal Perspective. Washington

D.C.: Georgetown University Press.

Castro, Amanda Lizet. 2002. Pronominal address in Honduran Spanish. Munich:

208

LINCOM.

Davies, Mark. (2016-). Corpus del Español: Two billion words, 21

countries. http://www.corpusdelespanol.org (accessed 20 October 2016).

Danet, Brenda. 2010. Computer mediated English. In Janet Maybin & Joan Swann

(eds.), The Routledge Companion to English language studies, 146-156. New

York, NY: Routledge.

Dresner, Eli & Susan C. Herring. 2010. Functions of the nonverbal in CMC:

Emoticons and Illocutionary Force. Communication Theory 20(3). 249–268.

Erker, Daniel & Gregory R. Guy. 2012. The role of lexical frequencly in syntactic

variability: Variable subject personal pronoun expression in Spanish.

Language 88(3). 526-557.

Fernández Martín, Elisabeth. 2012. La oposición vosotros/ustedes en la historia del

español peninsular (1700-1931). Granada: Universidad de Granada

dissertation.

García Mouton, Pilar. 1992. El Atlas Lingüístico y Etnográfico de Andalucía.

Hombres y mujeres. Campo y ciudad. Actas del Congreso Internacional de

Dialectología, Iker 7, 151-177. Bilbao: Real Academia de la Lengua Vasca.

Grupo Vernáculo Urbano Malagueño (VUM), Universidad de Málaga. (n.d.). Corpus

oral [Data file]. http://www.vum.uma.es/corpus_oral.php (accessed 6 March

2016).

Hernández Campoy, Juan Manuel & Juan Andrés Villena Ponsoda. 2009.

Standardness and Nonstandardness in Spain: Dialect Attrition and

Revitalization of Regional Dialects of Spanish. Int’l. Soc. Lang. 196-197. 181-

214.

209

Horn, Lawrence R. 1984. Toward a new taxonomy for pragmatic inference: Q-based

and R-based implicature. In Deborah Schiffrin, (ed.), Meaning, Form and Use

in Context: Linguistic Applications, 11-42. Washington D.C.: Georgetown

University Press.

Jaeger T. Florian & Neal Snider. 2008. Implicit learning and syntactic persistence:

Surprisal and cumulativity. In Bradley C. Love, Ken McRae &Vladimir M.

Sloutsky (eds.), Proceedings of the 30th Annual Conference of the Cognitive

Science Society (CogSci), 1061–1066. Austin, TX: Cognitive Science Society.

Jaime Jiménez, Elena. 2011. Lo que ustedes queráis: plural second-person pronoun

Variation in Andalusian Spanish. Unpublished Manuscript, Department of

World Languages, Literatures, & Linguistics, West Virginia University,

Morgantown, WV.

Jaime Jiménez. 2014. Si ustedes queréis: A quantitative approach to variation of

second person plural pronouns in Andalusian Spanish.

Columbus, OH: The Ohio State University unpublished manuscript.

Jaime Jiménez. 2016. Second Person Plural Variation in Andalusian Spanish.

Columbus, OH: The Ohio State University unpublished manuscript.

Jiménez Fernández, Rafael. 1999. El andaluz. Madrid: Arco.

Johnson, Mary. 2016. Pragmatic variation in voseo and tuteo negative commands in

Argentinian Spanish. In María Irene Moyna and Susana Rivera-Mills (eds),

Forms of Address in Spanish across the Americas (Issues in Hispanic and

Lusophonic Linguistics 10), 127-148. Amsterdan & Philadelphia: John

Benjamins.

Kalman, Yoram M. & Darren Gergle. 2014. Letter repetitions in computer-mediated

210

communication: A unique link between spoken and online language.

Computers in Human Behavior 34. 187-193.

Lamíquiz, Vidal & Miguel Ángel de Pineda. (eds.). 1983. Sociolingüística

andaluza 2: material de encuestas para el estudio del habla urbana culta de

Sevilla. Sevilla: Universidad de Sevilla.

Lara Bermejo, Víctor. 2015. Los tratamientos de 2PL en Andalucía occidental y

Portugal: estudio geo- y socio-lingüístico de un proceso de gramaticalización.

Madrid: Universidad Autónoma de Madrid dissertation.

Mayol, Laia. 2010. Contrastive Pronouns in Null Subject Romance Languages.

Lingua 120 (10). 2497-2514.

Martínez-Sanz, Cristina & Gerard Van Herk. 2013. Nothing Happening Here:

Frequency Effects and Dominican Spanish Pronominal Subject Expression.

Paper presented at New Ways of Analyzing Variation 42, Pittsburgh.

Millán, Mónica. 2011. Pronouns of address in informal contexts: a comparison of two

dialects of Colombian Spanish. Urbana, IL: University of Illinois at Urbana-

Champaign dissertation.

Miller, Jim. 1996. Clefts, particles and word order in languages of Europe. Language

Sciences 18 (1-2). 111-125.

Moreno de Alba, José Guadalupe. 2010. Notas sobre la cronología de la eliminación

de vosotros en América. In Rosa María Castañer Martín & Vicente Lagüéns

Gracia (coords.), De moneda nunca usada: Estudios dedicados a José Mª

Enguita Utrilla, 461-470. Zaragoza: Instituto Fernando el Católico, CSIC.

Morgan, Terrell A. & Scott A. Schwenter. 2016. Vosotros, ustedes, and the myth of

211

the symmetrical Castilian pronoun system. In Alejandro Cuza, Lori

Czerwionka & Daniel Olson (eds.), Inquiries in Hispanic Linguistics: From

theory to empirical evidence (Issues in Hispanic and Lusophonic Linguistics

12), 263-280. Amsterdam & Philadelphia: John Benjamins.

Narbona, Antonio, Rafael Cano & Ramón Morillo-Velarde. 1998. El español

hablado en Andalucía. Sevilla: Ariel Lingüística.

Narbona, Antonio, Rafael Cano & Ramón Morillo-Velarde. 2011. El español

hablado en Andalucía. Sevilla: Ariel Lingüística.

Nieuwenhujsen, Dorien. 2006. Vosotros: surgimiento y pérdida de un pronombre en

perspectiva diacrónica y diatópica. Proceedings from VI Congreso

Internacional de Historia de la Lengua Española, coord. by J. L. Girón

Alconchel and J. de Bustos Tovar, 949-960. Madrid: Arcos.

Otheguy, Ricardo, Ana Celia Zentella &David Livert. 2007. Language and dialect

contact in Spanish in New York: Toward the formation of a speech

community. Language 83.770–802.

Penny, Ralph. 2002. A history of the Spanish language. Cambridge: Cambridge UP.

Poplack, Shana. 1992. The inherent variability of the French subjunctive. In

Christiane Laufeur & Morgan Terrell (eds.), Theoretical Analyses in Romance

Linguistics, 235-263. Amsterdam: Benjamins.

Posio, Pekka. 2015. In Ana Carvalho, Ana M, Rafael Orozco & Naomi Lapidus Shin

(eds), Subject Pronoun Expression in Spanish: A Cross-Dialectal

Perspective, 59-78. Washington D.C.: Georgetown University Press.

RAE. 2017. Usted. In Diccionario panhispánico de dudas.

http://lema.rae.es/dpd/?key= (accessed 21 December 2017).

212

Raymond, Chase. 2016. Linguist reference in the negotiation of identity and action:

Revisiting the T/V distinction. Language 92 (3). 636-670.

R Core Team. 2015. R: A language and environment for statistical computing. R

Foundation for Statistical Computing, Vienna, Austria. http://www.R-

project.org/ (accessed 10 January 2016).

Rosemeyer, Malte. 2015. How usage rescues the system: Persistence as conservation.

In Aria Adli, Marco García García & Göz Kaufmann (eds.), System, usage,

and society, 1-26. New York: De Gruyter.

Rosemeyer, Malte & Scott A. Schwenter. 2017. Entrechment and persistence in

language change: the Spanish past subjunctive. Corpus Linguistics and

Linguistic Theory. 0(0). -. Retrieved 1 Nov. 2017, from doi:10.1515/cllt-2016-

0047.

Sánchez Méndez, Juan. 2003. Historia de la lengua española en América. Valencia:

Tirant lo Blanch.

Schwenter, Scott A. 2015. Priming effects and obsolescence in morphosyntactic

variation: Evidence from Spanish and Portuguese. Paper presented at the 4th

Annual Hispanic and Luso-Brazilian Linguistics Conference, Arizona State

University, 10-11 April.

Schwenter, Scott A. & Rena Torres Cacoullos. 2008. Defaults and indeterminacy in

temporal grammaticalization: the ‘perfect’ road to perfective. Language,

Variation and Change 20. 1-39.

Silva-Corvalán, Carmen. 2001. Sociolingüística y pragmática. Washington, DC:

Georgetown University Press.

Sinnott, Sarah. 2010. Address Forms in Castilian Spanish: Convention and

213

Implicature. Columbus: The Ohio State University dissertation.

Szmrecsanyi, Benedikt. 2005. Language users as creatures of habit: A corpus-based

analysis of persistence in spoken English. Corpus Linguistics and Linguistic

Theory 1(1). 113-149.

Tagliamonte, Sali. A. 2012. Variationist Sociolinguistics: Change, Observation,

Intepretation. Malden, MA: Whiley-Blackwell.

Tagliamonte, Sali A. & R. Harald Baayen. 2012. Models, forests, and trees of York

English: Was/Were variation as a case study for statistical practice. Language

Variation and Change 24. 135-178.

Thorsos, Eileen. 2002. Markedness and Morphological Change in Obsolescent

Languages. Swarthmore: Swarthmore College dissertation.

Tiersma, Peter Meijes. 1982. Local and General Markedness. Language 58 (4). 832-

849.

Torres Cacoullos, Rena & Catherine E. Travis. 2013. Assessing inter-linguistic

(di)similarity via intra-linguistic variability for subject expression. In Ana

Maria Carvalho, Rafael Orozco & Naomi Lapidus Shin (eds.), Subject

pronoun expression in Spanish: A cross-dialectal perspective, 1-20.

Georgetown: Georgetown University Press.

Vida Castro, Matilde (ed.). 2007. El español hablado en Málaga. Málaga: Sarriá.

Villena Ponsoda, Juan Andrés. 2000. Identidad y variación lingüística: Prestigio

nacional y lealtad vernacular en el español hablado en Andalucía. In Georg

Bossong & Francisco Báez de Aguilar González, Identidades lingüísticas en la

España autonómica: Proceedings from Jornadas Hispánicas held in Zurich,

1997, 107-150. Madrid: Iberoamericana.

214

Appendix A: Online Forced-Choice Survey

1) Te gusta la playa pero te da miedo el agua. Te pregunta alguien en la playa

que por qué no te bañas. Le dices:

a) Es que me da miedo la mar

b) Es que me da miedo el mar

c) Es que me da miedo el agua

You like the beach but you are afraid of water. Someone asks you in the

beach why you don’t swim. You say: “a) and b) I am afraid of the sea c) I

am afraid of water”

2) Tus primos te preparan una paella riquísima. Les dices:

d) ¡¡Qué bien cocináis ustedes!!

e) ¡¡Qué bien cocinan ustedes!!

f) ¡¡ Qué bien cocináis vosotros!!

Your cousins prepare a delicious paella for you. You tell them: “You guys cook so well!!”

3) Una vecina cotilla te pregunta por los estudios de tu nieta, que se llama María.

Tu nieta es muy buena estudiante, entonces le dices:

a) La María va muy bien

b) María va muy bien

c) Mi nieta va muy bien

A gossipy neighbor asks you about the studies of your granddaughter, who

is called María. Your granddaughter is a very good student, so you tell

215

your neighbor: “a) The María is doing very good b) María is doing very

good c) My granddaughter is doing very good”

4) La casa de tus padres está muy mal, muy dejada y ellos están muy mayores

para arreglarla o limpiarla ya. Entonces, os toca a ti y a tus hermanos, pero, al

tiempo de haber quedado en arreglarla, tus hermanos te dicen que no limpiaste

ni arreglaste nada al final, que sigue todo muy mal. Sabes que ellos tampoco

hicieron mucho. Les dices:

d) Claro, como que vosotros limpiasteis mucho

e) Claro, como que ustedes limpiaron mucho

f) Claro, como que ustedes limpiasteis mucho

Your parents’ house is in a very bad state, but your parents are very old to fix

it or clean it. Therefore, your siblings and you have to do it. However,

sometime after having cleaned and organized the house, your siblings tell you

that you didn’t clean much, that the house is still dirty and messy. You know

that they didn’t do much either. You tell them: “Right, as if you guys are very

clean.

5) Estabas bebiendo agua, pero tiras el vaso y se rompe. Tu madre, que está en la

habitación de al lado, te pregunta que qué pasó. Le dices:

a) Se me ha caído el vaso

b) Se ha caído el vaso

c) Se me cayó el vaso

d) Se cayó el vaso

e) Me se ha caído el vaso

216

You are drinking a glass of water but you drop the glass and it breaks.

Your mother, who is in the room next to you, asks you what happened. You

tell her: “I dropped the glass of water”

6) Los vecinos de la puerta de al lado llevan toda la noche con la música a todo

volumen porque tienen una fiesta en casa. No es la primera vez que lo hacen,

ya lo han hecho muchas veces. No puedes dormir por la música y al día

siguiente tienes que trabajar. Además, tienes a tu madre enferma que tampoco

puede dormir. Vas y les pides amablemente que le bajen el volumen a la

música, por favor, que tu madre está enferma. Te dicen que no, que no

molestes, que te vayas a dormir, que eres un pesado/a. Les dices:

d) ¡¡Y vosotros lo que sois es unos hijos de puta!!

e) ¡¡Y ustedes lo que son es unos hijos de puta!!

f) ¡¡Y ustedes lo que sois es unos hijos de puta!!

The neighbors next door have been playing music very loud all night because they have a party at home. They have done it before, many times. You can’t sleep because of the music and you have to work the next day. Furthermore, your mother, who is sick, can’t sleep either. You knock at their door and nicely ask them to lower the volume, since your mother is sick. The refuse to do it, and tell you not to bother them, that you are nuts. You tell them: “And you guys are motherfuckers, that’s what you are!!”

7) Tu sobrino salió ayer de viaje a Londres y no sabes aún si ha llegado bien a

Londres, no ha llamado todavía. Le dices a un amigo:

a) Espero que haya llegado bien

b) Espero que llegara bien

217

c) Espero que haiga llegado bien

Your nephew left on a trip to London yesterday. You don’t know if he arrived

safe and sound to London, he hasn’t called yet. You tell a friend: “I hope he

has arrived safe and sound”

8) Quieres ir al cine con tus amigos, a ver la película Batman vs. Superman pero

no sabes si tus amigos la han visto. Les preguntas:

d) ¿Ustedes la han visto?

e) ¿Vosotros la habéis visto?

f) ¿Ustedes la habéis visto?

You want to go to the theater with your friends, to watch the movie Batman vs.

Superman, but you are not sure whether your friends already saw it or not.You

ask them: “Have you guys seen it?”

9) Estás en una comida de Navidad y los familiares te preguntan si quieres

comer más. Les dices:

a) No, gracias, ya no quiero más nada

b) No, gracias, ya no quiero nada

c) No, gracias, ya no quiero más

d) No, gracias, ya no quiero nada más

You are at a Christmas lunch and your relatives ask you if you want to eat

more. You tell them: a) and d) No, thanks, I don’t want anything else b)

No, thanks, I don’t want anything c) No, thanks, I don’t want more

10) Tus cuñado/a y tu hermana/o se quejan todo el tiempo de que se cansan mucho

al subir las escaleras de tu casa. Les dices, para que no se quejen tanto:

d) Ustedes lo que tienen que hacer es un poquito de ejercicio.

218

e) Ustedes lo que tenéis que hacer es un poquito de ejercicio.

f) Vosotros lo que tenéis que hacer es un poquito de ejercicio.

Your brother/sister in law and your sister/brother complain all the time

because they get very tired when they have to go up the stairst of your

house. In order to make them stop complaining so much, you tell them:

“What you should do is some exercise”.

11) Cuando estabas en la escuela no estudiabas porque no te gustaba. Ahora te

arrepientes porque podrías tener un trabajo mejor. Le dices a tu compañero de

trabajo:

a) Ojalá hubiera estudiado más cuando estaba en la escuela.

b) Ojalá haya estudiado más cuando estaba en la escuela.

c) Ojalá fuera estudiado más cuando estaba en la escuela.

12) Vas conduciendo y ves que una pareja joven, en otro coche, de manera

peligrosa, te adelanta y se pone delante de ti. En un semáforo en rojo te pones

en el carril de al lado de ese coche y aprovechas para gritarle al conductor por

su irresponsabilidad, pero él te dice que van al hospital, que su mujer,

embarazada, ha roto aguas. Ves que es verdad y te sientes avergonzado. De

todas formas, ellos te piden perdón por adelantarte así. Les pides que te

perdonen ellos; les dices:

f) No, perdonarme ustedes

g) No, perdonadme vosotros

h) No, perdónenme ustedes

i) No, perdonadme ustedes

j) No, perdonarme vosotros

219

You are driving and see a young couple, in another car, pass you in a dangerous way.

They come in front of you with their car. Later, when stopped because of a red light, you get to the lane next to the car that just passed you. You use the opportunity to yell at the driver because of his irresponsibility. He tells you that they are going to the hospital, that his pregnant wife just broke water. You see that it is true and you feel ashamed. In any case, they apologize to you for having passed you like that. You apologize to them; you tell them: “No, forgive me, you guys”.

13) Tu hermano quiere sacar buenas notas pero no quiere estudiar mucho. Le

aconsejas que estudie mucho:

a) Contra más estudies, mejores notas tendrás

b) Cuanto más estudies, mejores notas tendrás

c) Contrimás estudies, mejores notas tendrás

d) Mientras más estudies, mejores notas tendrás

Your brother wants to get very good grades but does not want to study

much. You advise him to study a lot: “The more you study, the better

grades you’ll get”

14) Tus hijos quieren que llames al electricista para reparar un problema que hay

en la casa. Quieres que llamen ellos. Les dices:

k) Llamadlo vosotros

l) Llamarlo ustedes

m) Llámenlo ustedes

n) Llamarlo vosotros

o) Llamadlo ustedes

220

Your children want you to call the electrician to fix some problem that you have at home. You want them to call instead, so you tell them:“You guys call him”

15) Fuiste por la mañana al médico, pero te atendió muy tarde por la cantidad de

enfermos que estaba delante de ti. Llegas tarde a trabajar y le explicas a tu

jefe:

a) Lo siento, es que estaba en el médico y habían muchos enfermos hoy

b) Lo siento, es que estaba en el médico y había muchos enfermos hoy

c) Lo siento, es que estaba en el médico y encontré muchos enfermos hoy

16) Tu hijo es muy buen estudiante. Estás reunido con dos de sus profesores, que

son catedráticos. Les dices que piensas que tu hijo podría llegar lejos también,

como profesor, y les preguntas por su opinión. Les dices:

d) ¿qué pensáis ustedes?

e) ¿qué pensáis vosotros?

f) ¿qué piensan ustedes?

Your son is a good student. You are in a meeting with two of his

professors, who are full professors. You tell them that you think that your

son has a great future ahead too, as a professor. You ask them about their

opinion on the matter. You tell them: “What do you think?”

221

Appendix B: Oral Elicitation Task 1

1) Tu amigo te dice que por qué no te compras un coche, pero tú no tienes

dinero, aunque te gustaría comprar uno. Le dices que...

A friend of yours asks you why you don’t buy a car, but you don’t have the

money. You tell him that...

2) Unas vecinas, dos señoras mayores, se están quejando de que la boda de

Emilio y Belén sea al aire libre porque va a hacer mucho calor. Emilio les dice

a las vecinas pues que se lleven ellas una pamela y una cantimplora.

(Temporada 5, episodio 12, Aquí no hay quien viva.)

https://www.youtube.com/watch?v=CZ8eO3yj6KE (min. 11:20; mute: 11:30)

Some neighbors are complaining about the fact that the wedding of Emilio and

Belén will take place outdoors and it will be very hot. Emilio tells the

neighbors to just bring a Pamela hat and a canteen with them. (Season 5,

episode 12, No one could live here)

3) Vas a montar en un taxi con más gente. El taxista pregunta que cuántas

personas. Le dices que cuatro.

You are getting a taxi with more people. The taxidriver asks you how many

people. You tell him that four.

4) Publicas una foto en Facebook y la gente de tu facebook te dice que eres

guapísimo/a. Te sientes halagado y les dices que ellos sí que son guapos.

You post a picture of yourself on Facebook and your facebook friends tell you

that you are so beautiful. You feel flattered and you tell them that they are the

beautiful ones.

222

5) Tu hijo acaba de hacer un examen, tú, que quieres que tu hijo saque buenas

notas, le dices a tu marido/mujer, con respecto al examen, que esperas que el

niño….

Your son has just done an exam. You want your son to get good grades. You

tell your husband/wife, with respect to the exam, that you hope that your

boy…

6) Estás hablando con tus primos sobre las pasadas elecciones generales. Ellos

han votado a un partido diferente al tuyo y la discusión se vuelve acalorada, no

estáis de acuerdo para nada. Te insultan, te dicen: “¿¿cómo has podido votar a

esa mierda de político y de partido, es que no piensas con la cabeza, eres tonto

o qué te pasa??” Les dices que los tontos son ellos y su partido es el malo, y

que igual que ellos pueden votar a quien ellos quieran tú puedes votar a

quien tú quieras.

You are talking with your cousins about the past general elections. They voted

for a different party than you. The argument becomes heated, you don’t agree

at all. They insult you, they tell you: “How could you vote for that shitty

politician and party, do you even use your head? Are you stupid or what is

wrong with you?” You tell them that they are the stupid ones, and the same as

they can vote for whoever they want to, you can vote for whoever you want too

as well.

7) Quieres entrar a un baño de un bar, pero hay que pedir la llave al camarero. Le

dices a tu amiga que se lo diga ella.

You want to go to the restroom at a bar, but you have to ask the server for the

key. You tell your friend to tell him.

223

8) Tus vecinos son muy simpáticos y te llevas muy bien con ellos, pero no son

muy limpios y siempre te toca a ti y tu familia limpiar las escaleras y el portal

de tu bloque. Un día te encuentras a varios de ellos al bajar las escaleras.

Hablas con ellos y les dices amablemente que sería bueno que no sólo tú y

tu familia, sino que también ellos limpiaran la escalera y el portal, para

mantenerlo todo limpio.

Your neighbors are very nice and you get along with them very well, but they

are not very clean. You and your family have always to clean the stairs and

the hallway. One day you run into several of you neighbors when going down

the stairs. You talk to them and nicely and tell them that it would be good that

not only you and your family but also, they cleaned the stairs and the hallway,

to keep everything clean.

9) Ayer fuiste a una cena. Te preguntan que cuánta gente asistió a la cena. Dices

que mucha.

You went to a dinner last night. You are asked how many people attended the

dinner. You say that a lot.

10) Hay una cola larguísima para pagar en el Mercadona. Hay un grupo de

chavales jóvenes que se te cuela delante y se lo dices enfadada, ellos dicen que

no se han colado, que era su turno, que les tocaba a ellos, no los crees y

discutes con ellos, piensas que son unos descarados. La cajera del Mercadona

te dice después que, en efecto, les tocaba a ellos, que estaban de antes pero que

estaban en otra caja que cerró y los pasó ahí. Avergonzada, les dices a los

chavales que es verdad que ellos tenían la razón y tú estabas equivocada, y

les pides que te disculpen.

224

There is a huge line to pay at Mercadona. There is a group of young guys that

have come in front of you in the line. You tell them that angrily but they say it

was their turn. You don’t believe them and so you argue with them, you think

they are shameless. The Mercadona cashier tells you later than, actually, it

was their turn; she explains that they were waiting for another cashier who

closed and so she told them to come where you are. Feeling ashamed, you tell

them that they were right and you were wrong, and you apologize.

11) Tu hijo insiste para que le des más dinero para salir. Le dices que va a ser

peor si sigue insistiendo, que menos va a pillar.

Your son is intensively trying to get you to give him more money to go out. You

tell him that the more he asks for it the worse it will be, the less money he will

get.

12) En una boda, los amigos de los novios le hacen un vídeo en el que los

felicitan. Un amigo les dice en el vídeo que sean ellos muy felices que ellos

se lo merecen. (Temporada 3, episodio 10, Malviviendo)

https://www.youtube.com/watch?v=ihaNQhFQDJ8&index=10&list=PLgrmcS

DxLrZmNCzD1p1E9VpIkQFP827WI (Min. 1:33:40; mute: min. 1:33:57)

At a wedding, friend of the bride and the groom make a video for them with

best wishes messages. One of them says that he whishes that they are very

happy and that they deserve it. (Season 3, episode 10, Bad Living)

13) Tu hermano tiene una novia. Le dices a tus amigos que es muy apañada.

Your brother has a girlfriend. You tell your friends that she is really nice.

225

14) Tus tíos te están contando que fueron de viaje a Madrid el fin de semana

pasado. Les dices que tú fuiste una vez a Madrid en tren y les preguntas que si

ellos fueron en avión o en tren.

Your uncle and aunt are telling you about their trip to Madrid last weekend.

You tell them that you went once to Madrid by train and you ask them if they

traveled by train or by plane.

15) Ves un bebé hermoso en el Mercadona. Le dices a la madre lo hermoso que

es su bebé.

You see a baby in Mercadona. You tell his mother how beautiful her baby is.

16) Estás en el hospital llevando a tu anciana madre, que está enferma. Los

doctores te piden permiso para hacerle un tratamiento que necesita, porque es

peligroso. Les dices que lo que ellos decidan, que tú no sabes, que ellos son

los expertos.

You are at the hospital, with your old mother, who is sick. The doctors ask you

for permission to apply a dangerous treatment to her, a treatment she needs.

You tell them that whatever they decide, that you don’t know, that they are the

experts.

17) Un amigo de tu hijo viene a tu casa a preguntar por él, pero le dices que tu

hijo, que se llama David, ya se ha ido.

A friend of your son comes to your house asking about your son, but you tell

him that your son, who is called David, already left.

18) En una terapia de grupo, para drogadictos. El que dirige la terapia les dice a

todos que mientras él les habla ellos cierren los ojos. (Temporada 3,

episodio 2, Malviviendo)

226

https://www.youtube.com/watch?v=ylFH5DEpP6w (min. 23:1; mute: min. 24:

05)

It’s a group therapy for drug addicts. The therapy leader tells everybody else

to close their eyes while he talks to them.

19) Estás viendo en tu casa fotografías familiares, te pregunta tu madre si has

visto ya la de la boda de tu prima.

You are at home looking at family pictures. Your mother asks you if you have

already seen the picture of your cousin’s wedding.

20) Tus hijos/nietos/compañeros de piso están haciendo mucho ruido en casa y no

te dejan descansar. Te quejas y te dicen que te calles. Les dices que se callen

ellos.

Your children/grandchildren/roommates are being noisy at home and you

cannot get a rest. You complaint about it and they tell you to shut up. You tell

them to shut up.

21) Te gusta ir a caminar pero ayer no fuiste porque tenías muchas cosas que

hacer. Hoy te arrepientes de no haber ido a caminar, y se lo dices a tu

marido/mujer/pareja.

You like going for a walk, but you didn’t go yesterday since you were very

busy. You regret that today and you say that to your husband/wife/partner.

22) Tus amigos te dicen que has estado todo el verano sin trabajar ni hacer nada.

Tú sabes que ellos tampoco han hecho nada. Entonces, les dices que, claro,

que como que ellos no han descansado nada este verano, que como que

ellos no han parado de trabajar ni nada este verano…

227

Your friends are telling you that you didn’t work at all this summer, that you

didn’t do anything. You know they didn’t do anything either this summer. You

tell them that sure, that if they didn’t rest at all this summer, if they didn’t stop

working this summer at all.

23) Una persona te pregunta por la calle que dónde está correos. Está justo detrás

de la persona que te pregunta. Se lo dices.

Someone asks you in the street where the post office is. It is actually right

behind the person asking about it. You tell him.

24) Un grupo de vecinas está planeando la despedida de soltera de Lucía. Una de

ellas no está conforme con los planes que hacen. Las demás le dicen que qué

quiere que hagan. La vecina que no está conforme con los planes de despedida

de soltera les dice a las demás que es que ellas sólo buscan lo típico.

(Temporada 3, episodio 29, Aquí no hay quien viva).

https://www.youtube.com/watch?v=uRfWn4EsI-s (min. 9:08; mute min. 9:39)

A group of neighbors is planning the bachelorette’s party for Lucía. One of

them does not agree with the plan. The others ask her what she wants them to

do. She says that the thing is that they just like very common, ordinary things

(Season 3, episode 29, No one could live here).

25) Te gusta bailar y sabes bailar muy bien. Unos amigos te dicen que a ti no te

gusta ni sabes bailar. Les dices medio molesto…

You like dancing and you can dance very well. Some friends tell you that you

don’t like it or can dance. You tell them, annoyed…

228

26) Dos amigos critican la fealdad de la novia de otro. Él les dice que sabe que

ella no es muy guapa pero que ellos, sus amigos, tampoco es que sean

Kevin Costner o la gran cosa. (Temporada 1, capítulo8, Malviviendo)

https://www.youtube.com/watch?v=i--XduiLsI0 (Min. 32:14 mute: 32:20)

Two friends are saying that the other’s girlfriend is ugly. He acknowledges

that she is not very beautiful but tells them that they are not Kevin Costner

either or very beautiful either. (Season 1, episodie 8, Bad Living)

27) Tienes mucho calor este verano, le dices a alguien que no lo aguantas, le

dices…

You are very hot this summer. You tell somebody that you can’t stand it, you

tell him…

28) Vas paseando con tu familia por Granada. Te preguntan que dónde vamos y

qué hacemos, les dices que ellos, que son los que te han traído a Granada de

paseo, que ellos sabrán, que lo que quieran ellos, que lo digan ellos.

You are going for a walk with your family in Granada. They ask you where to

go and what to do. They brought you there, so you tell them that they should

know, that whatever they want.

Oral Elicitation Task 2

1) Te llama alguien desconocido para hacerte una encuesta por teléfono. Les

dices cualquier excusa, que no tienes tiempo, para no hacerla.

Some unknown person calls you to make you a survey via the phone. You tell

him some excuse, such as you don’t have time, to not do the survey

229

2) Una buena amiga/un buen amigo tuya te dice que está embarazada, que va a

tener un hijo, que va a ser padre/madre. La/lo felicitas.

A good friend of yours tells you that she is pregnant, you congratulate her.

3) Estás con amigos, viendo en qué restaurante vais a cenar, y les preguntas a

tus amigos si ellos quieren ir a una pizzería, que a ti te gustaría ir a una

pizzería, que ellos adónde quieren ir.

You are with friends trying to decide where to have dinner. You tell them you

would like to go to a pizza place and ask them if they want to go to have pizza

or where they want to go.

4) Estás en un restaurante para cenar, pero cuando te sirven ves que hay pelos en

la comida, en la sopa. Le dices al camarero que te cambie ese plato de

comida o que te vas sin pagar.

You are at a restaurant to have dinner, but when they serve you your food, you

notice there is hair in the soup. You tell the server to change your food or that

you will leave without paying.

5) Vas conduciendo por una rotonda y estás para salir de la rotonda, pero te

rebasan 2 taxistas en un coche por la izquierda, aunque tú tenías preferencia,

con riesgo de accidente. Les pitas a los taxistas y encima te sacan el dedo los 2

y te llaman “hijo de puta” y te dicen que mires por dónde vas y qué aprendas a

conducir, que dónde te sacaste el carnet. Te enfadas mucho y les dices que

los “hijos de puta” son ellos, que son ellos los que tienen que mirar por

dónde van, que son unos sinvergüenzas y que ellos son los que tienen que

aprender a conducir, no tú.

230

You are driving through a roundabout, but two taxi drivers pass you on your

left, even though you had the priority. They pass you dangerously, so you

honk, but they give you the middle finger, call you “motherfucker”, and tell

you that you should look where you are going, and that you should learn to

drive, and that where did you get your license. You get very mad, and tell them

that they are the motherfuckers, that they are the ones who have to look where

they are going, that they are assholes, that they are the ones who need to learn

to drive, not you.

6) Ya te ha llamado el de Sevillana Endesa muchas veces para actualizar tu

factura y si no estás con ellos, hacerte una oferta, pero ya estás harta de que te

llame sin parar y no te quieres ir con Endesa. Se lo dices educadamente, que

no te interesa y que no llame más.

The electric company Sevillana Endesa has called you many times to update

your bill or to make you an offer, and you have had enough of their calls. You

tell them nicely not to call you anymore, that you are not interested,

7) Haces de comer para tus cuñados y te dicen que eres muy buen cocinero/una

cocinera estupendo/a, que está todo muy bueno. Ellos te han hecho de comer

muchas veces y cocinan muy bien también. Les dices que ellos sí que

cocinan bien, que tú más o menos.

You prepare food for your siblings in law and they tell you that you are such a

good cook, that everything is delicious. They have cooked for you before and

you know they cook very well. You tell them that you are a mediocre cook, that

they are the ones who cook very good.

231

8) Una vecina cotilla te pregunta que a quién votas tú, y no quieres decírselo. Le

dices que lo sientes pero que es privado eso.

A gossipy neighbor asks you what political party you vote for, but you don’t

want to tell her. You tell her that you are sorry but that that is private.

9) 2 muchachillos jovencillos, testigos de Jehová, te paran por la calle para

hablarte de su religión y tratan de convencerte. No tienes ganas de escucharlos

y tienes prisa, pero te dicen que cómo no te puede interesar lo que te van a

decir y se ponen insistentes en tratar de convencerte de sus creencias. No

quieres ser maleducado con ellos, que son muy educados, pero les dices,

educadamente, que, igual que ellos tienen sus creencias, tú tienes las

tuyas, diferentes, y que no te interesa, pero que tú respetas sus creencias,

y que ellos respeten las tuyas, y que te disculpen que te tienes que ir.

Two Young men, Jehovah’s Witnesses, stop you in the street to talk to you

about their religion and to try to persuade you to convert. You don’t feel like

listening to them and you are in a hurry, but they are persistent. However,

they are very polite and nice, so you nicely and politely tell them that you are

sorry but you are not interested and have to leave, and that the same as you

respect their beliefs, you have your own ones, and that they should respect

them too.

10) Vienen los del gas a tu casa a decirte que tienen que cambiarte una goma, pero

sabes que hay mucha estafa con eso, y no te fías de que sean los del gas de

verdad. Les dices que no necesitas cambiar la goma.

232

Gas company staff come to your home to tell you that they need to change a

gas pipe. However, you don’t trust them since you know there is a lot of scam.

You tell them that you don’t need to change the pipe.

11) Un hombre te pide dinero por la calle, pero tú piensas que no es en realidad

para comer sino para droga. Le dices…

A guy asks you for money in the street, but you think he wants the money for

drugs instead of food. You tell him…

12) Te das cuenta de que tus padres ven unos programas de televisión malísimos,

que tú no verías en la vida. Les preguntas extrañado y mosqueado que si

ellos ven esos programas tan malos.

You realize that your parents are watching a very bad TV program, that you

would never watch. You ask them, bewildered, if they watch those TV

programs habitually.

13) Ves que tu madre/padre tiene mucho trabajo en la casa, pero no te quiere pedir

ayuda para no molestarte, te ofreces a ayudarla/lo de todas formas, se lo

dices.

You realize that your mother has a lot of house chores to do but she doesn’t

want to ask you for help not to bother you. You offer to help her anyways.

14) Estás en la fila para entrar en el autobús. Una pareja joven te quiere dejar pasar

delante, pero tú les dices que no, que entren ellos.

You are in the line for getting in the bus. A young couple in front of you tell

you that you can go first, but you tell them that it is not necessary, that they

can go.

233

15) Ves que este año está la playa muy mal, muy sucia, y tú que querías ir todos

los días, así no vas a poder. Me dices a mí la queja.

You realize that the sea is polluted this year and you wanted to go to the beach

everyday. You tell me your complaint.

16) Tus hermanos te piden perdón por no haberte llamado ni visitado mucho este

año, que no os habéis visto casi nada, pero tú sabes que tú tampoco has estado

muy pendiente de ellos, porque has estado muy ocupado, entonces les dice

que no, que te perdonen ellos, que tampoco has llamado mucho ni nada,

que tú también has estado muy ocupada.

Your siblings apologize because this year they haven’t paid much attention to

you, they haven’t called or visited, because they have been very busy. You said

that they are the ones who have to forgive you because you have also been

busy and haven’t called or visited either.

17) Antiguos compañeros de la escuela han organizado una cena para reunirse

todos después de muchos años. Te da pereza ir, le dices a la organizadora

que es que tienes mucho trabajo/que alguien de tu familia está enfermo.

Old schoolmates have organized a dinner for a class reunion. You don’t feel

like going and tell the organizer any excuse, such as someone from your

family is sick.

18) Tus compañeros de trabajo te decían que no ibas a ser capaz de terminar un

proyecto encargado por el jefe a tiempo, pero sí lo consigues, entonces les

enseñas a todos el trabajo terminado, diciéndoles que miren, que para que

ellos vean, cómo sí has sido capaz, que ellos decían que no y sí has podido.

234

Your workmates told you that you were not going to be able to finish a proyect

in time. Eventually, you are able to finish it, so you show them the proyect and

tell them to look how you were able to do it even though they were telling you

otherwise.

19) Tu hermano tiene la música puesta muy fuerte, a mucho volumen y no te deja

concentrarte en algo que estás haciendo, le dices que le quite voz.

Your brother is playing the music very loud and doesn’t allow you to focus, so

you tell him to lower the volume.

20) Tus primos te han hecho ir a la playa de noche a pasar las fiestas de San Juan,

a pasar la noche en la playa, y tú no tenías ganas, y ahora están aburridos, y te

preguntan qué hacer ahora. Tú les dices que ellos son los que te han hecho ir

de San Juan, que ahora que ellos hagan lo que ellos quieran.

Your cousins made you go to spend the San Juan festivities at the beach even

though you didn’t want to. Now that you are there, they are telling you that

they are bored and that they don’t know what to do. You tell them that since

the made you go, now they can do whatever they want.

21) Tu hijo te dice que lo dejes salir hasta más tarde por la noche y le dices que

no, que es por su bien, que aún es muy chico para eso.

Your son asks you for permission to stay later at night. You tell him that it’s

for his own good, that he is very young to stay until late.

22) Tus cuñados/as te quieren llevar de parranda; sabes que son muy fiesteros para

ti, beben mucho, y no te fías mucho de ir con ellos, y les dices que ellos es

que son muy peligrosos.

235

Your siblings in law want to take you to party; you know that they are party

animals and that they drink too much so you don’t trust going out with them.

You tell them that you don’t trust them and that they are very dangerous.

23) Tu tía te dice que te ha estado llamado y no has contestado. Imaginas que es

porque tu teléfono no ha estado funcionando bien. Se lo dices.

Your aunt tells you that she has been calling you and you didn’t answer. You

tell her that it might be because your phone hasn’t been working properly.

24) Tus compañeros de trabajo te dicen de broma que eres muy lentillo en tu

trabajo, aunque ellos son en realidad más lentos y lo saben; les dices que es

que ellos van muy rápido.

Your workmates are telling you that you are very slow, but they are joking

because they know that you are faster than them. You tell them, also joking,

that the thing is that they are very fast.

25) Estás intentando ver la tele pero tu nieto pequeño se pone todo el rato delante

jugando. Le dices que se aparte, que no puedes ver el fútbol.

You are trying to watch TV but your grandson is all the time in front of the TV,

playing. You tell him to move so you can watch your program.

26) Tienes 3 hijos ya mayores de edad, de 20 años, pero que no quieren hacer

nada, ni estudiar ni trabajar; encima te das cuenta ahora de que nada más te

han estado diciendo mentiras este año y sacándote mucho dinero, los 3. Les

dices a los 3 muy enfadada que se acabó, que estás muy decepcionada con

ellos, que ahora se busquen la vida ellos, que tú ya no les das más dinero

ni nada.

236

You have three kids, over 20-year-old the three of them, but they don’t want to

work or study. Moreover, you realize that they have been lying to you, to get

money from you. You are very mad at them and tell them that it is over, that

you are very disappointed in them, that you are not going to give them more

money and that now they have to earn their own livings.

27) Te das cuenta que frutas que compraste en el Mercadona están pasadas, se las

llevas de vuelta allí al encargado y le pides que te devuelva el dinero.

You realice that the fruits you bought at Mercadona supermarket are too ripe,

in bad state, so you ask the staff to give you the money back.

28) Dejas pasar delante tuya a dos hombres/dos mujeres muy guapos/as, te dan las

gracias y les dices, por echarles un piropo, que las que ellos/as tienen.

You let two beautiful men/women pass in front of you. They say thank you to

you and you tell them that they don’t have to say thank you to you, that you

thank them as they are so beautiful.

29) Vas a pagar en el Mercadona pero te das cuenta de que no has traído dinero

bastante. Le pides a la cajera que te aparte las cosas que vas a tu casa que

está al lado, a por el dinero.

You are about to pay at Mercadona supermarket but you realize that you

didn’t bring money enough, so you ask the cashier to save the things you

bought for you while you go home in a second to grab your wallet.

30) Les estás contando a unos vecinos tuyos que vas a ir a un concierto, de un

cantante que te gusta en tu pueblo, el fin de semana que viene, les dices que tú

vas a ir y les preguntas que si ellos van a ir también.

237

You are telling our neighbors that you are going to a concert next weekend, to see the singer/band you like. You tell them you are going and ask them whether they are going too.

238