Language Contact and Dialect Contact: Cross-Generational Phonological Variation in a Puerto Rican Community in the Midwest of the United States

Home , Belizean Spanish, Canarian Spanish, Central American Spanish, Language contact, New Mexican Spanish, Nuyorican

LANGUAGE CONTACT AND DIALECT CONTACT: CROSS-GENERATIONAL PHONOLOGICAL VARIATION IN A PUERTO RICAN COMMUNITY IN THE MIDWEST OF THE UNITED STATES

DISSERTATION

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

Michelle F. Ramos-Pellicia, B.A., M.A.

****

The Ohio State University 2004

Dissertation Committee: Approved by ______Professor Brian D. Joseph, Co-Adviser Co-Adviser Linguistics Graduate Program Professor Donald Winford, Co-Adviser

Professor Richard Janda Approved by ______Co-Adviser Linguistics Graduate Program

This project investigates a variety of Spanish found in Lorain, Ohio, focusing on the description of variation at the phonological level. The description considers the dynamics of the interaction of two minority dialects, Lorain Puerto Rican Spanish

(henceforth LPRS) and Mexican American Spanish (hereafter MAS), in the context of

English as a majority language. It seeks to find an explanation for phonological variation

across three different generations in the LPRS community by examining five variables

(raising of /e/ and /o/, and the realization of /dΩ/, /r/, and /b/). The role of the youngest

generation is given particular attention. Previous studies (Eckert 1997, 2000; Kerswill

1994; Kerswill and Williams 1994) on language variation have shown that this group

leads the changes their linguistic variety will undergo or eliminates the language

completely. This project also explores the phonology of rural Island Puerto Rican

Spanish (henceforth IPRS) transported to Lorain to determine if LPRS and IPRS display

similar or different patterns of use of the variables.

ii One hundred four people representing the Puerto Ricans and the Mexicans in Lorain, the

Puerto Ricans of the island and the American English speakers of Lorain (hereafter AE)

were recorded. The groups are also subdivided by generation and gender. All subjects

read a word-list and a paragraph and participated in a short informal conversation with the interviewer. Also, a long-term observation of the community was conducted. The phonological variables were correlated with extralinguistic factors using VARBRUL.

LPRS shows almost nonexistent (e) and (o) raising, preserving the pattern found in IPRS.

Lorain MAS and LPRS are converging in their use of (j), but the merger varies within each age and generation. The two groups are diverging in their use of (r), presenting different degrees of divergence depending on generation and age group.

The Spanish spoken by second and third generation LPRS speakers shows influence from

AE in the use of a labiodental [v] for (b) and a retroflex [] for (r). Indirectly, AE

influences first generation speakers not proficient in AE through contact with those

speakers of the first generation with more exposure to AE.

iii In general, the results reveal that LPRS exhibits continuity from IPRS preserving many of

the features transmitted from IPRS though it also diverges from the latter to some extent.

Although predictions are speculative, possibly in the future, due to the pressures of MAS and AE on LPRS, this pattern will change and LPRS will become a dialect different from

IPRS, if it survives at all.

Dedicated to my multilingual family…

my Grandma

Emerance Louise Hilversum Thompson, who unconsciously initiated me in linguistics

and

my son

Antonio José Rodríguez-Ramos, a productive and inquisitive bilingual speaker

v ACKNOWLEDGMENTS

Growing up in Puerto Rico, I was exposed mainly to two different languages, Spanish

and English. However, as I shared my childhood years with my beloved Grandma, I

discovered through her nurturing a language that is dear to my heart and yet it was lost in

my generation. Surinamese Dutch was used at home with my Grandma, my Mom and

her family. Unfortunately, my elders decided for me that I was not to continue my

acquisition of Surinamese Dutch because they thought this language would confuse me as

I was starting my schooling in English. Thus, whenever around me, my relatives made a

conscious effort to use Taki Taki, English, but mostly Puerto Rican Spanish.

As I continued growing up, I also discovered that my Dad’s side of the family had very

interesting language interactions. Having grown up in the military, my Dad and his

family were exposed to Spanish, English and French. I witnessed many interactions

where my uncles, my Dad and his parents will use these three languages in a magnificent example of code switching. I was always amazed at how quickly they switched from one language to the next just to make themselves understood.

vi My curiosity to explore language led me to study Linguistics at The Ohio State

University. Many are the people who expanded my world during my graduate school years. I will always be indebted to my Professors Brian D. Joseph and Donald Winford.

Their knowledge and insightful comments expanded my original project into what is now

this dissertation.

I am also thankful to Professors Richard Janda, Terrell Morgan, Mary Beckman,

Elizabeth Hume, Marcia Farr, Dennis Preston and Patricia Lunn who helped me at

different stages of developing this project and who also raised questions that I did not

foresee as I was collecting and analyzing the data.

I also want to thank the members of the Changelings discussion group. It was in this

warm and welcoming environment where I not only presented the earlier results of my

dissertation but also learned of the projects that scholars interested in language variation

and change were conducting.

Thank you to my classmates, colleagues and friends at the Department of Linguistics…

Shelome Gooden, Misun Seo, Steve Hartman-Keiser, Tsan Huang, Hope Dawson, Robin

Dodsworth, Ila Nagar, Craig Hilts, Wes Collins, Alejandra González-Pérez, Grant

McGuire, Na’im Tyson, David Durian, Anton Rytting, Maeve Eberhardt, Amanda

Boomershine, Robin Dautricourt, Magdalena Mejía who encouraged me at different

stages of my writing.

vii Also, thanks to Yanitza Marrero who introduced me to her relatives and friends in Lorain, to the Carrión Family, the Bigio Family, the Reverón Family, the Meléndez Family, the

Mendiola Family, the Ortiz Family, the DeGracia Family, the Maldonado Family,

Georgina and Salvador Carrión, Tomás Cabassa, Carmen and Wilson Meléndez,

Rigoberto and Judy Reverón, Jimmy and Evelyn Carrión, Trinidad Amador, Sister

Theresa Stegman, Sister Elisa Bonano, Christian Doyle, Rev. William Thaden, Rosa

Ramos, Milagros Pellicia, Ana Capetillo, Edna Colón Torres, Agustín López Oliveras,

Noa Rivera, Wilson Acosta, Wanda Aponte Rodríguez, Jorge Rivera and all the

American English speakers, Mexicans and Puerto Ricans in Lorain and all the Puerto

Ricans in Puerto Rico who participated in this study and who so kindly received me in their homes and patiently answered my questions. Without them, this project would have never been possible.

Thanks to my friends in Ohio… Leigh and Lee Österling, Karen Lynch, Socorro

Moores, Odemaris Irizarry, Lillian Olivieri, Rohini and Uday Dandavate, Celia James,

Steve Grill, Donna Villareal, Francie Buschur, Gilberto Serrano, Gary Fourman, Cristina

Pérez, Eileen Butler, Mary Flanagan, Luisa Lara, Eleonor Opitz, Beatriz Alvarado. All of them opened their homes and welcomed my family during the years I spent in

Columbus, Ohio.

To my friends in Puerto Rico… Rosie Aulet and Madalene Valentín who despite the distance kept up with our friendship. ¡Mil gracias!

viii To Christine Heinze and Christina DiLorenzo and all my past, present and future students

from whom I learn every day… They make me want to be a better teacher and a better

scholar.

I am grateful for the funding I received from the following sources: The Ohio State

University Department of Linguistics, Chair Peter Culicover made available funding to

cover my travel expenses to Lorain, The OSU College of Humanities Graduate Research

Council Small Grant covered for equipment expenses, The Ohio State University Council of Graduate Students Ray Travel Award for Scholarship and Service (formerly

Professional Development Fund, PDF) covered for travel expenses and The OSU

Graduate School Alumni Grant for Graduate Research covered my travel expenses to

Puerto Rico.

And last, but not least, my family, here in the United States and in Puerto Rico…

Thanks to Mom who always encouraged me with her questions about language and the problems Puerto Ricans face in the Island and in the mainland. To her, I owe many useful conversations that expanded my curiosity and my thirst to understand more about the different aspects of language in interaction and bilingual education.

ix To Dad and all his multiple interferences of English in his Spanish that always made me laugh with him, to his wit and his talent to switch back and forth between his tongues.

His positive attitude and encouragement at very obscure moments helped me in accomplishing many of my goals.

To my sisters Christine and Talía who were a constant source of encouragement during my journey through Graduate School, both of them are a wealth of bilingualism and productive code switches and borrowings.

To the memory of my Abuela Emi, who started me early on the road of language.

And to my son, who has experienced so many different languages since before he was one year old and who has become an avid code switcher, knowing who to switch with, and whether to explain a word to a person in Spanish or English. It is my hope that he continues being as inquisitive as he is now and that his Spanish, English and Spanglish abilities keep expanding.

x VITA

September 20, 1972………………….Born, Río Piedras, Puerto Rico

1995………………………………….B.A. Hispanic Studies, Minors: French,

English, Linguistics, University of Puerto Rico, Río

Piedras

1999…………………………………M.A. Linguistics, The Ohio State University

PUBLICATIONS

2001. ‘Oh, you want TETE?’ (‘Oh, ¿quieres TETE?’) Co-operación risas embarazosas y gestos en un grupo de apoyo para madres lactantes. Lengua, Discurso y Texto. Proceedings of the I Simposio Internacional Análisis del Discurso. Visor Libros. 1677-1696.

1999. Progressive Constructions in the Spanish Spoken in Puerto Rico. Ohio State University Working Papers in Linguistics. Volume 52. Language Change. Columbus, Ohio: The Ohio State University. 97-112.

1995. Childs, Monyette; Dale Fritter, Jennifer Harr, Michael Kester, Heidi Ramírez and Michelle Ramos-Pellicia. Early Intervention Programs and Public Education: Should TRIO Go? The Odyssey, ed. by Pai-Ling Yin and Aida Orenstein- Cardona, 80-83. Washington, D.C.: The Harry S Truman Scholarship Foundation.

1995. Spanish-English Code-Switching among Young Mexican Americans from North Lansing, Michigan. Contornos. 10-11.

xii 1995. Teaching English in Puerto Rican Schools: Public Policy Analysis. Contornos. 8-9.

FIELDS OF STUDY

Major Field: Linguistics

xii TABLE OF CONTENTS

Page Abstract………………………………………………………………………………...…ii Dedication…………………………………………………………………………………v Acknowledgments……………………………………………………………………..…vi Vita………………………………………………………………………………...…..…xi List of Tables………………………………………………………………….…...…....xvi List of Figures……………………………………………………………………...….…xx

Chapters:

1. Introduction……………………………………………………………………..…1

2. Theoretical Issues: Linguistic and social outcomes of language contact and dialect contact……………………………………………………………………12

2.1 Introduction…………………………………………………………...….12 2.2 Overview of the pressures, processes of formation and outcomes emerging out of language/dialect contact…..…………………………………...... 13 2.3 Superstrate influence……………………………………..……………....15 2.4 Processes……………………………………………………………...….15 2.4.1 Dialect accommodation…………………………………………….……15 2.4.2 Convergence……………………………………………………………..18 2.4.3 Results of dialect accommodation: levelling and konéization……….…19 2.5 The result of dialect accommodation: dialects and languages converging…………………………………………………………….....22 2.6 The role of the social networks……………………………………..…...24 2.7 The Lorain Puerto Rican community…………………………………....25 2.8 Summary……………………………………………………………...... 26

3. The Community: social and linguistic description of the Lorain Puerto Rican community……………………………………………………………….……...28

3.1 Ethnography of the Puerto Rican community of Lorain…………….….28

xiii 3.2 Linguistic description of the Puerto Rican community of South Lorain...38

4. Issues in data collection models. which model is appropriate for a minority speech community?……………………………………………………………...50

4.1 Introduction……………………………………………………………...50 4.2 Network analysis………………………………………………………...52 4.2.1 Network case studies…………………………………………………….54 4.2.1.1 Rural migrants to the urban community of Brazlândia….………….…...55 4.2.1.2 Rural communities in Transdanubia, Hungary and Oberwart, Austria….57 4.2.1.3 A Vernacular Cultural Index………………………………………….....58 4.3 Class-based approach……………………………………………………59 4.4 Methodological similarities and differences between a class-based approach to linguistic variation and social network analysis…….……...60 4.5 Assumptions about the social conditioning of linguistic variation associated with a class-based approach and a scial network analysis.…..63 4.6 Summary: a unified account of sociolinguistic variation in the Lorain Puerto Rican Spanish speech community……………………………...... 66

5. Methods in data collection. data collection with social variation………..…...... 69

5.1 Introduction………………………………………………….……...... 69 5.2 Data collection………………………………………………………...... 69 5.3 Three generations in the community………………………….……...... 70 5.4 The youngest generation……………………………………….…...... 72 5.5 Participants……………………………………………………...... 72 5.5.1 Lorain Puerto Rican speakers and data collection……………….…...... 74 5.5.2 Puerto Rican Islanders and data collection………………………...... 74 5.5.3 Lorain Mexican Americans speakers and data collection………….…....76 5.5.4 American English speakers and data collection…………………….…...76 5.6 Interviews………………………………………………………....…...... 77 5.7 Interaction with the participants………………………………...... ….....79 5.8 Data transcription………………………………………………..…...... 80 5.9 Data analysis………………………………………………………….....81

xiv 6. Analyses of phonological variables………………………………………….....84

6.1 Introduction…………………………………………………..…………84 6.2 The vowels…………………………………………………..………….84 6.2.1 Introduction to variable (e)………………………………...…………...85 6.2.1.1 Results of the VARBRUL analyses regarding variable (e) in IPRS, LPRS and MAS………………………………………….……….………....…88 6.2.2 Introduction to variable (o)……………………….…….…………...….96 6.2.2.1 Results of the VARBRUL analyses regarding variable (o) in IPRS, LPRS and MAS………………………………………….………………...…..99 6.3 The consonants………………………………………………….....…..106 6.3.1 Introduction to variable (r)………………….. ……………………...... 106 6.3.1.1 Results of the VARBRUL analyses regarding variable (r) in IPRS, LPRS and MAS…………………………………………………… ………...111 6.3.2 Introduction to variable (b)…………………………………………....131 6.3.2.1 Results of the VARBRUL analyses regarding variable (b) in IPRS, LPRS and MAS……………………………………………………….……...133 6.3.3 Introduction to variable (dΩ)…………………………………….….....149 6.3.3.1Results of the VARBRUL analyses regarding variable (dΩ) in IPRS, LPRS and MAS…………………………………………………….....154

7. Outcomes of language and dialect contact: Explanations………………...... 172

Appendices:

Appendix A Word list in Spanish…………………………………………………….181 Appendix B Reading passage in Spanish………………………………...………...... 186 Appendix C Questions for the interview in Spanish……………………………...... 189 Appendix D Questionnaire in Spanish………………………………………………..193 Appendix E Word list in English………………………………………………….….202 Appendix F Reading passage in English……………………………………….…….209 Appendix G Questions for the interview in English………………………………….212 Appendix H Questionnaire in English………………………...……………………...216

Bibliography………………………………………………………………………….224

xv LIST OF TABLES

Table Page

3.1 Race and Hispanic or Latino…………………………………………………..…29

3.2 Largest employers in the city of Lorain…………………………………...……..31

6.1 VARBRUL percentages for (e) in Island Puerto Rican Spanish………..……….90

6.2 VARBRUL weights, application values and input and weights for [e] in Island Puerto Rican Spanish…………………………………………………………….91

6.3 VARBRUL percentages for (e) in Lorain Puerto Rican Spanish…………..……92

6.4 VARBRUL weights, application values and input and weights for [e] in Lorain Puerto Rican Spanish…………………………………………………………….93

6.5 VARBRUL percentages for (e) in Mexican American Spanish in Lorain………94

6.6 VARBRUL weights, application values and input and weights for [e] in Mexican American Spanish in Lorain………………………………………………..……95

6.7 VARBRUL percentages for (o) in Island Puerto Rican Spanish……….………100

6.8 VARBRUL weights, application values and input and weights for [o] in Island Puerto Rican Spanish ………………………………………………………..…101

6.9 VARBRUL percentages for (o) in Lorain Puerto Rican Spanish……….….…..102

6.10 VARBRUL weights, application values and input and weights for [o] in Lorain Puerto Rican Spanish ……………………………………………………….….103

6.11 VARBRUL percentages for (o) in Mexican American Spanish in Lorain……..104

6.12 VARBRUL weights, application values and input and weights for [o] in Mexican American Spanish in Lorain …………………………………………………...105

xvi

6.13 VARBRUL percentages for [r], [l], deletion and [] in IPRS..…………...…….113

6.14 VARBRUL weights, application values and input and weights for the variant [r] in Island Puerto Rican Spanish…………………………………………………114

6.15. VARBRUL percentages for [r], [l], deletion and [] in LPRS………………….116

6.16 VARBRUL weights, application values and input and weights for the variant [r] in Lorain Puerto Rican Spanish ………………………………………………..117

6.17 VARBRUL percentages for [r], [l], deletion and [] in MAS………………….118

6.18 VARBRUL weights, application values and input and weights for the variant [r] in Mexican American Spanish in Lorain……………………………………….119

6.19 Summary of percentages and comparison of input probabilities and weight values for [r] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican American Spanish……………………………………………….…....120

6.20 Summary of percentages and comparison of input probabilities and weight values of [l] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican American Spanish …………………………………………………....127

6.21 VARBRUL percentages for [r], [l], deletion and [] in the ten speakers of the first generation of LPRS with exposure to AE..………………………………...... 128

6.22 VARBRUL percentages for [r], [l], deletion and [] in the conversations of the seven speakers of the third generation of LPRS who were Spanish dominant…………………………………….……….…………………………130

6.23 VARBRUL percentages for (b) in IPRS…………………………………...…..135

6.24 VARBRUL weights, application values and input and weights for [b]/[B] Island Puerto Rican Spanish ……………………………………………………...…..136

6.25 VARBRUL percentages for (b/B) in LPRS ……………………………………137

6.26 VARBRUL weights, application values and input and weights for (b) in Lorain Puerto Rican Spanish…………………………………………………………..138

xvii

6.27 VARBRUL percentages for (b/B) in MAS……………………………………..140

6.28 VARBRUL weights, application values and input and weights for [b/B] in Mexican American Spanish in Lorain………………………………………….142

6.29 Summary of percentages and comparison of input probabilities and weight values of [b]/[B] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican American Spanish……………………………………………..…143

6.30 Summary of percentages and comparison of input probabilities and weight values of [v] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican American Spanish…………………………………………………….146

6.31 VARBRUL percentages for [b] and [v] in the ten speakers of the first generation of LPRS with exposure to AE………………………………………………….147

6.32 VARBRUL percentages for [b]/[B] and [v] in the conversations of the seven speakers of the third generation of LPRS who were Spanish dominant………………………………………………………………………..148

6.33 VARBRUL percentages for (dΩ) in IPRS.…………………………………..…156

6.34 VARBRUL weights, application values and input and weights for [dΩ] in Island Puerto Rican Spanish……………………………………………………….…..157

6.35 VARBRUL percentages for (dΩ) in LPRS……………………………………..159

6.36 VARBRUL weights, application values and input and weights for [dΩ] in Lorain Puerto Rican Spanish………………….………………………………………..160

6.37 VARBRUL percentages for (dΩ) in MAS ……………………………………..162

6.38 VARBRUL weights, application values and input and weights for [dΩ] in Mexican American Spanish in Lorain……………………………………….....163

6.39 Summary of percentages and comparison of input probabilities and weight values of (dΩ) for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican American Spanish…………………….………………………………164

xviii

6.40 Summary of percentages and comparison of input probabilities and weight values of (j) for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican American Spanish……………………………………….……………168

6.41 VARBRUL percentages for (dΩ) in the conversations of the seven speakers of the third generation of LPRS who were Spanish dominant………………………...169

xix

LIST OF FIGURES

Figure Page

3.1 Map of Lorain, Ohio……………………………………….…………………….30

3.2 Map of Puerto Rico………………………………………………………………33

CHAPTER 1

INTRODUCTION

The Puerto Rican migration has brought to the American melting pot many intriguing

issues. Puerto Ricans are US citizens by birth, and, thus, their migration from Puerto

Rico is comparable to any interstate movement within the mainland. On the other hand,

their culture and language are as foreign to general American culture as the cultures and

languages of many other ethnic groups that migrate from Europe, Africa, the Middle East

and Asia.

Researchers have studied Puerto Rican1 migrants in the United States, as many other migrant groups. However, their studies discuss issues that "Nuyoricans", or Puerto Rican

migrants who live in New York, face. Although the term "Nuyorican" has been extended

to refer to any Puerto Rican who lives in the United States2, most of the

1 The label "Puerto Rican" refers both to people born on the island of Puerto Rico and those of Puerto Rican ancestry born on the US mainland. This is unlike the labels "Mexican", which only refers to those people who come from Mexico; and "Mexican American", which is the label used to refer to people who live in the United States and are descendants of Mexicans. Since Puerto Rican Islanders are American citizens, the term "Puerto Rican American" seems inappropriate and misleading, just as the term "Ohio American", for instance, would be. However, a distinction will be made between Puerto Rican Islanders and Lorain Puerto Ricans.

2 Similarly, Nuyores has come to mean the United States.

1 literature written about Puerto Ricans in the United States concentrates on their situation in New York City, or other east coast cities (Casiano Montañez 1975, Alers-Montalvo

1982, Institute for Puerto Rican Policy 1992, Zentella 1997). Nowadays, Puerto Ricans are following the pattern of other Latino groups who are moving away from the states where they have traditionally been concentrated, and founding new communities all throughout the United States (Western Interstate Commission for Higher Education

2004). Despite these changes in US demographics, Latino communities in the Midwest, for example, have been understudied. Many questions, thus, remain without answers.

Although issues such as bilingual education and access to higher education, among others, are beyond the scope of this study, I examine here a linguistic issue that can ultimately contribute to the development of educational policy and even materials for the

Latino community. The linguistic situation in any community involving immigrants is complex, and the complexity is transported to the classroom at any educational level.

Thus, before any bilingual policy in an immigrant community can be implemented, the linguistic situation needs to be documented and studied first.

The main purpose of this study is to document and explain a selected number of variable elements in the pronunciation (i.e., the phonology) of the Puerto Rican Spanish (hereafter

PRS) variety spoken in Lorain, Ohio, a city 30 miles west of Cleveland. Previous studies on Mexican American Spanish (henceforth MAS) and PRS have focused on these varieties as spoken in the Southwest of the United States and New York City (Casiano

Montañez 1975, Peñalosa 1980, Zentella 1997); notable exceptions are Attinasi 1982,

Figueroa and Hislope 1998, Ringer Uber 1998 and Mendieta and Molina 2000 which have paid attention to varieties spoken in the Midwest of the United States. However,

MAS and its speakers have a different history from the Puerto Ricans and their language.

Thus, one should not assume that the findings of the above studies on MAS will shed light on the history and development of PRS in the United States.

Studies on the Mexican American variety of Spanish have focused particularly on phonology, aspects of code switching, code switching strategies, language attitudes, language use and language education in the United States. Studies about other varieties of Spanish in the United States have neglected the area of phonology, paying more attention to issues related to education and code switching. By describing several phonological features characteristic of the PRS spoken in the Midwestern city of Lorain,

Ohio, this study fills a gap in the literature on PRS phonology in the United States.

Moreover, I pay particular attention to the region of origin in Puerto Rico for the PRS speakers, describing the phonology of the PRS that was spoken towards the end of the

1940s; this is a point of relevance because it was this phonology that was transported to

South Lorain and which developed there, either evolving differently or maintaining the same patterns of development that IPRS follows. Furthermore, I examine changes in progress across three different generations of Puerto Rican speakers of Lorain, Ohio, looking in particular at the role of each generation.

Sociolinguistic research focuses in large part on matters of language variation and change. However, there is an evident gap in the study of language variation and change

3 in the Spanish varieties spoken in the United States. In particular, as noted above, there is an absence of research into the varieties spoken in the American Midwest where many

Spanish speakers have migrated in recent decades. As the US Census figures show, the

Hispanic population in the Midwest has increased from 3,754,389 in 1990 to 5,254,087 in

2000, representing an increase of 1,499,698 or 39.9%. In Ohio alone, Hispanics increased 55.4%, or 77,427; in 1990, the Hispanic population in Ohio was 139,696, while in the 2000 US Census this group consisted of 217,123 people. Thus, this project is also unique in that it studies a region in the United States experiencing a rapid growth in its

Hispanic population that has not been extensively studied in this regard. More specifically, within the Hispanic population in the American Midwest, I examine the dynamics of interaction of two Spanish dialects spoken by two minority groups in the context of English as a majority language.

The PRS situation in Lorain, Ohio is complex and any account should not be overly simplistic. PRS speakers are in continuous interaction with speakers of another Spanish dialect, MAS. Both groups live in a community where American English (hereafter AE) is the dominant language. The degrees of bilingualism within the community of Puerto

Ricans are even more complicated. PRS speakers who first arrived to Lorain, bringing with them a rural variety of PRS, do not have an extensive command of AE. Their immediate descendants, however, because of their exposure to AE in the school system and in the community at large, are native speakers of this language, while –at the same time-- speaking the rural variety of PRS they have learned from their parents and other relatives. The linguistic situation for the following generation is drastically changed as

they are AE dominant with --at best-- a basic, in some cases even non-existent, command

of Spanish at all linguistic levels. So, those who are English dominant could not really be

leading a change as they are switching to AE. Odlin (1997) describes the targeted

language of the third generation as a second language variety spoken by the bilingual

second generation, i.e., their parents. These third generation speakers who have learned

and maintained the language are the ones introducing variation in LPRS.

The linguistic situation in other Spanish-speaking communities in the United States has been documented, but the relevant studies present mixed conclusions. For instance,

Zentella (1997) argues that there is no English influence in the Spanish spoken by second generation Puerto Ricans in New York City:

The influence of English on the Spanish pronunciation of the first generation was minimal, corroborating Casiano’s (1975) finding that the Spanish of PR immigrants who had lived in New York for 20 years or more revealed no evidence of English interference. Even second generation English-dominant bilinguals showed no signs of the kind of English phonological transfer that might be labelled Anglicized Spanish as a counterpart to the first generation’s Hispanized English. The children sounded like native Puerto Ricans when they spoke Spanish.

Thus, the situation of the Lorain Puerto Rican community differs from that of El Barrio in

New York City (Zentella 1997). By contrast, Peñalosa (1980) argues that some phonological features in MAS are indeed the result of English influence. Other features

in MAS and LPRS are found in colloquial dialects in Mexico and other Spanish-speaking

regions. In LPRS, there are several phonologically variable elements, among them the

realizations of the mid vowels (e) and (o), of (r), (b) and of (dΩ).

As far as the Spanish situation in Lorain is concerned, it may be that AE has influenced

the Spanish spoken in both the Puerto Rican and the Mexican communities. Thus,

different competing hypotheses need to be considered when one attempts to account for

the different phonological variables in PRS: (a) they are continuations (or “natural”

extensions) of features of Island Puerto Rican Spanish (hereafter IPRS), (b) they

represent influence from MAS, (c) they show effects of interference from AE on the

phonology of Lorain Puerto Rican Spanish (henceforth LPRS), or (d) they are

independent innovations unrelated to the developments in IPRS, MAS or AE. Moreover,

the factors represented in these hypotheses could be acting independently, or all, or some

subset, together to produce the variable phonology in LPRS. Furthermore, each variable

could have a different explanation --there need not be a single explanation that holds for all the variables.

The data for this project were obtained from voice recordings, video recordings and notes

taken during observations of social activities and interviews. The voice recordings were

collected from a total of one hundred and four speakers from four different ethnic groups.

The voice recordings consisted of an informal conversation with the researcher and two

readings that collected a more formal, careful sample. Furthermore, a long-term

observation of the community was conducted accompanied by note-taking and video-

recording. The video recordings were collected and the notes were taken during social

events such as festivals, birthday parties, quinceañeras (a rite of passage similar to the

American “sweet sixteen”) and weddings. The tokens analyzed for this study, however,

were collected in guided and controlled tasks. No tokens were drawn from the video recordings or from the notes taken; rather, these were used to round out the overall

picture of language use in the community.

Inasmuch as the focus of the project is the Puerto Rican community in Lorain, the largest

group recorded consisted of these speakers. However, these were not the only speech

samples collected. Samples of the phonology of the MAS spoken in Lorain was collected

and studied to test whether the phonological variables evident in LPRS were influenced

by this Spanish dialect. Recordings of the IPRS spoken in the rural areas of Puerto Rico

were collected as an example of the Spanish transported to Lorain in the late 1940s.

Another sample consisted of native speakers of AE with whom the Puerto Ricans and

Mexicans interact on a regular basis. Their English phonology constitutes a baseline

against which to compare AE influence on LPRS, essentially serving as a control group.

In my initial examination of LPRS, certain aspects of pronunciation, but not syntax or

morphology stood out as salient as an IPRS speaker and potentially interesting to study.

The data considered for this study include developments with mid-vowels, and some

consonants (realizations of /r/, of /b/ and of /dΩ/), and ways in which their treatment differs from the standard phonology of IPRS. Other variable elements, however, have not been considered here, for instance, variation between [tS] and [S], as well as variation

of [s] with aspiration and deletion. These were not considered because, in the case of [tS] and [S], it was not clear how AE was involved at all since AE has both [tS] and [S]. In the

case of variation of [s] with aspiration and deletion, it is known that this variation already

occurs in IPRS as well as in LPRS so no new developments were to be found. Moreover,

studies have already explored (s) variation, and so I felt the need to explore other areas of

the PRS phonology, particularly in the United States.

It is important to note that the source for IPRS is multiple as the Spaniards who settled in

Puerto Rico came from a diverse geographical background within Spain. For instance,

Spaniards from the Canary Islands settled in the rural coastal areas as well as the

mountainous center of Puerto Rico, while Catalonians settled predominantly in the

coastal urban centers and Spaniards from the Basque country, Navarra, Galicia, León,

and Aragón, could be found in the rural as well as the urban areas (Álvarez-Nazario

1992). Because of the incentives provided by “La Real Cédula de Gracia”3, many settlers

from Corsica moved to Puerto Rico and concentrated in the mountainous regions –but

particularly in the southwest— where coffee was cultivated (e.g. Yauco and nearby

3 “La Real Cédula de Gracias” was granted by the King of Spain in the early 1800s to increase the population in the island, awarding land grants to immigrants from Spain, other European countries and colonies in South America.

areas). Thus, even though not the source of the regional IPRS dialects, standard IPRS is more conservative with regard to the features considered for this study and, thus provides

a convenient basis against which to judge “correctness” and the degree of innovation for

dialectal IPRS and LPRS.

Regarding the features discussed above, previous pilot studies (Ramos-Pellicia 2000,

2002) investigating these sounds showed the raising of unstressed /e/ among speakers of

the second and third generations, the pronunciation of /r/ as a retroflex, and the raising of

unstressed /o/, all features shared with MAS, and thus providing evidence for MAS and

LPRS convergence. In the present study, however, very few cases of vowel raising were found among the participants. The results suggest that the variation is almost lost (see

Chapter 6). Contrary to the few cases of vowel raising, cases of retroflex [] and lateralization were found in the data. Features such as lateralization of /r/ and the raising of unstressed /e/ still maintain the same negative prestige for the first-generation Puerto

Rican speaker that is prevalent in IPRS. This negative evaluation of PRS phonological features is considered in the analysis. However, the speakers of the second and third generations from this community are not exposed to the same IPRS variety that the members of the first generation were exposed to in their formative years. The former are exposed to MAS and to the LPRS spoken by their relatives in Lorain, while the latter were exposed to rural IPRS.

As tempting as it may be to argue that what is emerging in Lorain is a new variety of

Spanish in which features transported from the rural areas of Puerto Rico, features from

MAS, and features from IPRS are combined to make what is referred to here as LPRS, I

consider whether it is sufficient to concentrate on only a limited number of phonological

features in order to define a new language variety. Accordingly, I discuss in this study

the criteria that should be considered to establish a given variety as different from

another, specifically, whether they can be limited to phonology or whether the analysis should be expanded to include syntax and morphology, and if so, how many features

should be considered when determining if a spoken system is in fact a new dialect.

Furthermore, because the group under study is a minority group, unlike many other

groups that have been observed in sociolinguistic studies, complications arise when it comes to data collection methods. In particular, a crucial question is whether a class-

based approach or a network analysis is the appropriate method of data collection for a

minority speech community. Although I do not pretend to offer a final answer to this

troubling question, I present several aspects of both methods, as a synthesis of both

proved most appropriate for data collection for this study.

This dissertation consists of seven chapters. The chapters to follow this introduction

cover several topics related to minority speech communities. Chapter 2 discusses the

different theoretical issues of the linguistic and social outcomes of dialect and language

contact. A historical, social and linguistic description of the Lorain Puerto Ricans is

presented in Chapter 3. Since minority language groups present a challenge when

developing a methodology for data collection, Chapter 4 discusses which model of data

collection is appropriate for a minority speech community and is followed by a thorough

10 discussion in Chapter 5 of the methods used in this study. Chapter 6 presents the description of the different phonological variables selected for this study as they are found in several Spanish dialects and discusses the problems in determining the emergence of a new dialect, correlating the phonological variables with the social aspects of language, and analyzing the social and linguistic data. Chapter 7 discusses the criteria considered when determining if a spoken system is in fact a new dialect. This chapter also lays out the overall conclusions of the study.

CHAPTER 2

THEORETICAL ISSUES: LINGUISTIC AND SOCIAL OUTCOMES OF

LANGUAGE CONTACT AND DIALECT CONTACT

2.1 Introduction

The framework of dialects in contact (Trudgill 1986) describes linguistic situations involving social and spatial mobility in ways that the Labovian framework (1972) has not been able to explain. Several outcomes, among them dialect convergence, dialect levelling and konéization, are found in dialect contact situations. However, because the outcomes depend so much on a myriad of social factors that may promote the preservation of a language or a dialect, or favor a group of speakers shifting to a different language or dialect or even foster the mixing or convergence of different linguistic systems it is impossible to list all the possible scenarios resulting from language/dialect contact. Among the sociocultural factors, Winford (2003) mentions as relevant not only the type of community settings, demographics of the different groups in contact and their numerical ratios, patterns of social interaction, and the speakers’ ideologies and attitudes towards their languages/dialects, but also factors such as degree of bilingualism, history of the contact, the frequency, length and intensity of the contact, cultural pressures (such as social and economic advancement, and better educational or employment

opportunities) to adopt certain foreign features, social values (the relative prestige of the

linguistic systems, degree of loyalty to each language/dialect), the structure of the social

networks in the community of speakers, degree of tolerance to foreign interference,

language mixture, social solidarity, social identity, the linguistic structure and linguistic ideology of the community, patterns of language usage and language accommodation between groups as well as the power relations between the linguistic groups in contact,

the direction of the change, degree of stability or shift of the languages/dialects involved.

The sociocultural factors are so relevant because they can supersede any internal

resistance to linguistic change.

Language contact, on the other hand, refers to the interaction that involves linguistic

systems not sharing many similarities in the domains of grammar and phonology. By

contrast, dialect contact deals with linguistic varieties that share sufficient similarities between the systems to foster intelligibility. The similarities could be shared in different domains, e.g., lexicon, phonology, but are not necessarily limited to these. In the Lorain community of Puerto Ricans, the possible outcomes of dialect contact between Puerto

Rican Spanish and Mexican American Spanish have to be studied, while not forgetting the outcomes that contact with AE may have on LPRS.

2.2 Overview of the pressures, processes of formation and outcomes emerging out of

language/dialect contact

The contact linguistic situation in the speech communities observed for sociolinguistic studies has been explained by referring to the superstrate pressure of a dominant language

(Thomason & Kaufman, 1991, Holmes 2000), by accommodation, leveling and konéization processes (Silva-Corvalán 1996, Thomason 2001) that result in convergence

(Kerswill 1994, 1996).

Before one can come to an understanding the sociolinguistic situation in the speech community of Puerto Ricans in Lorain, the different pressures, processes and end results used to explain dialect and/or language contact in different linguistic communities need to be examined.

The Lorain Spanish speakers are arguably experiencing pressures and undergoing processes that are typical when dialects interact in a community (e.g., the Puerto Rican and the Mexican American speech communities in face-to-face interactions). However, the AE influence as the superstrate should be added to their dialect interaction as these interactions take place in the context of a wider mainstream American English-speaking community.

2.3 Superstrate influence

Superstrate influence has often been exemplified by contact situations such as Prince

Edward Island French, Pennsylvania German (Anderson and Martin 1976, Keiser 2001),

Schwyzertüütsch, NYC Spanish (Zentella 1997), LA Spanish (Silva-Corvalan 1994), in which lexical and structural superstrate influence of English is evident. Other situations studied include French and Breton contact in Brittany (Dressler and Wodak-Leodolter

1977) French contact in Montréal, Canada and Alsace, Strasbourg, as well as Creole genesis, where one language is more dominant, sometimes having more prestige than the other language with which it has contact.

2.4 Processes

Convergence is the outcome of several processes taking place when two or more dialects interact. Initially, speakers of different dialects will accommodate to their peers’ dialect to foster communication. During interactions taking place over extended periods of time, speakers may replace certain features of their dialect with features from another dialect in the community, or speakers may eliminate differences by reducing the variation in salient features of their dialects to come up with a more homogeneous variety for the different dialectal speakers. However, depending on the scenario where contact takes place and the patterns of social interaction, there may be different outcomes (Winford 2003).

2.4.1 Dialect accommodation

The process of accommodating to a dialect (Giles and Smith 1979; Giles, Taylor and

Bourhis 1973) must come before another dialect is acquired or before dialects converge.

Speakers will first accommodate to features that impede intelligibility between speakers of similar dialects. Short-term accommodation will result in the acquisition of features that were originally accommodated to when speakers were trying to communicate at first.

Accommodation, however, may be incomplete because the speakers may be alternating between the targeted variant and their own as a result of their use of the variant in some words and not in others or because of approximations to the variant being accommodated.

For example, in a case where one dialect has /s/ and another dialect has /S/ in comparable words, speakers may take on the variant phoneme that is new for them while still retaining another phoneme from their own dialect. The phoneme is variably acquired since it is in alternation with the phoneme of the speaker’s dialect.

Speakers modify certain pronunciations through analogy and borrowing to accommodate their dialect to that of the target dialect. Some words will be affected first, then, others.

In the case of incomplete accommodation, the interdialect used (i.e., fudged dialect, interlanguage, approximation), is characterized by the creation of intermediate forms.

There are two types of intermediate forms, (a) those that occur in neither dialect, but that are produced as intermediate variables (forms) and (b) those that exist in the recipient dialect and are used because they are an intermediate realization of the one imitated and that to be replaced. The latter form is not the result of the contact between the two dialects. An example presented in Trudgill (1986) is Oslo Norwegian of the diphthong

/øy/, a phonetically intermediate compromise form between the upper-class stereotyped monopththong /ø:/ and the peasant stereotyped diphthong /æu/. Larsen (1907, quoted by

Trudgill 1986) found the /øy/ diphthong –without any historical justification— in words such as brøyt (‘broke’). The /øy/ diphthong was already found in e.g. /gøy/ (‘fun’).

Anderson and Martin (1976) studied two Old Order Mennonite communities, one in

Pennsylvania, another in Ontario. The Old Order Mennonite Pennsylvania community is in more intense contact with AE than the Ontario community. Andersen and Martin

(1976) found that the loss of the dative and the borrowing of non-trilled ‘r’ in

Pennsylvania Dutch are greater in the Pennsylvania community due to the increased contact with AE.

2.4.2 Convergence

For dialects to converge, certain new features must first be acquired by a speech

community. Trudgill (1986) predicts that those features that will converge are the ones

that are salient in face-to-face interaction. The saliency of a feature depends on whether

it has social value or not as well as on the feature’s frequency and transparency. In

intercommunity long-term accommodation, salience is not defined socially, but

linguistically. The salient feature has to have greater phonetic/grammatical distance, and

may alter surface phonetic contrasts producing communicative difficulties. In a contact situation, not all features that are different are selected for alteration, only those that are

salient (i.e., those that have greater linguistic distance, alter contrasts and make communication difficult) are modified.

While dialect accommodation is taking place, certain features (phonological, lexical, syntactical) may be converging. Convergence, according to Giles and his associates

(Giles, Taylor and Boorhis 1973) is the behavior that speakers exhibit when –in a broad conversational setting-- they attempt to adapt their speech patterns to those speech patterns around them. Convergence can occur not only in content, but also in form.

Wolfram (1973) illustrated long-term convergence or ‘long-term accommodation’ in a case where Puerto Rican immigrants in New York assimilated to either the white or black norms of English depending on the degree of contact with the larger community in question.

2.4.3 Results of dialect accommodation: levelling and konéization

According to Trudgill (1986), Kerswill (1994) and Kerswill and Williams (1994),

levelling is the “reduction of the number of variants of a particular phonological,

morphological or lexical unit”. As Kerswill (1994) explains:

Levelling refers to the disappearance of all but a couple of the variant forms of a particular linguistic unit (e.g. a phoneme or an inflectional morpheme) found in the original dialect mixture.

The reduction or attrition that occurs in the mixture is found among those variants that are geographically and demographically restricted or “marked” variants in a defined area

(Torgersen & Kerswill 2004). Levelling is closely related to the speech accommodation mechanism. People from different regions move to the same place, individual acts of accommodation occur and, over time, the individual accommodations lead to quasi- permanent changes in the speakers’ linguistic performance. As a result, differences inside the “melting pot” are leveled out. Then, this is a new variety with no strongly localized forms. The leveled varieties replace localized varieties that are distinctive at the larger, regional level. These new forms are intermediate between the older more marked forms.

Levelling is a factor in konéization, a process in which mixing among speakers of different dialects leads to the rapid formation of a new mixed and generally simplified variety, found in mobile populations where individuals do not have the opportunities to establish close and dense networks (Tuten 2003, Martinez 2000, Milroy & Milroy 1985).

These speakers have a lot of contact with speakers of other dialects. The situation is then

that these individuals come into contact in face-to-face interactions and they try to

accommodate their speech to that of the speakers of the other dialect, usually trying to

avoid marked features that could deter communication. With this accommodation

occurring repeatedly by numerous speakers a permanent language change is produced.

The marked features (stigmatized or otherwise strongly localized regional features) will

disappear and the widely accepted standard (i.e., sociolinguistically neutral) features will

prevail. Those forms that are the majority are preferred as well as the simple features

which are chosen over the more complex ones.

The members of the first generation of the Puerto Ricans in Lorain have been found to

accommodate their speech to MAS in minor ways (e.g., phonology, some morphology

borrowing, lexical borrowing, etc.) since they have already passed the critical language

acquisition stage (Lenneberg 1967, Kerswill 1996b). However, the members of the second generation (i.e., their children) are the protagonists of the levelling process by

reducing the amount of variation in their speech community.

The study of Milton Keynes, Reading and Hull in Great Britain illustrates the levelling

process mediated by different underlying processes (Williams and Kerswill 1999).

Another study that illustrates levelling is found in the formation of the Høyager dialect

(Kerswill 1994). A small area in western Norway is characterized by diphthongal

realization of /y:/, /i:/, /Ë:/, /u:/, as also in some areas in the west. Also, /å/ has

dipththongal realizations. It is in this typical diphthongal area that the Høyager dialect is

found. The Høyager dialects, however, have leveled out the diphthongal forms favoring

20 monophthongs. Torgersen & Kerswill (2004) argue that dialect contact and, more specifically, dialect levelling play a relevant role in the process of vowel shift in the

English towns of Reading and Ashford. In the case of Ashford, the observed vowel shift follows Labov’s Principles II (“In chain shifts, short vowels fall. In chain shifts, the nuclei of upgliding diphthongs fall”.) and III (“In chain shifts, back vowels move to the front”) of Vowel Shifting (Labov 1994), but the vowel shift in Reading does not follow a vowel chain shift, the vowels in “foot” and “strut” seem to be moving in opposite directions, while the other short vowels do not show any changes. The vowel shift in

Reading is rather a change, with a different starting point, in two vowels (i.e., the vowels in “foot” and “strut”), with a similar target for change as in Ashford, the backing of the vowel in the London area, at the same time that the front vowels in Reading do not change because they already had the positions targeted in Ashford.

Martinez (2000) argues that Tijuana Spanish has undergone koinézation as a result of the levelling process that arose through intense migration to the city of Tijuana. Moreover,

Tuten (2003) argues that Medieval (and Modern) Spanish came about as a result of konéization through three different stages of dialect mixing among speakers of different dialects.

Another factor in konéization is simplification. Simplification, sometimes representing the accommodation to the other dialect, involves the elimination of irregularities, e.g., the elimination of morphological irregularity in the vowel changes of irregular verbs. An example of simplification can be seen in the present tense of ‘ta’ in the Stril dialects of

Norway. There, ‘ta’ (take) is regularized to /tA:r/ among the Strils in Bergen (Kerswill

1994). Other cases of simplification result in the lack of acquisition of a complex rule.

For instance, the Stril migrants do not acquire the rather complex rule of vowel lengthening and the syllabification of /l/ in singular and plural definite forms for the nouns. Another case of simplification is found in the morphology of Trinidad Hindi

(Trudgill 1986). Trinidad Hindi has lost gender, number, case agreement as well as several forms in the verbal paradigm. In the latter case, the eighteen forms that were original in the dialect have been reduced to three. Due to the simplification in the verbal paradigm, not only the number distinction has been lost, but also second and third honorific inflections as well as the feminine inflectional forms.

2.5 The result of dialect accommodation: dialects and languages converging

When two or more variants have become alike as a result of several accommodating processes, languages become more similar as the end result. There are two possible cases of accommodating processes in language contact situations. One case is when two languages have mutual interference, making both of them similar to one another. Thus, this is not a unilateral process. The resulting convergent structures had no single source.

The resulting structures are very similar; neither is completely the same in any of the languages. The convergence found in Turkish-German bilingual children serves as an

example. The result of these two languages converging is that neither had the resulting feature in their intonation system. Turkish and German monolinguals have each a single

pattern of intonation for the same function. Turkish and German bilingual children in

Germany have two different patterns of phrase-final intonation in German and in Turkish.

The two intonation patterns in the two languages are differentiated semantically. Neither

German nor Turkish can be established as the unique primary source for the whole

pattern, i.e., the phrase-final intonation patterns can be traced back to only German or

Turkish; neither of the two languages has the two different semantic functions

(Thomason 2001).

Furthermore, Keiser (2001) argues that the Deitsch varieties spoken in the Midwest and

Pennsylvania show features that result of two languages converging. Schreier (2003)

discusses how the Maori dialect of New Zealand English has reduced word final

consonant clusters as a result of the substratum influence of Maori native structures that do not admit syllable-coda consonant clusters. In this case, the Maori speakers not only shifted to New Zealand English, but their language’s structure modified the structures already existing in NZE. These effects, however, were not maintained and were lost as

speakers’ accommodation to English increased.

The Balkan Sprachbund is a well-known case of convergence that came about as a result

of a long period of multidirectional contact facilitated by years of Turkish rule. The

sources of all the features that are converging are difficult to identify and much is

controversial as to their origins4. However, it is clear that they are not the result of

massive borrowing from a single source language. All the languages involved in the

Sprachbund (Greek, Rumanian, Albanian, Bulgarian, Macedonian, the Southeastern

dialects of Serbian) are not in an asymmetrical dominance relation, and show a

multilateral multilingualism (as opposed to one-way bi- and multilingualism).

Thomason and Kaufman (1991) propose that, in the case of a multilateral Sprachbund, it

is possible that convergence is developed similarly to code switching among bilinguals,

and multilinguals as proposed by Pfaff (1979). When speakers code switch, the

structures favored are those equivalent in both languages. For instance, periphrastic

verbal constructions are grammatically accepted in both languages (Spanish and English),

not raising any syntactic conflict. On the other hand, adjective switching must be

restricted to limiting adjectives before the noun. Postnominal adjectival constructions are

disfavored in Spanish/English code switching because there is a mismatch; English does

not allow postposed adjectives while Spanish does. Thus, adjective switches are limited

within the NP; they can be used instead as predicative adjectives (Pfaff 1979).

2.6. The Role of Social Networks

Recent changes in urbanization and the increase in ways of communication have had

consequences in the linguistic contacts between individuals. Researchers argue that,

because it is easier now to be in touch with different people everywhere (due to the

advances in communications and transportation), contact among members of different

4 For brief overviews of the facts concerning the Balkan Sprachbund, see Joseph (1992), Hock and Joseph (1996: 400-405), Thomason 2001; for a more detailed presentation, see Sandfeld (1930), Schaller (1975), among other works.

speech communities has relaxed the ties in what were once close-knit communities. The

laxity in such close-knit communities has brought more complexity to language contact

situations. In a close-knit community, the highly dense and localized networks

maintained the local norms and prevented new forms from entering the repertoire of the

speech community. With the changes in the structure of the social networks in a close-

knit community, uniplex networks are formed. These uniplex networks are argued to be

the facilitators of change (Milroy & Milroy 1985), and, consequently, the cause for

dialect levelling and konéization, possibly because newer forms are easily accepted now,

older forms are not preserved and marked forms are disfavored.

2.7 The Lorain Puerto Rican community

In Lorain, there are two groups of Hispanic speakers, i.e., Mexican Americans and the

Puerto Ricans. Each group arrived to Lorain with a different dialect. The details and

historical information about the communities are given in Chapter 3. Because the two

groups have been cohabiting side-by-side for a long period of time, their dialects are in

contact. Both Latino communities exist in the context of a wider AE speaking

community whose language is in contact with the two Spanish dialects.

Therefore, given the social situation of the Lorain community, several scenarios of language contact have been proposed to explain the differences between LPRS and rural

IPRS originally brought to Lorain by the first generation. One possibility is that AE is

shaping LPRS and having superstrate influence. Another possibility is that the dialects of

the two Latino speech communities have converged, and the two have become one

speech community with the same shared grammar, shared norms and the uniformity of

shared patterns of variation (Labov 1972).

2.10 Summary

To explain the linguistic situation in the Puerto Rican community of Lorain, I need to

consider superstrate influence and levelling, as these are taking place in the LPRS/MAS

dialect and Spanish/English language contact situation.

Convergence comes as a result of a long process of speakers accommodating their

dialects to those patterns of their interlocutors. As a result, the languages involved

become more similar. The end result of convergence is the development of a linguistic

system whose sources are not easily identifiable.

The data gathered for this study will show whether LPRS features have in fact converged with MAS features. Thus far, previous studies have shown that there is variation in the data. The Spanish of the Puerto Rican and Mexican American communities in Lorain is found in a society where AE is the dominant and more prestigious language. Thus, I explore whether some aspects of the phonology in LPRS (e.g., (b) ~ (v) variation) have been shaped by the influence of AE as the superstratum.

CHAPTER 3

THE COMMUNITY: SOCIAL AND LINGUISTIC DESCRIPTION OF THE LORAIN

PUERTO RICAN COMMUNITY

3.1 Ethnography of the Puerto Rican community of Lorain

Lorain is a city about 30 miles to the west of Cleveland, Ohio. According to the 2000 US

Census (Table 3.1), Lorain is inhabited by 68,652 people of whom 21.0% (14,438) are

Hispanic. Of this 21.0%, 15.3% (or 10,536) are Puerto Ricans; 3.5% (or 2,437) are

Mexicans; 0.1% or (64) is Cuban and 2.0% (or 1,401) belong to different Hispanic

groups. Thus, in Lorain, the Puerto Ricans outnumber by far all the other Hispanic

groups, including the Mexicans.

Race One Race Area Total Total White Black Ameri- Asian Native Some Two or His- Popu- or can of other more panic lation Afri- Indian Hawaii race races or can and and Lati- no of Ameri- Alaska other any can Native Pacific race Islan- der Lorain 68,652 65,911 47,848 10,943 304 227 24 6,565 2,741 14,438 city, Lorain county

Table 3.1 Race and Hispanic or Latino (Source: US Census 2000)

South Lorain (see Map 3.1) is the nucleus where the Puerto Rican community was

established in the late 1940s when the Ford Motor Company and the National

Tube/U.S./KOBE steel company5, the two major manufacturing companies in the area, hired a large number of Puerto Rican males. The majority of the Lorain Mexicans also lived in South Lorain (Decker 1952).

5 Hereafter, I will use the “steel company” to refer to the National/U.S./KOBE steel company, which is how the participants in this study refer to the steel plant.

Figure 3.1 Map of Lorain, Ohio The Puerto Ricans who came to Lorain concentrated in

South Lorain after their arrival to the region. (Source: http://mapquest.com)

Of these two companies, the steel company played a major role in hiring Puerto Ricans.

The steel company was in search of individuals who were U.S. citizens, were hard- working, and were willing to be underpaid, and who --at the same time-- could pass for

“Americans”6. Also, another concern in the company was the need to replace the many

Mexicans who were at risk of being deported because of a threat from the Immigration

and Naturalization Service. The INS threatened a raid and the expulsion of all the non-

American workers, a group into which many Mexicans fell. It was imperative, then, to

6 I use the term “American” to refer to race not citizenship, as in “White Caucasian”, or “Anglo American” English-speaking individuals.

look for American replacements who would work hard and accept low wages as the

Mexicans and Mexican Americans (Valdes 1991) had. Then, as now, the steel company dominated the Lorain labor scene. Table 3.2 presents official information provided by the city of Lorain that shows the ten largest employers in the city of Lorain today and there is good reason to suppose that the steel company has been in a leading position for decades.

======Employer Nature of Activity or Business Employees ======USS/KOBE Steel Company Steel Manufacturing 2,800 Community Health Partners Health Care 2,178 Ford Motor Company (Lorain) Automobile Manufacturing 2,000 RELTEC-Lorain Products Division Telecommunications 1,250 Lorain City Schools Education 1,200 May Department Stores, Inc. Computer Programming/Credit Center 1,600 City of Lorain Government Government 574 P .C. Campana, Inc. Steel mold manufacturing and repair 550 Lorain County Community Education/Community Services 290 Action Agency W.G. Nord Community Mental Health Care Health Care 250 ======

Table 3.2 Largest employers in the city of Lorain (Source: http://www

.Ci.Lorain.Oh.Us/nfLorain/largeemploy.htm)

The steel company went to Puerto Rico looking for individuals with the aforementioned characteristics. The political and economic situation in the Caribbean island could not

have been more favorable for the steel company needs. First of all, the government was

trying to solve the large socio-economic problem of unemployment among thousands of

Puerto Ricans. The government encouraged the establishment of US companies on the

island with the economic project of Operation Bootstrap (Dietz 1986; Silvestrini and

Luque de Sánchez 1988). However, the economic problem was not solved. The

unemployment rates were high, and became higher after Operation Bootstrap. Thus, the

Puerto Rican government started to encourage --directly or indirectly-- the migration of thousands of Puerto Ricans to the continental United States, and even to Hawaii

(Maldonado-Dennis 1980). Puerto Ricans migrated to New York and many metropolitan cities in the east coast. Of those who migrated to the east coast, very few were from the island's center (Alers-Montalvo 1982); rather, the Puerto Rican migrants were typically from the metropolitan cities of San Juan, Ponce and Mayagüez. The Lorain Puerto Rican profile, however, differs from that of the Puerto Ricans on the east coast. According to

O’Brien (1954), when the first group of Puerto Ricans arrived in Lorain, the groups consisted of rural men because the steel company that traveled to Puerto Rico to take advantage of Operation Bootstrap went specifically to the Puerto Rican mountainous center to hire workers. It did not go to the metropolitan area of San Juan, nor to big cities like Ponce or Mayagüez; see Map (3.2). In fact, San Juan accounts for only 3% of the first Puerto Ricans who migrated to Lorain, and Ponce and Mayagüez account for 2% of the total of Puerto Ricans brought to Lorain (O'Brien, 1954). The rural men, or jíbaros

had many reasons to immigrate to the United States. Not only was there a great scarcity of employment in the island, but also those who worked received very low wages. In general, rural areas had deplorable sanitary conditions and rustic residences that did not even offer suitable places to live.

Figure 3.2 Map of Puerto Rico The Puerto Ricans who moved to Lorain came from the mountainous center of the Island (in towns located in La Cordillera

Central) and towns outside the main metropolitan cities of San Juan, Ponce and

Mayagüez. (Source: http://www.lonelyplanet.com)

It is a common belief among the Puerto Ricans in Lorain that, due to the ethnic

composition of the residents of the mountains, these people were more likely to pass as

white Americans. Basically, the ancestors of these Puerto Ricans were not as racially

mixed as those who lived in the coast (Silvestrini and Luque de Sánchez 1988). Because

of their European ancestry (see Chapter 1), they were white with light-colored hair and

eyes. Furthermore, the previous employment experience of these individuals worked to

their advantage. They were hard workers who have been involved in the coffee and

tobacco industries.

Those who were hired by the steel company came to the United States in 1947. The first

group consisted of about fifty Puerto Rican males. Later on, more Puerto Ricans were

brought to Lorain. Between 1947 and 1950, approximately 500 Puerto Rican males

moved to Lorain. Many of these males were single, and those who had families later sent

for their relatives, increasing the number of Puerto Ricans in the area. In 1953, it is

reported that the Puerto Rican population in Lorain consisted of 3677 individuals; 2119

were males, while 1558 were females (O'Brien 1954). Later, more Puerto Ricans were

brought to the city of Lorain. Although it is calculated that around five hundred Puerto

Rican males moved to Lorain between the years 1947 and 1950, one is left to wonder

about how many Hispanics (i.e., Mexicans, Mexican Americans and Puerto Ricans) lived in Lorain in the late 1940s.

Several ethnic groups such as Hungarians, Polish, Slovenians, among others, used to live

in South Lorain. South Lorain, or el campito (“the small countryside”), is known to have

been very poor: inadequate housing conditions such as the lack of appropriate sewage

systems made it a very undesirable area to live. In fact, even before the Puerto Ricans

and Mexicans inhabited el campito, the area was looked down upon by other Lorain

citizens because it was very isolated. Decker (1952: 71) described the area as:

South Lorain was looked down on by citizens of most other parts of the city even before it was taken over by the Puerto Rican and Mexican laborers, the latter being the first to arrive. This was due to the fact that it is practically isolated from the rest of the city to the west by railroad tracks and small industry, to the east by a river and by the corporation line, to the south by the city limits, and to the north by the immense National Tube Company, several railroads and the Black River. The houses of South Lorain are, at best, humble, most of them completely in needs of repair.

As the members of the different ethnic groups improved their economic situation, these

people moved away from the area looking for better housing opportunities and a better

quality of life in general. Because of their low income, the newly arrived Puerto Ricans

had no option but to settle in el campito7. Thus, the first location of the Puerto Rican

community in Lorain was in the same area where Hungarians and other ethnic groups

used to live. El campito eventually started to be subdivided into different sections

depending on the towns or cities from where they originally came. Thus, sections of el

campito were named after Humacao, Utuado, Yauco, Lares, etc.

The Puerto Ricans are not the only Latinos in Lorain. As mentioned above, prior to the

arrival of Puerto Ricans, substantial numbers of Mexican and Mexican-American

workers were already established in Lorain. The Mexican-American community was

7 It is said that the landlords in the area agreed among themselves to limit the location of the new migrant group to the area between Pearl and Vine avenues. Any Puerto Ricans who looked for a place to live outside of this perimeter would not be successful.

established following World War I, earlier than the Puerto Rican arrival to Lorain. The

Mexican Americans came to Lorain from the southern United States (especially, Texas

and California, among other states) searching for better job opportunities, just like the

Puerto Ricans. Also, due to similar economic reasons, some members of this Mexican-

American community came directly from Mexico (Gregg 1954). However, unlike the

Puerto Ricans, the Lorain Mexican Americans could be second- or third- generation

individuals whose ancestors came from Mexico, or were already in the U.S. Southwest

before it was incorporated to the United States. Thus, many were already acquainted

with American culture and the English language.

The Puerto Ricans do not form an exclusive group, but interact with Mexicans and

Mexican Americans (and other non-Latino ethnic groups) as neighbors, coworkers,

classmates, and worshippers as well as through other types of interpersonal relationships

in their daily affairs. The interactions between the Puerto Rican and the Mexican-

American communities started in the late 1940s, soon after the Puerto Ricans arrived in

the region. Different circumstances promoted the contact between the two communities.

When the Puerto Ricans first arrived to Lorain, they initially rented rooms from Mexican

families, and even shared the living space with them, their families and other Puerto

Rican laborers (Decker 1952)8. Moreover, in the late 1940s, there was a great scarcity of

Puerto Rican stores, and it was very difficult for the Puerto Ricans to find the culinary

ingredients needed in the preparation of their preferred foods. Thus, they bought similar

Mexican products in Mexican stores. Another interaction between the Latino groups

8 Of the first generation Puerto Ricans interviewed for this study, some mentioned that they shared a very small space with a whole family or with many other single Puerto Rican males.

took place in social organizations, clubs, work and school environments. These types of interactions, along with the involvement of the Catholic Church and other religious denominations, served as the basis for establishing the earlier relations between the two

Latino groups. The Catholic Church, aware of the increasing needs of the Latino population in South Lorain, founded in September 1952 a new Spanish-speaking church:

The Sacred Heart Church/la Parroquia del Sagrado Corazón. This parish has a Spanish- speaking priest who conducts Spanish, English and bilingual services. However, this

Catholic Church does not identify with either Spanish-speaking group, but rather identifies itself as a Latino church with festivities and activities inclusive of both the

Mexican and the Puerto Rican cultures. An awareness of the growth of the Latino community was raised a couple of years before the arrival of the Puerto Ricans because the number of Mexicans and Mexican Americans in Lorain was increasing. It was not until the arrival in large numbers of Puerto Ricans, though, that the needs of the growing community were fully evident. The Sacred Heart still remains to this day the largest religious congregation, not only of Puerto Ricans, but also of Latinos in general.

However, other religious denominations (e.g. Pentecostals, Baptists) were also active back in the early 1950s, and are still active today within the Puerto Rican community and within the Latino population in general. Likewise, these religious denominations do not identify with a Latino ethnic group in particular, but target the broader Hispanic community in Lorain. These religious congregations also have Spanish-speaking pastors as head of their churches and conduct Spanish, English and bilingual services.

3.2 Linguistic description of the Puerto Rican community of South Lorain

The Lorain Puerto Rican community is bilingual, but with different levels of bilingualism

evident among the speakers. The lack of uniformity in their bilingualism is observed in

the differing linguistic behavior of the three Puerto Rican generations in Lorain. Each of

the three generations in the community has had a different exposure to English and

Spanish. Even within each generation, it is possible to find individuals with different

levels of bilingualism depending mostly on their age and social circumstances. Many of

the Lorain Puerto Ricans still maintain Spanish, but do not resist English. Others achieve

a level of bilingualism, while having a greater fluency in English.

The community, in general, can best be described as an open community, with a constant

migration flow between Lorain, Puerto Rico, New York and other US cities. Low

airfares as well as the increase in the number of flights within the United States and

between Puerto Rico and the United States encourage Puerto Ricans to travel. The flow

of Puerto Ricans allows for a constant input of PRS from Puerto Rico and other varieties

of PRS spoken in the mainland.

Moreover, the Puerto Rican community maintains constant interaction with the Mexicans

regardless of the speaker’s generation. Members of the older generation have been

exposed to speakers of Mexican Spanish and Mexican American Spanish in several

different environments: work, social clubs, stores, and church. The interactions, however, depended mostly on the gender of the individual. The males were more exposed to MAS since they had to interact with this ethnic group, as well as with

Mexicans, in their jobs in the steel mills. By contrast, women of the older generation did not have that much exposure, except when they had to go to the Mexican grocery store, or when they were interacting with Mexican (American) speakers in the church setting.

Today, members of the same generation interact with Mexican Americans, although in more limited environments, namely church, or in the neighborhood, because they are

retired. Likewise, members of the second and third generations interact with Mexicans

and Mexican Americans, not only in the same environments where the members of the

older generation used to do so, but also in school and in other social contexts. In general,

the community revolves around the church, an institution that has a dramatic and pervasive influence on members of all the generations. Thus, different generations of

Puerto Ricans interact constantly with MAS speakers in religious activities such as

Sunday school, funerals, festivals, and religious clubs (e.g. Hijas de la Virgen de

Guadalupe, Hijas de María, Cursillistas, Círculo de Oración). Furthermore, the interactions taking place at home and work are scenarios as important as the church and church-related activities for contact between Puerto Ricans and Mexicans.

The Puerto Ricans I interviewed in South Lorain are aware of some differences between their PRS and MAS. That the Spanish of the Mexicans and Mexican Americans differs from PRS and had some English influence was reported by members of the Puerto Rican community I interviewed9. Comments such as the ones presented in (1) thru (7)10:

9 A second generation LPRS speaker referred to the Mexican Americans in Lorain as mojados (“wetbacks”), using the derogatory term to refer to the illegal status of some Mexicans in the United States. The same speaker added, shortly after making this comment, that everybody gets along with the Mexicans in Lorain. However, comments such as his indicate subtle tensions in the community.

(1)

Ellos hablan, tú sabes, un Spanglish. (M, 1.1)11 They speak, you know, a Spanglish.

(2) Ellos patean el español. (M, 1.1) They corrupt Spanish.

(3) El español de ellos tiene un slang… con el accento… (Lorain_17_F_2_)12 Their Spanish is like a slang … with the accent…

(4) … los mejicanos mucho de ellos olvidaron su lenguaje por completo… (Lorain_5_F_2) ... the Mexicans many of them completely forgot their language.

(5) … I found myself watching Spanish stations, but the Mexicans have a different dialect… (Lorain_14_M_2)

(6) Different words mean different sayings… The way they say things… a Mexican probably speak with a long accent to the end of the word like “como estáááá”, you know. Spanish people are real real quick when they are talking… (Lorain_7_M_3)

10 The first four comments were made in Spanish, while the last three were in English. Note that the ones made in Spanish came from first and second generation speakers, while the ones in English came from second and third generation speakers.

11 This abbreviation stands for M(ale) speaker, l(st) generation, l(st speaker interviewed of that generation).

12 This abbreviation stands for Lorain seventeenth female speaker of the second generation.

(7) Some of the words are different, like, am, we say pahteles or something and that might mean something different for them. (Lorain_7_M_3)

were among the comments collected from our conversations. Similar comments were made about the Spanish spoken by the third generation.

(8) ... y nuestros hijos, los míos saben el español peor tal vez que yo... y trataba de hablar español a veces y automático vamos al inglés y ellos saben español... le entienden y lo hablan, pero no tanto. (Lorain_5_F_2) … and our children, mine know Spanish probably worse than me… and I tried to speak Spanish sometimes and automatically we switched to English and they know Spanish… they understand it and speak it, but not as much.

(9) ... mucha gente que es así ya… si es lo que le llaman Espangliashion... hablan un par de palabras en inglés pues muchas de las personas de edad en la Iglesia los viejolos hablan español. La juventud es ahora la que... los jóvenes hablan un chispito español les gusta hablar inglés que el español... (Lorain_6_M_2) ... many people are already like that... it is what is called Spangliashon… they speak a couple of words in English because many of the elderly in church, the elderly speak Spanish. The youth are the ones that now… the young speak a little bit of Spanish they like to speak English better than Spanish.

(10) No, hoy día los los jóvenes están perdiendo la la el el el idioma español parte por eh, por lo regular el el inglés a la larga domina, tú sabes, a veces nosotros mismos estamos hablando y se nos vienen palabras en inglés. (Lorain_11_M_2) No, today the young are losing the Spanish language, in part, because, usually, English is the one that is preferred, you know, sometimes even when we are speaking some words come in English.

(11) … hablan más ingles que espa..., pero hablan español también, pero

hablan mucho inglés. (Lorain_12_F_2) … they speak more English than Spa…, but they speak Spanish too, but they speak a lot of English.

Second and third generation speakers also acknowledge that their Spanish is different from the Standard or from that spoken by other more fluent speakers in the community.

Self-evaluations such as the one below in (12) and (13) were collected in my conversations with second and third generation speakers.

(12) … como nosotros no no sabemos el español muy bien a veces... (Lorain_5_F_2) ... like us, sometimes we do not know Spanish very well…”

(13) … just that my words are going to come a little pateao [messy] … okey ‘cause I learned my Spanish the hard way. (Lorain_5_F_3)

These observations demonstrate that PRS speakers are aware of differences between their

Spanish and MAS in Lorain. Moreover, it is clear that they all interacted with one another. Thus, some influence from MAS on LPRS is certainly a possible source of change and variation in LPRS. However, the influence is subtle since Puerto Rican speakers feel that the Mexican and Mexican Americans of their community are not good language models. I propose that Puerto Rican speakers may shift in the direction of MAS but not by completely accommodating to the speech of the Mexican and Mexican

Americans for the purposes of solidarity. Instead, they are avoiding certain features of

MAS because Mexican Americans are not regarded as representing a particularly prestigious culture since the stigma is that they are not US citizens and work illegally in

the country13. Their Spanish variety, as a consequence, is stigmatized among Puerto

Ricans because they are seen as socially and financially less attractive. However, LPRS speakers may choose other features from MAS that do not have a negative prestige or that are not stigmatized, and thus, are less salient, making communication easier, or they may choose features that are the result of AE influence in MAS and LPRS and thus do not carry the negative stigma that some LPRS speakers avoid.

The first generation turns to the use of certain phonological variables as a less stigmatized option in more formal environments, since they are acquainted with the acceptable, less stigmatized norm of the Island. The situation for the second and third generation is rather different since they do not have a clear idea of the norm, i.e., they are increasingly more exposed to English as the prestigious language, as well as the LPRS that their relatives use when speaking to them, MAS, and even other Spanish varieties if one considers television as a further possible source of influence on their speech.

As previously mentioned, over six decades, Puerto Ricans have been exposed to the

Spanish spoken by Mexican and Mexican Americans with whom they live and interact in the community. Because Puerto Ricans from the Island, from New York and from other

US cities move to Lorain, the Puerto Ricans of this community have been in uninterrupted contact with Island and New York Puerto Rican Spanish. Puerto Ricans in

Lorain also interact with AE speakers. Thus, they have been exposed, for several generations, to Mexican Spanish and to MAS, to AE, and to Island and New York Puerto

13 Even when the Mexican Americans are U.S. citizens and have been here for many generations, they are still seen as mojados (“wetbacks”) a Spanish pejorative expression used to refer to those Mexicans who crossed the U.S./Mexico border and get wet when crossing the Río Grande.

Rican Spanish. As a consequence LPRS has more phonological variation and more

variants for each variable than IPRS. The variability is due to a number of factors that come as a result of all these interactions.

The first generation participants have a very tight and limited network of friends. These networks of friends are usually limited to Spanish speakers who belong to the same generation. In only one case did a first generation speaker have a network that included several friends who were AE dominant, because their offspring preferred AE in their conversations due to their particular family situation, and because some of their friends or neighbors were AE native speakers. This will be explained in detail later on in Chapter 6.

The individuals who were interviewed in the Puerto Rican community of Lorain had different levels of education. On average, the first generation/first migrants had completed a third-grade level of primary school, while the second generation had achieved a higher level of education, with some having finished high school. Many of

the second generation Puerto Ricans have obtained a college degree or a Masters degree

and/or a doctorate. The same description holds for the third generation.

AE and Spanish cohabitate in Lorain in what is tantamount to a diglossic situation.

Puerto Ricans and Mexican Americans use AE in environments where the language is

commonly used in the mainstream culture, for instance, in schools, day care centers,

department stores, and banks. Spanish, however, is not completely excluded in those

environments, but AE is the language preferred. Spanish, however, is preferred in

44 settings such as home and church, when talking about topics such as child-rearing or experiences in Puerto Rico. Furthermore, in order to accommodate older speakers who do not know AE, even speakers who have mostly only a passive command of Spanish will switch to the latter, but only in limited environments and on a limited number of subjects.

Currently, Puerto Rican speakers in Lorain are bilingual, but with differing levels of bilingualism. They have limited active use and different abilities in Spanish and/or AE.

Younger third-generation speakers usually prefer AE, if they are not required to use

Spanish at home with their elderly relatives. These speakers tend to have a passive knowledge of Spanish, otherwise limited to listening comprehension14. Second- generation speakers have become bilingual and can communicate in both languages. The first generation, on the contrary, has either little command of AE, or just enough knowledge of AE to survive a daily errand. The second generation is bilingual because they need to communicate with the first generation speakers who have little knowledge of

AE and who are native Spanish speakers. Moreover, the second generation also grew up in Lorain at the time when the first generation was just being formed with their members having recently moved from Puerto Rico. If they wanted to communicate with their parents, grandparents and other older relatives, they had to use Spanish. Spanish was also the preferred language choice in their social networks present in the community.

The same second generation also had to use AE in environments outside the home, e.g., school, day care, and stores. However, the third generation nowadays does not need to

14 Almost half of the third generation (or 7) speakers I interviewed for this study were bilingual in Spanish and AE. The remaining thirteen third generation speakers preferred to use AE in our conversations, but were able to read in Spanish.

use Spanish with their parents because their parents know AE. The social networks of

the third generation consist of mostly bilingual or monolingual AE speakers. The only

situation when third generation speakers need to use Spanish is when they talk to their

grandparents, accommodating in this way to their grandparents’ native language.

Four speakers from the third generation serve as good examples of how social networks,

individual motivations to learn a language and positive attitudes towards the culture of

the language acquired (Brown 1993) play a role in language usage. Alexandra [all names

are fictitious] comes from a family who used AE on a daily basis to conduct all their

activities in and out of the house. As a school teacher working in her community, she

realized that she needs to know and to use Spanish in the classroom and has returned to

school to learn Spanish15. In my conversation with her father, AE was the preferred

language, but he used Spanish when prompted to do so when I switched to this language

in the conversation. By all accounts, Alexandra can be considered an AE dominant speaker with a passive command of Spanish because AE was the language preferred at

home during her formative years.

Miguel also exhibits a similar linguistic behavior to Alexandra. His preferred language is

AE, and as a college student counselor he realized that he needs Spanish in his work.

Although he is taking college classes in this language and makes efforts to use it with his

parents and with me, he has many ties outside and inside the community with speakers

who are AE dominant, but only weak associations with Spanish speakers. During the

time spent with his family, I observed their language usage. Miguel’s parents used

15 It is important to mention that the Spanish taught in schools is not IPRS, but Standard Spanish.

Spanish with him, and so did Miguel with them. But, when communication was not

possible, all of them switched to AE. Like Alexandra, Miguel can be considered an AE dominant speaker with a passive knowledge of Spanish, although maybe somewhat more natively Spanish than Alexandra.

Speakers like Miguel and Alexandra, who learned part or most of their Spanish in school

are problematic as far as classifying them as Spanish speakers. Although they are using the language, it is nonetheless relevant to understand that their Spanish is not the same as the one transmitted in the community of Puerto Ricans in Lorain, but rather a more neutral variety as it is "Standard" Spanish. The variety of Spanish taught in the schools and colleges where these individuals are enrolled will have the variation in the phonology, morphology and syntax of the Spanish dialect of whoever is teaching the language in the classrooms at those institutions, i.e., a Mexican teaching Spanish will use a Mexican accent, while an AE speaker teaching Spanish will have an American accent.

Then, these individuals present a problem when it comes time to classify them as a

Spanish speaker because of most or all of their Spanish learning took place in school not by interacting with members of their community.

Nicole and Phillip are the opposite side of the coin. Both of them were raised in single-

parent households. The preferred language at home was Spanish and this usage was

observed in the interactions with their families in which I participated. Furthermore, in

the interactions with their friends, they consciously chose to be friends with people who

were Spanish dominant so that they could “practice” Spanish. The comments in (14) and

(15) were made by Nicole (Lorain_5_F_3):

(14) En, yeah, todo el mundo tenían los grupos de ellos, pero yo jangueaba más con los puertorriqueños. (Lorain_5_F_3) And, yeah, everybody had groups [of them], but I hung out more with the Puerto Ricans.

(15) Oh, no no eran todos nosotros los mejicanos y los puertorriqueños juntos y los blancos estaban solos, pero yo me quedaba más con los puertorriqueños y los mejicanos porque a mí me gustaba hablar más español que inglés, así yo aprendía más. (Lorain_5_F_3) Oh, no no, it was all of us, the Mexicans and the Puerto Ricans together, and the whites were alone [by themselves], but I hung out more with the Puerto Ricans and the Mexicans because I liked to speak Spanish more than English, that way I would learn more.

Both speakers were able to follow and maintain a conversation in Spanish when I switched to this language in our interactions or when another speaker participated in the

interaction. Both Nicole and Phillip can be considered Spanish/English bilingual

speakers because they not only participated and used Spanish in Spanish-dominant social

networks, but they also used Spanish as their preferred language during their formative

years.

Another relevant factor to consider is the type of Spanish brought to Lorain by the original migrants. The PRS dialect they brought has features that are characteristic of the mountainous central area of Lares, Utuado, and other non-metropolitan towns. Thus, its phonology was already different from Standard PRS, with sound variation including, but not limited to, the raising of the middle vowels (e) and (o). This non-standard variety was the one used in conversations with their offspring, and with Mexican Americans in

Lorain. Thus, this study also explores the changes introduced by second and third generation speakers and/or by the interactions with Mexican Americans and AE speakers, as well as the question of whether or not the second generation has maintained the rural

PRS variety of their parents. The pages to follow this chapter will help in clarifying all of these questions.

CHAPTER 4

ISSUES IN DATA COLLECTION MODELS. WHICH MODEL IS THE

APPROPRIATE ONE FOR A MINORITY SPEECH COMMUNITY?

4.1 Introduction

A class-based approach and a social network approach are methodologies that have been used to explore linguistic variation. A class-based approach to linguistic variation

explains the conditioning of linguistic variation in terms of social factors such as class,

age, gender, ethnicity, socioeconomic status and style.

Network analysis can explain areas that have otherwise been impossible to explain by a

class-based approach to linguistic variation. In particular, social network analysis, as

illustrated by the Belfast studies of the Milroys (Milroy 1987, 1992a, 1992b, Milroy &

Milroy 1985, Milroy & Milroy 1992), helps in explaining: (a) patterns of resistance to change, (b) patterns of maintenance, and (c) how language changes are adopted by communities. However, the network analysis fails to take into account education levels, ethnicity, socioeconomic status and other information that a class-based approach otherwise considers (Labov 2001).

A class-based approach, however, misses important information obtained through a social

network analysis particularly in working class minority speech communities. In these

communities, a different array of possibilities for social network is possible since

bilingual individuals interact with speakers of the dominant language, with speakers of

their own minority group and with speakers of dialects of their own language. Because of all these language and dialect interactions, there are different sources and pressures that come into play. Thus, in order to capture all these possible scenarios of linguistic contact, it is not enough to only use one model for data collection. Therefore, in this chapter, I argue that a sensible model of data collection in a minority speech community includes, but is not limited to, a class-based approach. This method of data collection is complete only if it is combined with a study of the social networks of a community.

4.2 Network analysis

In the sociolinguistic literature, much has been said about methods of data collection.

Lesley Milroy (1987, 1992a, 1992b) and Milroy & Milroy (1992) have explained the

impact of studying the networks of the members of a community. In their research, the

Milroys based their analysis on previous sociological research (Mitchell 1969, Boissevain and Mitchell 1973, Boissevain 1974), paying particular attention to the personal relationships of the speakers with other individuals. Instead of emphasizing the inter- group interactions, they emphasize contact within the group. Thus, individuals should be understood not only in relation to the people of their own ethnic group, or to their neighbors, but also in the context of the society in which they live. Their lives are not isolated, but are intertwined in the complicated cultural and social relations of the city.

By using a number of indicators, the Milroys (Milroy & Milroy 1985, Milroy & Milroy

1992) facilitated the network analysis. The indicators are based on the concepts of

“multiplexity” and “density”. A network where everyone knows each other is “dense”.

When individuals from a community know each other in different capacities (e.g. friend, co-worker, member of a social club, etc.) within their community, this community is referred to as having a “multiplex” relationship.

Based on these studies, the Milroys posited that individuals who participate in close-knit networks have less of a chance of accepting a new linguistic change from the outside.

Because the ties among individuals are strong in such a network, the vernacular norms of the network are strongly enforced. Thus, a close-knit network can be a conservative

52 force since it refuses the changes that come from outside their network. Therefore, vernacular uses are maintained over time, regardless of the negative prestige, stigma and/or low status that they may have in the general community because the external society does not define the norm for a close-knit community; rather, the members of such a community determine this. These vernacular forms persist in their use regardless of the pressure from the outside standard norms. Thus, the forms used in the close-knit communities are maintained due to the covert prestige within the community. On the other hand, individuals who participate in an open network are more likely to admit new linguistic changes in their language variety. A loose-knit network is often associated with linguistic change. Because the individuals’ ties do not encourage language enforcement, individuals who are members of loose-knit communities are more likely to accept linguistic changes that originate from outside their linguistic network. These changes, however, are not necessarily in favor of the standard dialect. Milroy & Milroy

(1985) have suggested that those changes could be in the direction of the non-standard dialects. In loose-network communities, innovators are those individuals who are most likely to establish many weak association ties outside their network. These individuals are more likely to introduce innovations in the community norms. A particular characteristic of loose-knit networks is the “language missionary/språkmisionœr

(Steinsholt 1962, cited in Trudgill 1986). These speakers play a very important role since the linguistic changes they bring with them are not discarded as would have been the case for innovations brought to the community by outsiders because other members of the speech community view them as insiders and also as locals to their speech community.

The speakers of such a loose-knit speech community, then, accommodate to the

innovations brought to the area by the language missionaries. These innovations are judged to be prestigious and sophisticated as opposed to the dialect spoken by the people who stay for long periods of time in the community. Steinsholt (1962) explains

(Trudgill’s translation 1986):

The urban dialect spreads into Hedrum partly as a result of the influence of particular individuals living in different parts of the area. Such individuals –we call them “language missionaries”— may be village people who have been particularly heavily influenced by the urban dialect. The most important language missionaries are first the young girls who come home after living for a while in the town, and secondly the whalers.

The possible role of “language missionaries” in the Lorain community was explored.

There are several Puerto Ricans in the community who travel very frequently between

Puerto Rico and Lorain, as well as other Puerto Ricans who visit Lorain either from urban and rural Puerto Rico, from other Puerto Rican communities within the United States, or who travel to and from Lorain to work in other surrounding urban centers. They keep the language alive and thus may influence Lorain Spanish with the new forms they may adopt from outside and bring to Lorain.

4.2.1 Network case studies

Examining several case studies on urban linguistic variation and the shift from rural to urban settings is useful because of the comparison they offer with the situation in Lorain.

4.2.1.1 Rural migrants to the urban community of Brazlândia

Bortoni-Ricardo (1985) conducted a quantitative social network analysis in an urban

Brazilian monolingual but multi-dialectal situation. In this study, patterns of adaptation to the urban dialect by rural migrants were revealed. Bortoni-Ricardo’s theoretical framework relied on the assumption that linguistic patterns of behavior are explained by the individuals’ social network. These contacts are susceptible to statistical analysis.

The rural migrants to the urban community of Brazlândia went through two different stages of linguistic adaptation according to Bortoni-Ricardo’s research. In the first stage, the migrants belong to a close-knit migrant group that does not admit outside linguistic influence from the rural community. At a later stage, the migrants who originally were part only of an insulated rural community interact with members of the mainstream urban community where they are immersed. The “integration index” and the “urbanization index” were used as predictors for the degree of dialect diffuseness. The integration index is based on the number of acquaintances an individual has within the migrant community and with members outside of this community who are part of his original group of acquaintances. Thus, based on this quantification, the index defines the indicator of the individual’s transition from a migrant transitory community in an insulated network of fellow migrant acquaintances into a less homogeneous and more integrated urban community. Bortoni-Ricardo’s hypothesis is that the more integrated into the urban community an individual is, the more diffuse is the dialect. The more integrated the migrant individual is in the urban community and the more complex the ties become with other members of that community, the closer the dialect will become to the more prestigious standard dialect and the farther it will be from the less prestigious

rural dialect. The urbanization index is a profile of each of the contacts of the

individual's network. An average was obtained from seven variables that were the

indicators of the individuals’ transition from a rural to an urban network: schooling level, work category, spatial mobility, and participation in urban events, media exposure, political awareness and recruitment social framework.

Bortoni-Ricardo’s work is relevant here because it is possible that the level of integration of the Puerto Ricans in Lorain may be a factor in evaluating their linguistic changes.

Therefore, I propose that more complex ties within the community encourage the speakers to use a rural PRS variety, resulting in the maintenance of the local dialect within the community. In the Methods Chapter (Chapter 5), I explain how I developed an index to determine their level of integration in the community.

In the particular community under study, the more integrated a person is, the closer the

Spanish is to IPRS. However, when a person has many Mexican American acquaintances, that individual's dialect is either more homogeneous/more standard or more Mexican American-like despite how integrated (she) he may be. Furthermore, factors such as language attitudes, identification, solidarity, in-group influence and dialect maintenance need to be considered when studying the Lorain community of

Puerto Ricans. It is through direct yes/no questions (i.e., about their behavior, etc.) (see

Chapter 5) that I assess these issues among the Puerto Ricans and the Mexicans in Lorain.

4.2.1.2 Rural communities in Transdanubia, Hungary and Oberwart, Austria

Gal (1998) studied the networks of a rural community in Transdanubia, Hungary (Bóly)

where German has become increasingly salient. German is the symbol of a superior form

of society and opens doors for better economic opportunities.

Gal (1998b) demonstrated that a network-based approach is suitable for analyzing a

situation of bilingualism, language contact and language shift in Oberwart, Austria.

Oberwart has undergone urbanization and industrialization following World War II, a

process that is characteristic of the European countryside. With this process, languages

have acquired new values that were not associated with them prior to the war.

Nowadays, Hungarian represents peasant status, which is not as respectable as working-

class status. German, on the other hand, has a strong symbolic value representing money

and prestige only available to workers, not peasants. The language choice that members

of the Oberwarter community make depends on their social networks. The number of

peasant acquaintances in an individual’s network is a reliable predictor of (her) his choice

of language. The more peasants an individual knows, the more likely the individual will use Hungarian in more environments. If an individual interacts more with workers, the use of German is more likely.

Lorain, as Oberwart, has undergone industrialization since the 1940s. Many if not all of the Lorain participants in this study consider AE as a way to improve their quality of life, just as the Oberwarter community viewed German after the war, as discussed above. The

language choice an individual makes, thus, depends on (her) his life expectations, the

number of Spanish and/or AE speakers in the network of friends, and the type of life- style (i.e., school, work, etc.). In Chapter 5, I explain how I obtain information about the speakers’ language choice, and about the Spanish/AE speakers in their network by using a questionnaire (see Appendices D and H).

4.2.1.3 A Vernacular Cultural Index

Edwards (1992) developed a proto-network analysis. His “Vernacular Cultural Index” included items such as where the respondent’s kin, relatives and friends live; where the respondents work; the frequency of interaction with neighbors, black and white; desire to remain in the neighborhood; disapproval of street culture; and suitability of the neighborhood for child-rearing. Edwards found that younger speakers used fewer Black

English forms than older speakers in the group. The results from Edwards (1992) show that the older males distanced themselves from the young males of the community, forming a dense social network. It thus offers a potential parallel to the Lorain situation of language maintenance in the first generation and language loss in the younger third generation. The social networks of the younger members of both communities also help to explain the situation since they are in contact with the mainstream society, while the elders are more isolated and form a close-knit network.

4.3 Class-based approach

The class-based approach was proposed by Labov (1972) and conceives of society as

having class divisions. Thus, in this view, all language changes are to be analyzed in

terms of class differences; even gender differences are interpreted in this way. This

approach takes as a background the idea of a dominant market, later articulated by

Bourdieu (1977, 1982), to which speakers aspire to belong.

Labov (1966, 1972) based his views on social pressures that continuously operate in the

language. Labov argues that changes and developments within a language cannot be

understood if the social aspects of the community are not considered. With these

purposes in mind, natural data is collected from speakers with a diversity of

socioeconomic backgrounds. A linguistic variable is chosen using the following criteria:

frequency of usage, integration in the linguistic system and saliency of the variable under

study. Then, the interviewer obtains data through observation of casual situations and

through a formal interview.

In Labov (1972), the variable (th) was studied in New York City in words like three, thing, where a standard variant [θ], and two non-standard variants, [t] and [tθ], are

possible. Labov correlated these variants with style and social stratification. A sharp

difference between the working class groups and the middle class groups was found. The

middle class groups had higher rates of incidence of the standard variant, while the

working class groups had lower incidence rates. Although this difference was found between these socioeconomic classes, there was an internal pattern within each group,

59 i.e., casual style yielded more cases of the non-standard variants for all social groups, while the standard variant was found more among all the social groups in the most formal style. Several other studies have been conducted using a class-based approach to explain patterns of linguistic variation, including Shuy, Wolfram and Riley (1968), Wolfram

(1969), Sankoff and Cedergren (1971), Cheshire (1978) and Trudgill (1974, 1983) among many others.

4.4 Methodological similarities and differences between a class-based approach to linguistic variation and social network analysis

The difference between a class-based approach (Labov 1972, Labov & Harris 1986) and a social network analysis is that in the latter, study groups are based on group interactions. The class-based studies are based on the hierarchical class/economic inequality of society, which arise from an inequality in wealth and power. A social network analysis is anchored in the interpersonal relationships of the individual, where the analysis looks at the solidarity relationships based on the daily contacts in the usual daily lives of the speakers. The argument for using a class-based approach in urban sociolinguistics is that language is a form of social and cultural capital that could potentially become economic capital (Dittmar, Schlobinski & Wachs 1988). A social network approach assumes that linguistic behavior is affected by a diversity of alternative markets, instead of one single dominant market assumed by Bourdieu (1977, 1982).

Thus, in the latter, all speakers have the same model they aspire to imitate, while in the former, speakers have a choice of markets to which they can aspire. Furthermore, in a class-based approach, it is assumed that the model to follow is the language assumed by

society as being the standard promoted in the school system. On the other hand, a social

network approach assumes that the speaker will have ties with other people with whom they will alternate the code choice. More likely, the informal nature of such interactions encourages the use of the non-standard dialects. This non-standard dialect maintenance is the answer to the question of why non-standard dialects are maintained despite the constant pressure to substitute them with the standard dialect.

A class-based approach neglects the differences within a social class and assumes homogeneity within it, while, at the same time, it assumes that there are differences between social classes. The concept of social class is designed to explain structures and processes in society, politics and economics. A social network approach sees an interpersonal relationship within and between the social classes. These social networks between individuals provide them with ways of solving their daily problems. The social class approach works at the macro-level, while the social network approach works at the micro-level.

A class-based approach allows for generalizations, but overlooks important information.

A social network analysis does not merge individuals’ speech, while the class-based approach loses the information about each individual’s speech when they are merged in averages. A social network analysis studies the social ties of the individuals, exposing the complex relationships between them, while a class-based approach assigns individuals to discrete groups, ignoring the rather complicated interpersonal relations.

There are also some similarities between these two methodologies. In the class-based

approach, one collects a random sample and focuses on the linguistic behavior of isolated

individuals. The researcher does not have information about the individual’s networks of

linguistic interaction. This must be hypothetically reconstructed. A similar situation

happens with the study of networks. Although these studies take into account the

behavior of an individual and the individual's contacts, a given study typically is limited to a small selected group of several isolated small networks within a community.

Labov (2001) suggests creating a “judgment sample” of different main ethnic groups,

social classes, and neighborhoods. In the case of a minority speech community, the

researcher could create a “judgment sample” of the different groups within the

community (i.e., age, gender, social classes, and religious or community groups). With a

“judgment sample” that is representative of the minority group, the researcher will obtain

reliable data and explain the community’s complex linguistic situation.

4.5 Assumptions about the social conditioning of linguistic variation associated with a

class-based approach and a social network analysis.

One common misunderstanding is to think that a class-based approach and a social

network approach are in conflict. On the contrary, the two approaches complement each

other. A random sampling of a community concentrates on individuals and yields

unbiased results. With a careful and in-depth study of the speech community and its

social networks, the fieldworker can observe participants and the individuals (i.e., relatives, friends, co-workers, etc.) with whom (she) he usually speaks. With the use of this method of data collection, the researcher can obtain a complete picture of the whole community (Labov 2001).

Moreover, a study focusing only on the networks of a community will not yield the needed information about the individuals’ level of education, income, occupation, and ethnic group that can be obtained from a class-based approach (Labov 2001).

Milroy & Milroy (1992, 1999) clarify how both approaches are integrated into an

analysis similar to the one proposed by Højrup (1983). In general, a social-network analysis studies the individuals’ relationships among each other. But, these interactions

are interpreted within a larger socioeconomic structure. An approach that looks at the

individual’s socioeconomic status by considering level of education, income, occupation,

and ethnic group will yield a more complete and reliable sociological description, not

only of the participant, but also of the community; e.g., an individual with a high income

level may not necessarily have achieved a higher level of education; as a consequence,

the individual may have a lower proficiency when reading and writing (her) his native

language, although (she) he is able to speak and comprehend it. Moreover, life within

each social group is dependent on the class to which an individual belongs. For instance,

a member of the upper middle class is more likely to belong to a sports club than a

member of the working class. Likewise, holding two jobs or working night shifts is more

common among members of the working class than among members of the upper middle class, although not completely unlikely.

Also, using the social network approach is particularly helpful in communities that are

more heterogeneous, such as minorities and immigrant communities, e.g., Black and

Latino (Edwards 1992: 94-95):

The SNT [Social Network Theory] model is capable of revealing intracommunity sociolinguistic patterns that social class stratification would have missed (as the respondents all belong to the working class) and that groups within speech communities could adopt nonprestige linguistic habits as group norms.

Both the upper/elite class and the lower working class consist of social networks that are similar in their interactions characterized by their close-knit, highly dense, kinship- oriented nature. However, the population that falls in between those two social classes participates in a more loose-knit community connected by weaker ties. Thus, it is more reasonable to propose an integration of a class-based approach with a social network approach to account for the linguistic differences in all these groups. A social network analysis to study these loose-knit networks can be extended. But, the middle classes that have these networks can only be identified when using a class-based approach. A social network approach gives the researcher an idea of how the interpersonal interactions give rise to correlations between language and class which in turn can give rise to linguistic change.

Research based on an individual’s social networks, such as that conducted by the

Milroys, thus should not replace a social class-based approach to linguistic variation.

Both methods complement each other, coexisting to explain linguistic variation.

Speakers interact in their daily lives in small social networks embedded in larger social classes in the wider society. Members of a close-knit network with strong ties among them are more likely to maintain their dialect and resist change. Those speakers with less strong ties within their communities and with a wider number of ties outside their community are more likely to be at the forefront of language change.

A class-based approach assumes that the social conditioning of linguistic variation is associated with class divisions. Moreover, linguistic behavior is analyzed from a group

perspective. Linguistic patterns of behavior are analyzed as related to extra-linguistic

factors such as style, socioeconomic class, education, age, gender, etc. Then,

generalizations are made in terms of patterns of behavior of the different groups (i.e.,

based on age, gender, etc.) with respect to the use of variables.

4.6 Summary: a unified account of sociolinguistic variation in the Lorain PRS speech community

The quantitative methods used in class-based approaches are valuable in explaining without any bias different aspects of a structure of a community as a whole (i.e., gender, socioeconomic class, style and age) and the distribution of linguistic forms across a

defined population. Qualitative methods are essential to uncover the individuals’

characteristics in their complex social and psychological background. It is by studying the particular individuals that researchers can complement the quantitative methods.

Milroy & Milroy (1992) is an attempt to link the class-based approaches to linguistic variation with a social network analysis of linguistic variation. These researchers have shown that social class organization is related to tight or loose social networks. By considering the weak and strong social ties, different social classes can be explained. For instance, a close-knit network is characteristic of the lower and upper classes. Weaker network ties in density and multiplexity are characteristic of the middle classes. These are consequences of the different life styles of the social classes; the weaker network ties are not accidental, and are what distinguish them.

The two approaches can be integrated into a unified account of sociolinguistic variation

in the Lorain Puerto Rican speech community. Without neglecting the role of the host

community, but, at the same time, not assuming that it will necessarily play a dominant

role, the parameters for the changes that the language of the migrant community is

undergoing can be set. Several sociolinguistic studies have neglected the inclusion of

immigrant communities presuming that it is the host community that will set the direction

for change in the language. It is nonetheless true that in many cases the younger generations tend to acquire the dominant language without an accent, as is the case of the

Greeks and Italians in Sydney, Australia (Horvath 1998). It is also important to take into consideration that immigrant groups have unstable life histories when compared to speakers of the dominant language communities. Thus, their patterns of linguistic variation may be different from those of other speech communities (Kerswill 1994). In the case of the Puerto Ricans in Lorain, the second generation’s speech can also be influenced by the other dialects spoken around them. Their speech shows a great similarity to that of the first generation, but there is also great variability between and within speakers.

In Horvath (1998), the role of the immigrant community is questioned in terms of its contribution to change in the dominant language. It is assumed, Horvath argues, that the host community is homogeneous in their speech patterns. In this way, the role of the immigrant group is regarded. Horvath (1998) demonstrates the contrary. Moreover, taking this argument into consideration, then, the role of the host community must be considered in terms of its role in language change in the immigrant community. That is,

one must recognize a speech community (i.e., the immigrant community) within a larger

one (i.e., the host community). Therefore, social networks of the individuals within the

Puerto Rican community of Lorain need to be explored, as well as the social networks outside the community connecting the LPRS speakers with members of the AE and MAS communities. It should also be taken into consideration that a single speaker can belong to different speech communities, e.g., both the local community and the larger speech community. Such a multiple membership is typical of ethnic groups, which relate to the larger mainstream community, e.g., Puerto Ricans share linguistic norms that are characteristic of a ‘smaller’ ethnic community, but they also share linguistic norms that are characteristic of the mainstream community. The types and the density of networks of the Puerto Rican speech community shed light on the phonological variation found already in their variety of PRS.

In conclusion, by making use of a combined approach of linguistic variation based on social network and class-based approaches to linguistic variation, researchers obtain a

more comprehensive view of the community’s language use, and gain a clearer view of

the language/s and the social interactions within society.

CHAPTER 5

METHODS IN DATA COLLECTION: DATA COLLECTION WITH SOCIAL

VARIATION

5.1 Introduction

The data gathered for this project are of two types: (1) socio-historical data on the community and patterns of interaction within the members of the group collected by long-term observation, and (2) linguistic data collected by questionnaires and sociolinguistic interviews.

5.2 Data collection

A total of a hundred four people were recorded for this study. These participants represent four different groups: (1) Puerto Ricans, (2) Mexicans, and (3) American

English speakers in Lorain and (4) Puerto Ricans in Puerto Rico.

The speakers who participated in this study were self-identified as members of their

communities and these identifications were comfirmed by their acquaintances and

friends. The participation of all these individuals depended on their availability and the contacts that I, as the interviewer, was able to make. Also, some of the participants had already introduced their relatives, friends and/or acquaintances to me. Many of these individuals volunteered to participate in the study too. The sample is balanced in terms of gender and generational group.

5.3 Three generations in the community

I pay particular attention to the three different “generations” of Puerto Ricans already found in the community of Lorain, Ohio, all of whom display variation in their phonology. The data are obtained from speakers of each “generation”, defined in a way that differs from lay knowledge, i.e. not in the same sense as age and parentage, but rather, following the same criteria that Silva-Corvalán (1994) followed in her study of the

Mexican American community in Los Angeles: (1) the first generation consists of people who were born in Puerto Rico and who moved to the United States after age 1216, (2) the

second, of those who were born in the United States and whose parents came from Puerto

Rico, and (3) the third, of those people who were born in the United States, whose

parents were also born in the United States, but whose grandparents were born in Puerto

Rico. Age twelve was the dividing point, following Lenneberg (1967), who claims that,

after this age, it is more difficult to acquire a language without an accent due to

neurological changes such as neurofunctional reorganization, loss of brain plasticity,

16 This means that a person who arrived to Lorain in the late 1940s when (she) he was a young adult will belong to the same generation as an individual in (her) his mid-thirties who has moved to the area in the last decade.

hemispheric specialization, or “lateralization of function” (Lenneberg 1967; Flege 1999).

It is also assumed, following Lenneberg (1967), that the period of linguistic maturity in a

speaker’s native tongue for the most part has been completed17.

However, other considerations have to be taken into account when discussing language

acquisition. According to Brown (1993), the success that a speaker has in acquiring a

second language depends on (her) his motivation and highly positive attitudes towards

the culture of the second language being acquired. For instance, some highly motivated

individuals have begun learning a second language past the age that has been traditionally

identified as the end of the critical period to acquire a second language and these

individuals have acquired a foreign language without a foreign accent (Bongaerts 1999).

This is a factor that must be considered, particularly for those participants who are third-

generation speakers. These individuals may not necessarily have been exposed to

Spanish at home, and may not be motivated to use this language because they do not see the need to do so. However, there may be other speakers of the same generation who had a similar linguistic situation at home, but who, because of their work environment, or because they need to accommodate to the Spanish of their elderly grandparents, had either learned a heritage variety of Spanish from their parents and/or grandparents, or have learned a standardized Spanish in school (see Chapter 3 for some examples of speakers in Lorain). Moreover, other factors that play a role in acquiring a language

17 Much controversy has been raised with regard to the age of acquiring a language without an accent. I refer the reader to studies on the Sensitive Period Hypothesis, a weaker formulation of the Critical Period Hypothesis, that describes the stages in acquiring a second language as gradual with more variations in its end result (Greenough 1986, Selinger 1978, Oyama 1976, Kim 1994, and Eubank & Gregg 1999 offer an explanation about the distinction between the two formulations) and on alternate explanations (Flege 1999, Bongaerts 1999). Age twelve was a convenient cut-off point considered at the data collection stages of this project, although it turned out not to be as essential as expected as participants could be clearly classified as members of their respective generations.

pertain to the community itself, and not so much the individual; among these community

factors Winford (2003) mentions the speaker’s attitudes to the community and culture of

the language being acquired and the degree of integration and acceptance among the

speakers of that language.

5.4 The youngest generation

The youngest generation is usually identified as the one at the “cutting edge” of new dialect formation (Trudgill 1986, Chambers 1995, Kerswill 1994, Eckert 1997, 2000).

Due to the social pressures that adolescents experience during puberty (e.g. peer pressure, the need to assert their individuality, the desire to gain their independence from parental supervision, etc.), teenagers tend to adopt new forms of speech. Thus, in some communities, the members of this younger generation lead the changes that occur in their language. In the community under study, I looked at the phonological variation among all the different groups identified earlier to explore their role in the phonological variation of the community.

5.5 Participants

The sample for this study was selected through random sampling and through the networks of friends and relatives. This sample is not the result of “judgment sampling”

(Milroy 1995). I did not determine and choose the speakers based on a number of factors pertaining to their speech and their social situation. Rather, the participants in this study

–after having been selected through random sampling or through social network— identified themselves as either Puerto Ricans or Mexicans and were identified as such by

other members of the community. There were no cases where a Puerto Rican or a

Mexican did not consider (him) herself Puerto Rican or Mexican, but was identified as a

member of that group by another individual in the community. However, there were

occasions when individuals acknowledged not having a good command of Spanish or not having knowledge of Spanish at all. In the latter case, the individual was not recorded

since several sections of the recording involved the use of Spanish. However, in the

former, the speakers were still asked to read the word list and the paragraph, while the

conversation continued its course in AE, if the participant felt more comfortable doing so.

The recordings of third generation speakers who do not have a proficient command of

Spanish raise questions as to whether this group is even worth studying if they do not

know Spanish. At best, their data present evidence for the argument that Spanish in

Lorain is slowly dying, being threatened by the pressures of AE and by the lack of its

transmission from the older to the younger generations. Although conscious efforts are

taking place individually, in religious congregations and in the school system, in practice,

the majority of the third generation is AE dominant, with passive comprehension of

Spanish.

The speakers who participated in this study belong to different ethnic groups: Lorain

Puerto Ricans, Mexican Americans, AE speakers and Puerto Rican Islanders. These

groups were subdivided in terms of gender and generation.

5.5.1 Lorain Puerto Rican speakers and data collection

The members of the Puerto Rican community were divided in three different generations.

All sixty subjects were divided into evenly distributed groups, twenty – ten males and ten females for each generation. A final aspect of the data collection in the Puerto Rican community of Lorain was an examination of the networks in which each individual participates.

From data previously collected in Lorain, it had been found that the few younger speakers recorded from the first generation use more AE than Spanish in their conversations. On the other hand, members of the third generation usually use AE in their conversations, but some of them use Spanish instead. The reason for this difference in language usage may lie in their networks, i.e., who they are associated with in the community, as well as in their age group where, depending on their age, different pressures could be at play when deciding which language to use. Another aspect of language to be explored is whether Spanish is maintained in the community, if the youngest speakers are shifting to AE completely, or if a more stable bilingualism develops in the community.

5.5.2 Puerto Rican Islanders and data collection

In addition, speakers from different rural towns in the island of Puerto Rico were recorded in order to document the phonology of Island Puerto Rican Spanish as it is pronounced today and to develop a sense of the state of the dialect close to the time of migration to Lorain. My interest was in collecting samples of formal and colloquial IPRS

particularly of those speakers with a rural and/or working-class background who live in the same towns where the Lorain Puerto Ricans used to live before moving to the United

States. To these ends, I traveled to five of the rural hometowns of the Puerto Ricans who

moved to Lorain, namely Lares, Utuado, Guánica, Guayama and Aibonito. Due to the

limitations of this project, I had to restrict myself to collecting data from only four people in each of these five towns. From each town, I recorded two females and two males who belong to two different generations, for a total of twenty speakers, altogether. They were

divided in terms of gender and age, and moreover, individuals within each age group

matched each other in terms of level of education and exposure to AE. The only

exception was in the case of the older male and female in Guánica, where the female had a higher degree of education than the male, i.e., she was a teacher, while he finished only

eighth grade. I made contacts with the participants through people who were either their

friends or their acquaintances. In cases where I could not find a specific match for a

participant, I went to the town in question and asked several individuals if they would be

willing to read and converse with me about their life in Puerto Rico while I recorded our

conversation.

An in-depth analysis of each individual network was not done for the Puerto Rican

Islanders since the purpose of the data collection there differed from that for the Puerto

Ricans in Lorain.

5.5.3 Lorain Mexican American speakers and data collection

In order to understand the phonology of Mexican American Spanish in Lorain today and

to determine if this phonology has influenced in any way the phonology of Lorain Puerto

Rican Spanish, speakers from the Mexican American community were recorded. With

these purposes in mind, a total of twenty speakers were recorded. These speakers were

divided in terms of gender and age (with one younger and another older group).

The levels of bilingualism for this ethnic group present similar variation to that for the

Puerto Rican group, but with a different language maintenance situation. For instance, in the older group, I interviewed a woman who is fifth generation Mexican American in the

United States and speaks Spanish natively, but who has no proficiency in AE. On the other hand, there are also balanced bilinguals in the older generation who can switch between both languages, showing full command of both AE and Spanish.

5.5.4. American English speakers and data collection

For the purposes of this project, a small number of Lorain AE speakers, specifically four, were recorded. These speakers were evenly divided in terms of age and gender. Their recordings provide an idea of the character of AE as spoken in the area. They offer a baseline, a control, against which to judge any possible influences from AE in Lorain

Puerto Rican Spanish. Similar to the division established for the Lorain Puerto Rican group, the four AE participants represent two different groups: the old and the young.

The age divisions in the Puerto Rican Islanders, the Mexican Americans and the AE speakers groups differ from the way the Puerto Ricans in Lorain were divided by generation group. Thus, to compare these groups, the three generations in Lorain were recoded for age, taking 49 years as the dividing age between old and young.

Since the AE recordings were to provide an idea of specific characteristics of AE, these samples were not used for the same statistical analysis as the IPRS, LPRS and MAS18.

5.6 Interviews

Several interviews were conducted, one per individual, for a total of a hundred four. All of these were recorded on 65-minute digital tapes using a SONY digital audio recorder

PCM-M1 and a SONY lapel electret condenser microphone. The recordings were normalized with Cool Edit, burned and stored on compact discs for transcription and acoustical analysis with Cool Edit, PRAAT and/or Audacity.

18 I, however, listened to all the recordings and determined that AE, as spoken in Lorain, does not offer any cases of (e) or (o) raising, and that it has a retroflex [], a labiodental [v] and a semiconsonant [j]. As mentioned above, the results are a control which I used to judge AE influence in LPRS.

The main goal of these interviews was to obtain both the formal and the most casual speech from the participants, as well as their intuitions, judgments and information about language choice. With this in mind, I followed the sociolinguistic paradigm of interviewing techniques for data collection developed by Labov (1966) and the participant observation techniques first developed by Blom and Gumperz (1972) and later adapted by Labov and his colleagues (Labov, Cohen, Robins and Lewis 1968).

For all the groups, I followed the same steps, i.e., I asked the participants to read a series of three word tables and a short paragraph (see the Appendices). The word tables were read twice so that there were two instances of each word in case the first word recorded was not clear enough. Immediately after the reading of the paragraph, I engaged the participants in a short informal conversation dealing with different topics: hobbies, school experiences, and the like. These conversations were casual enough so that even the participant’s friends and relatives would also sometimes join us in our interaction.

These interactions usually lasted for an hour, though some were as short as forty-five minutes, and others were longer, the longest ones being one hundred five minutes.

During the data collection, there were some times when the participants started immediately talking to me, instead of reading the word tables; in such cases, I continued conversing with the participants, later asked them to read the word tables and the paragraph, and –soon after-- continued with our casual interaction.

I started with the word elicitation because our interaction was usually formal at the beginning, which is similar to the type of context that this exercise establishes. From

there, changing into a less formal environment was easier than moving from a formal

context into an informal one and then back to a formal context. However, in several

cases, the interaction was informal from the very beginning, and it only switched to a

more controlled speech when the participant was asked to read the words and the

paragraph.

During our interaction, the participants filled out a questionnaire, a copy of which is

given in Appendices D and H. The questionnaire collected self-reported data dealing with their language usage with different people and in different environments and situations in the community. Also, this questionnaire asked for a self-evaluation of speakers’ command of AE and Spanish, as well as their language education in both languages and their attitudes toward language use in the community. The participants were asked to mention the names of people from the Puerto Rican, Mexican American and Anglo communities with whom they meet most frequently.

5.7 Interactions with the participants

Mostly through interactions with members of different religious congregations, through contacts with several steel mills workers and through the school system, I established the networks that helped me to develop this ethnographic study and to understand the patterns of contact and their influence on language variation. The information obtained through this type of contact completes the information obtained with the questionnaire,

the participant observations, and the sociolinguistic interviews.

Many of the interviews took place at the individual’s home, but others were conducted in

church, at festivals, in convenience stores or in schools. Those interviews at home were carried on even when a speaker was preparing lunch, doing the dishes or having visitors.

Interviews that took place at home, in church, convenience stores, festivals, work or school often included interactions with the speaker’s relatives, friends, classmates/coworkers and/or acquaintances. During these interactions, I also noted the language choice made by the participants with the people with whom they spoke.

Moreover, I also looked at the relationships informants had within their families and at the languages they used at home. For these purposes, I talked directly about the topic, not only with the people who I already knew from the community, but also with several

of their relatives (e.g. offspring, spouse, siblings, and others), in order to have a complete account of the language/s used.

5.8 Data transcription

To control for the transcription, I selected only ten minutes of our conversation after the

first five minutes that immediately followed the paragraph reading. In those cases when

the participant conversed with me from the beginning, instead of reading the word list

and the paragraph, I waited fifteen minutes and transcribed the following fifteen minutes

of conversation.

Different techniques can be employed when transcribing data: orthographic

representation, “pseudo-phonetic”, and completely phonetic transcription. Since the

focus of this study is the sound variation in certain LPRS variables, a “pseudo-phonetic”

transcription sufficed since it captures an approximate pronunciation of those variables

not under study, while giving a precise phonetic transcription of the features observed for

the study, expediting in this way the transcription process. In cases where the feature

under study is identified, this specific sound is transcribed with an International Phonetic

Alphabet (IPA) symbol.

5.9 Data analysis

Once the data was collected, it was transcribed. Then, a spectrographic analysis of these

sounds was done using PRAAT4022, Cool Edit, and/or Audacity. This analysis made

possible the identification of the variants of each variable.

On cases that were hard to determine if the sound in question was a retroflex [], I looked

at the higher formants because retroflex sounds have been observed to affect these (Fant

1968) with a lowered F4 coming closer to F3 for the alveolar retroflex sounds. Stevens

& Blumstein (1975) point out that the retroflex effect is equivalent to grouping F2, F3

and F4 in a somewhat narrow frequency space. Dave (1977) also adds that there are

more extreme transitions when a vowel sound follows a retroflex than when a retroflex

precedes a vowel. More specifically, I looked at the transitions of the formants of the

preceding vowels into the possible retroflex variant. The effect, according to Ladefoged

and Maddieson (1996), is indicative of the tip of the tongue first moving backwards into the retroflex position, then, straightening during the closure phase, followed by a less extreme position after the closure release.

To insure that the variable (b) was correctly labeled and to distinguish between the

labiodental fricative and the labial fricative, I looked at the formant transitions. A

bilabial fricative has been characterized as having an F2 with an earlier onset and moving

to a lower frequency, as opposed to a labiodental fricative which has greater energy in its

spectra (Ladefoged & Maddieson 1996).

To identify the affricate [dΩ]19 in the spectrograms, I chose the sound in question and

verified that a constriction was observed followed by a period of instability or frication,

this would indicate that it was indeed an affricate since an affricate consonant sound has

been described as a sequence of a stop and a homorganic fricative (Ladefoged 1993,

Ladefoged and Maddieson 1997).

Another type of analysis was made of the data. Once the data was collected, an analysis

of the variation within the speech of each participant was done (Milroy 1995, Silva-

Corvalán 1989). Each variant of each variable was correlated with a range of speaker

variables (e.g. gender, generational group), as well as a range of linguistic variables (e.g. previous and following phonetic environment, stressed vs. unstressed syllables, stylistic context). In order to perform a statistical analysis of the data and correlate the dependent variables with the independent variables, a statistical package was used.

19 Although there is discussion in phonetics about what constitutes a real affricate, e.g. how much frication is involved, whether a sequence needs to be a homorganic stop and fricative constitute an affricate (Ladefoged 1993, Ladefoged and Maddieson 1997), I identified the affricate if I observed a homorganic stop followed by a fricative with the stop having a slight opening of its constriction to produce the following frication.

The statistical analysis of variance in the data was performed with GoldVarb (Rand &

Sankoff 1990), a multiple regression package. GoldVarb is an application that carries out descriptive statistical variable rule analysis. The VARBRUL statistical analysis

(Cedergreen & Sankoff 1974; Young & Bayley 1996) tests if extralinguistic factors, classified in different independent factor groups, have a significant statistical effect when two or more linguistic variants of a linguistic variable, assigned to a dependent factor group, are considered. For this project in particular, VARBRUL was used to determine the relation of age, generation, gender and style have to the different linguistic variables selected. The factor groups consisted of the independent factor groups of the phonological variables and their phonetic environment and the extralinguistic social factors (e.g., age, generation, gender and style) as dependent factors.

The application assigns a weight of 0 to 1 to the factors in each of the independent factor groups. If a factor is assigned a weight higher than 0.5, then this factor favors the application of the variable rule. An assigned weight of 0.5 means the independent factor has no significant statistical effect on the variable rule. If the assigned weight is lower than 0.5, it indicates the disfavoring of the variable rule. Chapter 6 provides the details of the statistical analysis for the LPRS, MAS and IPRS data.

CHAPTER 6

ANALYSIS OF PHONOLOGICAL VARIABLES

6.1 Introduction

In this chapter, I discuss five phonological variables in LPRS. For each variable, I first review the literature on their use in various dialects of Latin American Spanish, and then compare the distribution of their variants in IPRS, LPRS and MAS in Lorain. My aim is to document the phonological variation for all these phonological variables and to investigate if LPRS has preserved IPRS features, if it has merged with MAS, or if AE has had any influence in LPRS. The vowel variables include: (e) raising of unstressed /e/ to

[i]; (o) raising of unstressed /o/ to [u]. The consonant variables are (r) with variants [r],

[l], deletion and []; (b) with [b]/[B] and [v] as its variants; and (dΩ) with variants [dΩ], [j] and [i]. For these variables, Standard Puerto Rican Spanish (SPRS) is used as the basis to compare the PRS spoken by the participants as justified in Chapter 1. After each literature review I also present the results of the VARBRUL analyses for Island Puerto

Rican Spanish (IPRS), Lorain Puerto Rican Spanish (LPRS) and Mexican American

Spanish (MAS).

6.2 The vowels

The raising of /e/ to [i] and of /o/ to [u] could potentially be considered as the same phenomenon, inasmuch as both involve the raising of mid vowels. However, I discuss

them separately in the following subsections to facilitate the presentation. Moreover, the

results from the VARBRUL analyses which will be discussed below show that both

vowels present similar statistical patterns.

6.2.1 Introduction to variable (e)

The raising of (e) has usually been described in the Spanish linguistic literature as

carrying a negative prestige (Lipski 1994). The variable (e) occurs in the following words that have the affected sounds in boldface:

LPRS SPRS Gloss "no.če] "nightا] [no.čiا] "βir] "to surviveا.βir] [so.bre.βiا.su.bri.βi] "rior] "anteriorا.rior] [an.teا.an.ti] "te.rio] "cemeteryا.te.rio] [se.menا.se.min] "δor] "aroundا.δor] [al.re.δeا.al.ri.δi]

Analogy plays a role in some instances of the raising of PRS (e). For instance, e and i occur in related verbal forms, for example, the past participle of "to dress" vestido vs. vistío (past participle of "to dress"). Thus, vistío with [i] in the initial syllable could

result from analogy regularizing the verbal paradigms, as for example, vestido>vistido, with root vocalism based on the past form vistió.

In Peninsular or Peninsular-based Spanish, /e/ realized as [i] is found in Asturias and the

Ribera Salmantina. The raising of (e) is also attested in the Canary Islands of Tenerife

and La Graciosa; for instance, Alvar (1959) claims that (e) is más cerrada ("more raised")

in Canarian Spanish than in Castillian Spanish when a nasal follows or when [e] is in

final position (e.g. nochi/noche "night", binagri/vinagre "vinegar", dimi/dime "(you sg.)

tell (imperative) (it to me)", dali/dale "(you sg.) give (imperative) (it to him/her)",

mándeli/mándele "(you sg.) send (imperative) (it to (her) him)".

Matluck (1963) points out that the realization of /e/ as [i] is found in final position in the

central regions of Costa Rica; in Mexico, (e) is described as laxed in final closed syllables

in words such as después, "later". Peruvian Spanish spoken in the highlands reduces the

vowel system by eliminating the distinction between [e] and [i], though Cerrón-Palomino

(1988) attributes this development to the influence of Quechua and Aymara, adding that

it carries a negative stigma. Lipski (1994) describes the fronterizo dialect spoken in the

border of northern Uruguay and Brazil as having variable unstressed-vowel-raising

approximating the Portuguese vowel system, which has /e/-to-[i] raising in word-final

unstressed contexts.

The literature most relevant to the LPRS situation consists of studies of IPRS and MAS.

Álvarez-Nazario (1972, 1990, 1991) and Navarro Tomás (1948) have identified (e)

raising as a feature commonly found in the countryside of Puerto Rico (please see the

map of Puerto Rico presented earlier in Chapter 3) in the rural towns of Lares,

Quebradillas, Maricao, San Germán, Mayagüez, Adjuntas, Utuado, Jayuya, Orocovis,

Vega Baja, Comerío, Caguas, San Lorenzo, and Fajardo. Álvarez-Nazario (1972) points

out that – with the exception of Fajardo - the phenomenon is observed precisely in those

towns in Puerto Rico settled by Canarian immigrants. Álvarez-Nazario (1972) argues for

Canary Island Spanish influence due to the heavy migration by its speakers to rural

Puerto Rico and their continuous contact with PRS speakers in the area. The

mountainous center of Puerto Rico is not the only area of the Spanish-speaking world

where the development of /e/ to [i] is found. Researchers have observed (e) raising in

other Spanish dialects.

With regard to MAS, Espinosa (1930) pointed out that the Spanish of New Mexico has

(e) raising and that it has been attributed to several different factors: when /e/ follows a

palatal (e.g., lechi/leche "milk", cochi/coche "car", nochi/noche "night", calli/calle

"street", valli/valle 'valley'); when it is next to /s/, /n/ or /≠/, (e.g., siguro/seguro "sure", asigurar/asegurar "reassure", incontrar/encontrar "to find", mintir/mentir "to lie",

siñor/señor sir, siñal/señal "sign"); regressive assimilation (e.g., pidir/pedir "to ask",

amiricano/americano "American", lisión/lesión "injury"); prefix confusion (e.g.,

discontento/descontento "unhappy", discubrir/descubrir "to discover")20; by analogy with

other forms in the verbal paradigm (e.g., sirvir/servir "to serve", dicir/decir "to say")21.

With regards to LPRS, Decker (1952) described this dialect as having a closed [e] “half

as often as in S(tandard) C(astillian) S(panish)”, open [e], and “a variety which is

articulated with a glide, beginning as a closed e and terminating open”; the latter was also

used among the Puerto Ricans in Lorain. Although Decker (1952) mentioned that raising

of (e) is found both in Puerto Rico and in Judeo-Spanish, he did not find evidence of this

20 The prefix "dis-" is used in these examples due to analogy with words such as: discontínuo "discontinous" disconforme "not in agreement" and causes confusion in the cases already mentioned.

21 The forms that provide analogical pressure for these forms in the verbal paradigm are: yo sirvo "I serve", tú sirves "you serve", (ella) él sirve, "(she) he serves", yo digo "I say", tú dices "you say", (ella) él dice "(she) he says"'.

vowel change except in two of all the eighteen Puerto Ricans he interviewed in Lorain,

despite the fact that the majority came from Lares, a region where Navarro-Tomás (1948)

and Álvarez-Nazario (1972, 1990, 1991) have pointed out that (e) raising can be found.

Later in the section on “Random observations” in the same study, Decker (1952: 126)

mentions that:

Qui and hombri were heard from subject #10 and “um pato di lagua” from #13. Relaxed e is often closed almost to i; only in the above examples, however, was the change heard to be complete.

Since subjects #10 and #13 are from Lares, Decker’s comment leads me to want to explore the possibility that (e) raising was transported to Lorain, although with a low frequency of usage.

6.2.1.1 Results of the VARBRUL analyses regarding variable (e) in IPRS, LPRS and

MAS

To best describe the problems of distribution of the variants of (e) in the three dialects studied here, I carried out a VARBRUL analysis of the data (Cedergren & Sankoff,

Young & Bayley 1996). VARBRUL was used to determine the relation between the realization of /e/ as [e] or as its raised variant with the phonetic environment, age, generation, gender and style. I also tested the effect of stress or lack of stress since (e) raising has been found in unstressed syllables in IPRS (Navarro-Tomás 1948; Álvarez-

Nazario 1972, 1990, 1991), the factor group “phonetic environment” consisted of stressed vs. unstressed syllables. The factor group “style” consisted of reading and speaking styles. For the IPRS and MAS speakers, age was considered as a factor group. This

88 factor group consisted of an older and a younger group, with 49 years of age as the dividing point between the two. The LPRS group of speakers was divided into three generations, as described in Chapter 5. The results of the VARBRUL analysis presented in Tables 6.1 - 6.6 are from three separate runs, one for each ethnic group.

Tables 6.1 and 6.2 show that (e) was produced categorically in stressed and unstressed syllables among all the Island Puerto Rican speakers considered here, regardless of age, gender and speech style.

Dialect Island Puerto Rican Spanish Variable [e] [i] Total % Factor groups Phonetic Environment Stressed N 3128 0 3128 42 % 100 0 Unstressed N 4256 17 4273 58 % 100 0.001 Total N 7384 17 7401 % 100 0.001 Style Reading N 3375 6 3381 46 % 100 0 Talking N 4009 11 4020 54 % 100 0 Total N 7384 17 7401 % 100 0.001 Gender Male N 3706 14 3720 50 % 100 0 Female N 3678 3 3681 50 % 100 0 Total N 7384 17 7401 % 100 0.001 Age Older N 3887 16 3903 53 % 100 0 Younger N 3497 1 3498 47 % 100 0 Total N 7384 17 7401 % 100 0.001 Total N 7384 1722 7401 % 100 0.001

Table 6.1 VARBRUL percentages for (e) in Island Puerto Rican Spanish (Percentages are rounded out.)

22 See footnote 24 on a comparison of this figure with corresponding figures in LPRS.

Factor Group23 Factor Weight App/Total Input & Weight Style Reading 0.553 1.00 1.00 Talking 0.455 1.00 1.00 Gender Male 0.315 1.00 1.00 Female 0.686 1.00 1.00 Age Older 0.221 1.00 1.00 Younger 0.803 1.00 1.00 Input: 0.999

Table 6.2 VARBRUL weights, application values and input and weights for [e] in Island

Puerto Rican Spanish

Table 6.3 shows the descriptive statistics for (e) among speakers of LPRS, while Table

6.4 shows the corresponding weights. Similar to the Puerto Ricans in the Island, the

Lorain PRS speakers made near-categorical use of (e) in both stressed and unstressed syllables. Lorain males and females, young and old, used (e) in all speech styles.

23 The VARBRUL weights, application values and input and weights for phonetic environment are not presented here because of knockouts, i.e., factors for which applications have 0% or 100% frequency in the data.

Dialect Lorain Puerto Rican Spanish Variable [e] [i] Total % Factor groups Phonetic Environment Stressed N 6418 10 6428 43 % 100 0 Unstressed N 8447 119 8566 57 % 99 1 Total N 14865 129 14994 % 99 1 Style Reading N 8079 96 8175 55 % 99 1 Talking N 6786 33 6819 45 % 100 0 Total N 14865 129 14994 % 99 1 Gender Male N 6943 44 6987 47 % 99 1 Female N 7922 85 8007 53 % 99 1 Total N 14865 129 14994 % 99 1 Generation First N 6288 29 6317 42 % 100 0 Second N 6254 25 6279 42 % 100 0 Third N 2323 75 2398 16 % 97 3 Total N 14865 129 14994 % 99 1 Total (all factors) N 14865 12924 14994 % 99 1

Table 6.3 VARBRUL percentages for (e) in Lorain Puerto Rican Spanish

24 Given that the percentage that the 129 tokens represent, though small in absolute terms, is much larger than the corresponding percentage (0.001) in IPRS, I used a chi-square test on the total for LPRS versus the corresponding totals for IPRS and for MAS, in order to evaluate the possibility of the admittedly small percentage in LPRS representing an incipient innovation. The results, however, were not significant, so no conclusions as to a nascent trend can be drawn.

Factor Group Factor Weight App/Total Input & Weight Phonetic Stressed 0.777 0.99 0.99 environment Unstressed 0.281 1.00 1.00 Style Reading 0.444 0.99 1.00 Talking 0.567 1.00 1.00 Gender Male 0.512 0.99 1.00 Female 0.490 0.99 1.00 Generation First 0.555 1.00 1.00 Second 0.596 1.00 1.00 Third 0.167 0.97 0.98 Input: 0.996

Table 6.4 VARBRUL weights, application values and input and weights for [e] in Lorain

Puerto Rican Spanish

Tables 6.5 and 6.6 show, respectively, the percentages and weights for realization of /e/

as [e] or [i] in Lorain MAS.

Dialect Mexican American Spanish Variable [e] [i] Total % Factor groups Phonetic Environment Stressed N 4180 2 4180 44 % 100 90 Unstressed N 5315 0 5405 56 % 98 0 Total N 9495 90 9585 % 99 1 Style Reading N 4530 70 4600 48 % 98 2 Talking N 4965 20 4985 52 % 100 0 Total N 9495 90 9585 % 99 1 Gender Male N 4390 60 4450 46 % 99 1 Female N 5105 30 5135 54 % 99 1 Total N 9495 90 9585 % 99 1 Age Older N 4615 45 4660 49 % 99 1 Younger N 4880 45 4925 51 % 99 1 Total N 9495 90 9585 % 99 1 Total (all factors) N 9495 9025 9585 % 99 1

Table 6.5 VARBRUL percentages for (e) in Mexican American Spanish in Lorain

25 See footnote 24.

Factor Group26 Factor Weight App/Total Input & Weight Style Reading 0.337 0.98 0.99 Talking 0.651 1.00 1.00 Gender Male 0.397 0.99 0.99 Female 0.589 0.99 1.00 Age Older 0.498 0.99 0.99 Younger 0.502 0.99 0.99 Input: 0.993

Table 6.6 VARBRUL weights, application values and input and weights for [e] in

Mexican American Spanish in Lorain

As Tables 6.1 – 6.6 show, use of variant (e) is near-categorical in all three dialects. IPRS and LPRS are similar in the use of (e), with LPRS maintaining the same pattern of (e) usage found in IPRS. The use of (e) in LPRS is simply a reflection of the patterns found in IPRS and MAS. The data collected from the older generation in Puerto Rico and the data from the first generation of Lorain Puerto Ricans do not show any evidence that [i] is produced in their conversations. More evidence favoring the absence of (e) raising among the first generation to arrive in Lorain is Decker (1952) who found very few cases of (e) raising, although many of the Puerto Ricans he interviewed came from the same regions identified by Navarro-Tomás (1948) and Álvarez-Nazario (1972, 1990, 1991) as the areas where raising of (e) is found. Furthermore, no (e) raising was found among the

Mexicans in Lorain. My IPRS data provide no support for the claim made by Navarro-

Tomás (1948) and Álvarez-Nazario (1972, 1990, 1991) that (e) raising is a feature characteristic of the mountainous center of Puerto Rico, but rather the standard variant is

26 See footnote 23.

95 preferred for /e/ in my data. However, although I collected data from the oldest speakers to have an idea of the type of dialect spoken in c.1950, Navarro-Tomás (1948) was describing the PRS of those speakers who were already the oldest generation of the late

1930s. Thus, there is already an age difference between his data and my data and conceivably a difference in language use between these two age groups.

6.2.2 Introduction to variable (o)

The raising of the variable (o) occurs in the following words where the affected sounds are in boldface:

LPRS SPRS Gloss "βir] "to surviveا.βar] [so.bre.βiا.su.bre.βi] "βie.ra] "to moveا.βie.ra] [moا.mu] "ril] "to dieا.ril] [moا.mu]

Verbs like morir have a /u/ in its allomorph (i.e., murió/murieron/muero). The same situation holds for podrir. Thus, for those verbs at least it is possible that speakers are regularizing the verbal paradigm, via analogy, perhaps in addition to phonetic conditioning for the realization of [u].

The literature relevant to (o) raising in this study includes studies in IPRS, MAS in the

Southwest, LPRS and few other varieties of Spanish. Lope Blanch (1990) found (o) raising in the MAS spoken in different towns and cities in the United States: San

Marcos, Texas; Mora, New Mexico; Tucson, Arizona. Espinosa (1930) argues that the raising of /o/ to [u] is the result of regressive assimilation to /u/ (e.g.,

revulución/revolución "revolution", cumunidad/comunidad 'community'), dissimilation

(e.g., ucioso/ocioso "lazy", usión/opción "option", also with ps>s), analogy with other

forms in the verbal paradigm (e.g., durmir/dormir "to sleep", murir/morir "to die") or

bilabial assimilation; i.e., when /o/ is next to a labial consonant like /m/, it becomes more

labial (a word like almuhada for almohada "pillow"). In a work by Post (1934), a noun

such as buñiga for boñiga "cow dung" and the verbs murir for morir "to die", and durmir for dormir "to sleep" were reported for the MAS of Southern Arizona. Note, here, that the o in dormir does not appear next to a labial consonant, and that [o] is already a

rounded vowel. The explanation here of the labial consonant influence, thus, is not the

best or only possible explanation for the raising of (o), unless what is at issue is the

degree of gravity, because both [o] and [u] are rounded vowels, and thus, [+ grave],

though round to different degrees.

Similar to the realization of /e/ as [i], the phenomenon of (o) raising is found in the central regions of Costa Rica. Agüero (1960) notes that this occurred more frequently when a palatal consonant (e.g., [tS]) is in the preceding environment and when the vowel

is devoiced, and that the phenomenon was more frequently found among women. In

another article, Agüero (1962) reports (o) raising only in hiatus. Lipski (1994) found (o)

raising among older inhabitants who live in isolated areas within Costa Rica. Lipski

(1994) noted that the change is also found in Galician Spanish, as well as in the

“fronterizo” dialect spoken on the border of Uruguay and Brazil. Peruvian Spanish in the

highlands has (o) raising as a result of the influence of Quechua and Aymara accoding to

Cerrón-Palomino (1988), who also notes that (o) raising carries a negative stigma.

The raising of (o) is found in IPRS as well as in MAS. Álvarez-Nazario (1972, 1990,

1991) and Navarro Tomás (1948) explain that (o) raising is found in the rural areas of

Puerto Rico. The development is also argued to be a result of Canary Island Spanish influence on IPRS according to Álvarez-Nazario (1972) because of the extensive migration of such speakers to rural Puerto Rico and the continuous contact between the two groups in the area. Navarro Tomás (1948) and Álvarez-Nazario (1990) point out that

(o) raising occurs in syllable final position.

In LPRS, Decker (1952) found that (o) was closed “according to the S(tandard)

C(aribbean) S(panish) norms” with some exceptions, and was open according to the norm

“in over 50% of the responses”.

6.2.2.1 Results of the VARBRUL analyses regarding variable (o) in IPRS, LPRS and

MAS

I used VARBRUL to examine the relation between the realization of /o/ as [o] or as [u] and the effect of phonetic environment, age, generation, gender and style. Navarro-

Tomás (1948) and Álvarez-Nazario (1972, 1990, 1991) have pointed out that the raising of (o) is found in unstressed syllables in IPRS, thus, the factor group phonetic environment was classified as either stressed or unstressed. I also considered speech style as an extralinguistic social factor, as well as gender. For the IPRS and MAS speakers, age was considered as a factor group, with two age groups, an older and a younger one. LPRS speakers were classified in three different generations, as described in Chapter 5. The results of the VARBRUL analyses presented on Tables 6.7 – 6.12 are from separate runs.

The VARBRUL percentages in Table 6.7 show that the use of (o) among female and male speakers of all ages in Puerto Rico was near-categorical in all speech styles.

Regardless of whether the syllable was stressed or unstressed, IPRS also used [o], instead of (o) raising.

Dialect Island Puerto Rican Spanish Variable [o] [u] Total % Factor groups Phonetic Environment Stressed N 1959 1 1960 30 % 100 0 Unstressed N 4551 42 4593 70 % 99 1 Total N 6510 43 6553 % 99 1 Style Reading N 3303 2 3305 50 % 100 0 Talking 3207 41 3248 50 % 99 1 Total N 6510 43 6553 % 99 1 Gender Male N 3262 30 3292 50 % 99 1 Female N 3248 13 3261 50 % 100 0 Total N 6510 43 6553 % 99 1 Age Older N 3333 43 3376 52 % 99 1 Younger N 3177 0 3177 48 % 100 0 Total N 6510 43 6553 % 99 1 Total (all factors) N 6510 4327 6553 % 99 1

Table 6.7 VARBRUL percentages for (o) in Island Puerto Rican Spanish

27 See footnote 29.

100

Factor Group28 Factor Weight App/Total Input & Weight Phonetic Stressed 0.880 1.00 1.00 Environment Unstressed 0.300 0.99 1.00 Style Reading 0.817 1.00 1.00 Talking 0.179 0.99 0.99 Gender Male 0.406 0.99 1.00 Female 0.595 1.00 1.00 Input: 0.998

Table 6.8 VARBRUL weights, application values and input and weights for [o] in Island

Puerto Rican Spanish

A similar behavior is found among Lorain Puerto Ricans males and females of all

generations, as shown on Tables 6.9 and 6.10. These speakers chose [o] near-

categorically instead of a raised variant [u].

28 The VARBRUL weights, application values and input and weights for age are not presented here because of knockouts, i.e., factors for which applications have 0% or 100% frequency in the data.

101

Dialect Lorain Puerto Rican Spanish Variable [o] [u] Total % Factor groups Phonetic Environment Stressed N 3695 4 3699 29 % 100 0 Unstressed N 9064 48 9112 71 % 99 1 Total N 12759 52 12811 % 100 0 Style Reading N 7717 34 7751 61 % 100 0 Talking N 5042 18 5060 39 % 100 0 Total N 12759 52 12811 % 100 0 Gender Male N 5937 24 5961 47 % 100 0 Female N 6822 28 6850 53 % 100 0 Total N 12759 52 12811 % 100 0 Generation First N 5382 30 5412 42 % 99 1 Second N 5410 13 5423 42 % 100 0 Third N 1967 9 1976 15 % 100 0 Total N 12759 52 12811 % 100 0 Total (all factors) N 12759 5229 12811 % 100 0

Table 6.9 VARBRUL percentages for (o) in Lorain Puerto Rican Spanish

29 As with the e > i developments (see footnotes 22, 24 and 25), I used a chi-square test to evaluate the possibility of these figures representing an incipient innovation in LPRS. The results of the test were not significant.

102

Factor Group Factor Weight App/Total Input & Weight Phonetic Stressed 0.758 1.00 1.00 Environment Unstressed 0.386 0.99 1.00 Style Reading 0.475 1.00 1.00 Talking 0.539 1.00 1.00 Gender Male 0.495 1.00 1.00 Female 0.504 1.00 1.00 Generation First 0.397 0.99 1.00 Second 0.614 1.00 1.00 Third 0.469 1.00 1.00 Input: 0.997

Table 6.10 VARBRUL weights, application values and input and weights for [o] in

Lorain Puerto Rican Spanish

The use of [o] was also near-categorical among the Mexicans in Lorain, regardless of phonetic environment, style, gender or age of the speaker, as shown in Tables 6.11 and

6.12.

103

Dialect Mexican American Spanish Variable [o] [u] Total % Factor groups Phonetic Environment Unstressed N 6378 40 6418 69 % 99 1 Stressed N 2858 0 2858 31 % 100 0 Total N 9236 40 9276 % 100 0 Style Reading N 4716 30 4746 51 % 99 1 Talking N 4520 10 4530 49 % 100 0 Total N 9236 40 9276 % 100 0 Gender Male N 3830 30 3860 42 % 99 1 Female N 5406 10 5416 58 % 100 0 Total N 9236 40 9276 % 100 0 Age Older N 4635 20 4655 50 % 100 0 Younger N 4601 20 4621 50 % 100 0 Total N 9236 40 9276 % 100 0 Total (all factors) N 9236 4030 9276 % 100 0

Table 6.11 VARBRUL percentages for (o) in Mexican American Spanish in Lorain

30 See footnote 29.

104

Factor Group31 Factor Weight App/Total Input & Weight Style Reading 0.401 0.99 1.00 Talking 0.603 1.00 1.00 Gender Male 0.320 0.99 0.99 Female 0.631 1.00 1.00 Age Older 0.500 1.00 1.00 Younger 0.500 1.00 1.00 Input: 0.997

Table 6.12 VARBRUL weights, application values and input and weights for [o] in

Mexican American Spanish in Lorain

The raising of (o) as not found in the data collected in Puerto Rico, and it did not exist in

the dialect transported to Lorain. Otherwise, a conversation style would have yielded

cases of (o) raising among Lorain first generation Puerto Ricans as well as among the older generation of Puerto Rican Islanders. More evidence in favor of the argument that

(o) raising did not exist in IPRS is that Decker (1952) did not find this type of vowel raising among the first Lorain Puerto Rican migrants. Moreover, the Mexicans also do not use (o) raising, contrary to what was found in the MAS of Southern Arizona.

Although these two MAS dialects differ, it must be borne in mind that similarly derived

dialects need not show identical phonologies.

The realization of /o/ was variant [o] is near-categorical in all three dialects. Thus, the

use of [o] in LPRS reflects the same patterns found in IPRS and MAS. Those few cases

31 The VARBRUL weights, application values and input & weights for phonetic environment are not presented here because of knockouts, i.e., factors for which applications have 0% or 100% frequency in the data.

105 where /o/ is produced as [u] are found in the conversations of older Puerto Ricans and the first generation of Lorain Puerto Ricans. This finding suggests a trend toward raising in informal style despite the scattered cases found. Very few cases of (o) raising were found among the Mexicans in Lorain. The data I collected for IPRS do not provide any support for the claim made by Navarro-Tomás (1948) and Álvarez-Nazario (1972, 1990, 1991) that (o) raising is a feature characteristic of the mountainous center of Puerto Rico. In my data, [o] is preferred instead. Moreover, Decker (1952) did not find any evidence of (o) raising. Thus, this variation did not exist in the IPRS transported to Lorain in the late

1940s and early 1950s. Even if there has been any trace of (o) raising in the IPRS transported to Lorain, it seems to have been lost, either because of lack of reinforcement for (o) raising from PRS in Lorain, or from its absence in MAS in the area. Thus, IPRS and LPRS are similar in the use of (o), LPRS has maintained the same pattern of (o) usage found in IPRS.

6.3 The consonants

In each of the next subsections, I present a brief review of the literature and analyze the distribution of the variants of the consonants /r/, /b/ and /dΩ/ for IPRS, LPRS and MAS.

6.3.1 Introduction to variable (r)

The variable (r) in syllable-final and word-final positions covers several variants, including a tap [|] produced with the tongue touching the alveolar ridge once, in free variation with a trilled or multiple [r] articulated on the same region with the tongue

106

touching the alveolar ridge multiple times.32 Other variants are [l] and deletion which

both occur in word and syllable final position. In certain dialects such as Dominican

Spanish, /r/ vocalization is found as another variant. Interestingly, an [r] sound like

American English [] produced with the tongue’s apex going backwards without touching the roof of the mouth has been found in several Central American Spanish dialects. The

retroflex [] articulation is not found in other varieties of Spanish.

IPRS IPRS IPRS IPRS LPRS Gloss [be.Ber] [be.Be|] [be.Bel] [be.Beø] [be.Be] "drink" [fa.Bor] [fa.Bo|] [fa.Bol] [fa.Boø] [fa.Bo] "favor" [a.par.te] [a.pa|.te] [a.pal.te] [a.paø.te] [a.pa.te] "apart" [lar.go] [la|.go] [lal.go] [laø.go] [la.go] "long"

32 The distribution of /r/ in syllable-final and word-final position differs from /r/ in other positions. In syllable-final and word-final position, [r] is in free variation with [|], while in other positions, its variants are conditioned by the phonetic environment, i.e. /r/ has allophones.

107

Many researchers of Spanish in the Southwest have argued that the retroflex [] is a result of AE influence. Ornstein (1974) claims that the retroflexed [] is due to AE interference

and explains that this phoneme is pronounced more like the corresponding AE sound than

the corresponding one found in most other Spanish dialects. Furthermore, Lastra de

Suárez (1975) also attributes to AE interference the substitution of a retroflex [] for

Spanish [|] or [r] in this variety. Sánchez (1973) states that the retroflex [] found in the

informal speech of Texas MAS speakers is a common instance of AE phonological

interference in this dialect (e.g., [carne] [caRne] [R] = (retroflex)) (Standard Spanish

carne ‘meat’). Cassano (1973, 1977) argues that New Mexico Spanish speakers

substitute an AE retroflex for the Spanish flap /r/. In the Spanish of New Mexico, Alonso

(1930) heard an intervocalic and syllable-final /r/ similar to the AE /r/ with less

retroflection. Espinosa (1930: 141) provides a similar description of New Mexico

Spanish.

… Se oye en posición intervocálica o final de sílaba. Se parece un poco a la r final del inglés hablado en el oeste de los Estados Unidos, pero la lengua no se arquea tanto hacia arriba y hacia atrás como en el sonido angloamericano.33

We will see shortly if Lorain Mexican Spanish has also a retroflex [] and if this is similar

to the MAS varieties found in the Southwest.

33 It is heard in intervocalic or syllable-final position. It is somewhat similar to r in final position in the English spoken in the west of the United States, but the tongue does not arch as far upwards and back as in the English sound. [The translation is mine.]

108

As well as MAS, other varieties of Spanish have a retroflex as a variant of /r/ in coda

position. The retroflex variant is found in the central areas of Costa Rica. The retroflex

occurs in syllable-final and especially phrase-final position. Speakers of Costa Rican

Spanish extend the environments where the retroflex appear to those where trilled /r/ is

produced in fast speech (Lipski 1994). Belizean Spanish and Mexican Spanish as spoken

in the Yucatan Peninsula are other Central American Spanish varieties that realize /r/ as a

retroflex [] (Hagerty 1996). Apparently, according to Lipski (1994), the retroflex [] is

more common among Belizean Spanish speakers than among the Spanish speakers of the

Yucatan. The retroflection of /r/ is an innovation in these Spanish dialects when

compared to other dialects of Spanish.

IPRS, like the Caribbean Spanish dialects of the Dominican Republic and Cuba, has [l] and deletion as /r/ variants. Navarro-Tomás (1948) and Álvarez-Nazario (1972, 1990,

1991), in their studies of IPRS, note that /r/ is often pronounced as [l] in syllable-final

position or word-final position, so that, for example, hablar "to speak" is pronounced

.blal]. Retroflection has not been documented as a variant of /r/ in IPRSا.a]

109

In an earlier study of LPRS, Decker (1952) mentions several (r) variants that are

intriguing. In this study, Decker found [r], [l] as allophones of /r/, as well as an intermediate variant34. Furthermore, Decker (1952) mentions that [r] was articulated on

the alveolar ridge, and that “a markedly relaxed r (similar to SAE)” occurs after a vowel,

“so that the first element properly belongs to the preceding r-colored vowel.” The

description that Decker (1952: 116) provides of this [r] variant refers to a tap with a

preceding r-colored vowel, not a retroflex.

R-colored vowels are most common in SAE words like: bear, fur, and perfect; in these words the vowel and the r are a single sound and are [or can be -MFRP] written as such in I.P.A. and SAE phonetic transcription. This r-coloring was found in 10% of all responses, but never without the addition of a true alveolar or prealveolar r.

Decker (1952) also observes that lambdacisms -- "equalization" (in his terms, i.e. neutralization) of /l/ and /r/ -- and /r/ vocalization are found with less frequency in LPRS than in previous studies of PRS.

Figueroa & Hislope (1998) found “other” variants (e.g. semivocalization, the mixed variant, deletion, gemination and the English alveolar flap) and suggest that PRS in

Indiana is undergoing change in its phonetic inventory.

34 “Also may be found a variety of l and r which bears articulatory characteristics of each, the variety being an allophone of the two phonemes l and r. One can best approximate this articulation by forming the phoneme l with the posterior and lateral areas of the tongue while allowing the tip of the tongue to articulate an alveolar relaxed or lightly flapped r.”

110

6.3.1.1 Results of the VARBRUL analyses regarding variable (r) in IPRS, LPRS and

MAS

I tested the effects of extralinguistic factors and phonetic environment on all phonological variants of /r/ using VARBRUL’s statistical analysis. Navarro-Tomás (1948) and

Álvarez-Nazario (1972, 1990, 1991) have pointed out that lateralization of /r/ is found more often in syllable final position in IPRS, thus, the factor group phonetic environment included syllable final word internal and syllable final word final position. The social factors remain the same as before. Tables 6.13 – 6.22 present the results of the

VARBRUL analyses drawn from separate runs.

In Tables 6.13 - 6.14, I present the findings for the distribution of the possible variants of

/r/ in IPRS. Lateralization, [r], retroflection and deletion were the variants found in the data. There were few instances of retroflection and deletion among the Puerto Ricans in

Puerto Rico. The lateralization of /r/ was found in word-final position as has been previously described in the literature (Navarro-Tomás 1948; Álvarez-Nazario 1972,

1990, 1991).

More formal style (i.e., reading) yielded more cases of [r], among IPRS speakers, while talking did not favor the occurrence of [r], favoring [l] instead (70%, N=495). This should not be surprising since [r] is considered to be the more formal variant of /r/, while

[l] is described as found more often in casual speech (see Table 6.13).

111

IPRS females favored the use of [r] more than males, a result that is not surprising, either, and that has been described in the linguistic literature; i.e. women tend to favor more the use of the prestigious variants, as long as they are exposed to them. The Puerto Rican males were almost evenly divided in their usage of [r] (46%, N=476) and [l] (49%,

N=504). Both age groups were evenly divided between [r] and [l]. The older Puerto

Ricans preferred [r] (55%, N=555) over [l] (39%, N=393), just as the younger group, with 54% for [r] and 43% for [l].

112

Dialect Island Puerto Rican Spanish Variable [r] [l] deletion [] Total % Factor groups Phonetic Environment Word final N 549 499 0 3 1051 51 % 52 47 0 0 Syllable final N 566 344 85 1 996 49 57 35 9 0 Total N 1115 843 85 4 2047 % 54 41 4 0 Style Reading N 950 348 36 4 1338 65 % 71 26 3 0 Talking N 165 495 49 0 709 35 % 23 70 7 0 Total N 1115 843 85 4 2047 % 54 41 4 0 Gender Male N 476 504 48 3 1031 50 % 46 49 5 0 Female N 639 339 37 1 1016 50 % 63 33 4 0 Total N 1115 843 85 4 2047 % 54 41 4 0 Age Older N 555 393 53 2 1003 49 % 55 39 5 0 Younger N 560 450 32 2 1044 51 % 54 43 3 0 Total N 1115 843 85 4 2047 % 54 41 4 0 Total (all factors) N 1115 843 85 4 2047 % 54 41 4 0

Table 6.13 VARBRUL percentages for [r], [l], deletion and [] in IPRS

113

Factor Group Factor Weight App/Total Input & Weight Phonetic Word final 0.430 0.52 0.51 Environment Syllable Final 0.581 0.62 0.66 Style Reading 0.681 0.73 0.75 Talking 0.184 0.25 0.24 Gender Male 0.384 0.49 0.46 Female 0.616 0.65 0.69 Age Older 0.513 0.59 0.59 Younger 0.488 0.55 0.57 Input: 0.579

Table 6.14 VARBRUL weights, application values and input and weights for the variant

[r] in Island Puerto Rican Spanish

Tables 6.15 - 6.16 present, respectively, the descriptive statistics and the VARBRUL weights, application total and input and weight values for the realization of /r/ as [r], [l],

[] or deletion in LPRS. All these variants were found in the data collected for LPRS.

As Table 6.15 shows, LPRS speakers follow a similar pattern of usage of [r] as the Puerto

Ricans Islanders. LPRS speakers favored [r] in syllable-final position (52%, N=1082) more than in word-final position (40%, N=889), where [l] was used more (54%,

N=1186). This usage was also observed in IPRS by Navarro-Tomás (1948) and Álvarez-

Nazario (1972, 1990, 1991).

Similar to the IPRS speakers, the Lorain Puerto Ricans preferred [r] in their more formal style (56%, N=1702), but not in their conversations (21%, N=269) where [l] was preferred (61% N=767). Thus, they maintain the use of the more conservative variable in

114 their most monitored form of speech, just as the Puerto Ricans in Puerto Rico do, while preferring to use [l] in their conversations (see Table 6.15). Also, the Lorain Puerto

Ricans make more use of more variants for [r], as opposed to the IPRS speakers. LPRS speakers are deleting [r] in conversations (17%, N=216), while choosing to make use of a retroflex [] when they are asked to read in Spanish.

Both males and females trend with the IPRS speakers. LPRS males and females are evenly divided in their usage of [r] and [l], with the females slightly preferring [r] (49%,

N=1088) over [l] (42%, N=932), while the males present a very slight difference between their usage of [l] (45%, N=932) and [r] (43%, N=883).

The preference to use [r] declines across the three different generations in Lorain, as the first generation favors [r] usage (52%, N=956), while the second generation is barely divided between [r] (45%, N=769) and [l] (47%, N=802), but not so the third generation speakers who prefer [l] (51%, N=381). A retroflex [] was found among third generation speakers (10%, N=78) more than among second generation speakers (2%, N=33). First generation speakers also used retroflex [] (3%, N=52). I will discuss the VARBRUL probability results later.

115

Dialect Lorain Puerto Rican Spanish Variable [r] [l] deletion [] Total % Factor groups Phonetic Environment Word final N 889 1186 4 132 2211 52 % 40 54 0 6 Syllable final N 1082 678 289 31 2080 48 % 52 33 14 1 Total N 1971 1864 293 163 4291 % 46 43 7 4 Style Reading N 1702 1097 77 163 3039 71 % 56 36 3 5 Talking N 269 767 216 0 1252 29 % 21 61 17 0 Total N 1971 1864 293 163 4291 % 46 43 7 4 Gender Male N 883 932 162 74 2051 48 % 43 45 8 4 Female N 1088 932 131 89 2240 52 % 49 42 6 4 Total N 1971 1864 293 163 4291 % 46 43 7 4 Generation First N 956 681 144 52 1833 43 % 52 37 8 3 Second N 769 802 110 33 1714 40 % 45 47 6 2 Third N 246 381 39 78 744 17 % 33 51 5 10 Total N 1971 1864 293 163 4291 % 46 43 7 4 Total (all factors) N 1971 1864 293 163 4291 % 46 43 7 4

Table 6.15 VARBRUL percentages for [r], [l], deletion and [] in LPRS

116

Factor Group Factor Weight App/Total Input & Weight Phonetic Word final 0.419 0.43 0.43 Environment Syllable Final 0.595 0.61 0.61 Style Reading 0.603 0.61 0.61 Talking 0.245 0.26 0.25 Gender Male 0.470 0.49 0.48 Female 0.527 0.54 0.54 Generation First 0.590 0.58 0.60 Second 0.466 0.49 0.48 Third 0.351 0.39 0.36 Input: 0.511

Table 6.16 VARBRUL weights, application values and input and weights for the variant

[r] in Lorain Puerto Rican Spanish

Tables 6.17 - 6.18 show the percentages and weights for the distribution of variant [r] in

MAS respectively. The variants [r], [l] and deletion were found in the MAS data. No cases of retroflex [] were found35. Lorain MAS preferred [r] overwhelmingly over [l]

and other possible variants for /r/. In word-final position and syllable-final position, [r] was chosen in 90% and 89% of the cases, respectively.

Contrary to IPRS and LPRS, MAS in Lorain chose [r] not only for reading (91%,

N=1169), but also in the more informal speech style (87%, N=1055) targeted in the recordings.

35 The study of possible reasons as to why Lorain MAS diverges from Southwestern MAS on this feature in particular is an interesting question in its own right which merit further study. However, such a study is beyond the scope of this project. I refer the reader to page 124 where I discuss some possible explanations for the lack of a retroflex in Lorain MAS.

117

Dialect Mexican American Spanish Variable [r] [l] deletion [] Total % Factor groups Phonetic Environment Word final N 1008 113 0 0 1121 45 % 90 10 0 0 Syllable final N 1216 83 68 0 1367 55 % 89 6 5 0 Total N 2224 196 68 0 2488 % 89 8 3 0 Style Reading N 1169 81 28 0 1278 51 % 91 6 2 0 Talking N 1055 115 40 0 1210 49 % 87 10 3 0 Total N 2224 196 68 0 2488 % 89 8 3 0 Gender Male N 1115 110 25 0 1250 50 % 89 9 2 0 Female N 1109 86 43 0 1238 50 % 90 7 3 0 Total N 2224 196 68 0 2488 % 89 8 3 0 Age Older N 1014 86 18 0 1118 45 % 91 8 2 0 Younger N 1210 110 50 0 1370 55 % 88 8 4 0 Total N 2224 196 68 0 2488 % 89 8 3 0 Total (all factors) N 2224 196 68 0 2488 % 89 8 3 0

Table 6.17 VARBRUL percentages for [r], [l], deletion and [] in MAS

118

Mexican females and males in Lorain made a similar use of [r], with 90% and 89%

respectively. Age did not present any difference, with the younger Mexicans using [r]

(88%, N=1210) just as often as the older generation (91%, N=1014).

The only intriguing finding from the descriptive statistics presented on Table 16.17 is that

MAS in Lorain has [l] (8%, N=196) and deletion (3%, N=68) as possible variants of /r/, and it is similar, although with a lower frequency, to LPRS, suggesting convergence of

MAS and LPRS. Contrary to LPRS, however, no cases of retroflex [] were found in the data.

Factor Group Factor Weight App/Total Input & Weight Phonetic Word final 0.426 0.90 0.90 Environment Syllable Final 0.564 0.94 0.94 Style Reading 0.563 0.94 0.94 Talking 0.433 0.90 0.90 Gender Male 0.469 0.91 0.91 Female 0.532 0.93 0.93 Age Older 0.512 0.92 0.93 Younger 0.490 0.92 0.92 Input: 0.923

Table 6.18 VARBRUL weights, application values and input and weights for the variant

[r] in Mexican American Spanish in Lorain

119

Community IPRS LPRS MAS Total % 54 46 89 Input Probability 0.579 0.511 0.923 Factors Phonetic Environment Syllable Final 0.581 0.595 0.564 Word Final 0.430 0.419 0.426 Style Reading 0.681 0.603 0.563 Talking 0.184 0.245 0.433 Gender *** Male 0.384 0.527 0.469 Female 0.616 0.470 0.532 Generation First 0.590 Second 0.466 Third 0.351 Age * *** Older 0.513 0.512 Younger 0.488 0.490

* = Not chosen as significant in IPRS by VARBRUL ** = Not chosen as significant in LPRS by VARBRUL *** = Not chosen as significant in MAS in Lorain by VARBRUL

Table 6.19 Summary of percentages and comparison of input probabilities and weight values for [r] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican

American Spanish

A summary comparing percentages, input probability and weight values for [r] across the different ethnic groups is presented on Table 6.19. The input values show that [r] is the

most preferred variant by the Mexican speakers (0.923), than by the Island Puerto Ricans

(0.579) and the Lorain Puerto Ricans (0.511).

120

As far as phonetic environment is concerned, there is no difference between IPRS, LPRS and MAS. LPRS speakers follow a similar pattern of usage of [r] as the Puerto Ricans

Islanders. MAS, LPRS and IPRS speakers used [r] in syllable-final position (0.564,

0.595 and 0.581, respectively) and in word-final position (0.426, 0.419 and 0.430, respectively). The preference for [r] to appear in syllable-final position was also observed in IPRS by Navarro-Tomás (1948) and Álvarez-Nazario (1972, 1990, 1991).

Similar to the IPRS speakers, the Lorain Puerto Ricans preferred [r] in their more formal style (0.603), but not in their conversations (0.245). Thus, both groups maintain the use of the more conservative variable in their most careful form of speech. The Mexicans in

Lorain, on the other hand, prefer to use [r] in both speech styles (reading 0.563 and talking 0.433). The IPRS and LPRS prefer to use [l] in their conversations (see Tables

6.13 and 6.15). Moreover, in addition to using deletion and lateralization, LPRS speakers also make use of retroflection, exclusively in reading, which is not found in MAS, and almost never in IPRS (see Tables 6.13, 6.15 and 6.17).

Taking a closer look at the probability weights for gender, the probability weights for [r] among IPRS males (0.384) is rather low when compared to the IPRS females (0.616).

There is a very slight difference in the probability weights of the LPRS male and female use of [r]. These groups show probability weights of 0.527 and 0.470, respectively.

Gender was not selected as a factor group for MAS, however, MAS females and males are also slightly divided in the weights for [r], with the females having a probability weight of 0.532 and the males, 0.469.

121

The preference to use [r] seems to decline across the three different generations of Puerto

Ricans in Lorain, as the first generation favors [r] usage with 0.590 as probability weight, while the second generation is barely divided between [r] and [l] with a probability weight for [r] of 0.466, but not so the third generation speakers who prefer [l] and have a probability weight for [r] of 0.351.

Age was not considered significant for IPRS and MAS. Both groups show similar probability weights (see Table 6.19) for both the older and the younger generations.

Lorain Puerto Ricans are choosing to make use of several /r/ variants, including retroflex

[], when they are asked to read in Spanish. Reading in Spanish is the only style that triggers the use of retroflex [] more frequently among the first and third generations than in the second generation. At least for the first generation, the speakers are aware of the negative prestige that [l] has in Insular PRS, so they use retroflex [] as one of the alternatives to avoid using [l]. The retroflex [] was found among third generation speakers more than among second generation speakers. The retroflex in LPRS can be attributed to the influence of AE, but not to Lorain MAS influence, which does not have a retroflex []. I consider AE to be the source for the retroflex [] in LPRS for several reasons. First of all, AE has a retroflex []. Second and third generation speakers are bilingual in AE and Spanish and these speakers are more literate in AE than in Spanish, thus, they use the English phonology when reading. Already in the early 1950s, Decker

122

(1952) found an [r] that had characteristics similar to the AE (r). Thus, for the first generation, those speakers who had more contact with AE were already including the retroflex [] in their Spanish. First generation speakers interact with speakers of the second and third generations who are bilingual and use both languages with the first generation, particularly with those who have a basic command of AE. For the first generation, then, the retroflex [] has two sources, AE from the AE speakers in the community and from the second and third generations’ AE and LPRS used when speaking in the former and reading in the latter.

MAS in Lorain shows a very different pattern of [r] usage from either LPRS or IPRS. In

Lorain MAS, [r] is the preferred /r/ variant for both speech styles, contrary to the usage of the Puerto Rican Islanders who preferred [r] only in formal speech. Lorain MAS has an input of 0.923 for [r] usage, considerably higher than IPRS (0.579) or LPRS (0.511). As

I discussed above, IPRS uses laterals and LPRS uses laterals and other variants instead of

[r], while MAS makes an almost categorical use of [r] in all the factor groups with an overall 89% as opposed to 8% for [l] and 3% for deletion. On the other hand, contrary to what has been found in the MAS linguistic literature about the usage of a retroflex in the

Southwest (see section 6.3.1), I found not one single case of retroflection among the

Mexicans in Lorain.

Despite the fact that MAS in Lorain prefers [r] over all the other alternatives that LPRS uses, it is important to pay attention to the cases of deletion and lateralization collected in the data. The occurrence of lateralization is higher in more informal conversations (10%,

123

N=115), while reading yielded fewer cases (6% N=81). Similarly, deletion was more often found in conversations (3%, N=40) than in a more formal speech style (2%, N=28).

Thus, it is possible that MAS in Lorain is slowly converging with LPRS in these variants.

The retroflex was not found in MAS. This is surprising, not for IPRS, but for MAS, which has been said to have a retroflex []. In the past (Ramos-Pellicia 2000), I have attributed to MAS influence the retroflex [] in LPRS because this variety of Spanish in the Southwest has been described as having a retroflex [] in the Spanish linguistic literature. However, this literature describes MAS as spoken in the Southwest. Although

MAS in Lorain may ultimately derive from Mexican Spanish because its speakers came from the Southwest, it does not necessarily mean that MAS will have exactly the same sound system. Therefore, MAS in Lorain may well be different in at least that respect from Southwestern Mexican Spanish because each community of speakers has not been in continuous contact with each other and, thus, develop a different version of their dialect; that is to say, it is a result of internal development. On the other hand, the very low frequency of the retroflex [] in LPRS cannot be attributed to MAS influence. The

MAS data herein shows that the situation in the Lorain dialect is different, with the

Mexican speakers not using a retroflex []. Thus, the retroflex [] in LPRS is due to the influence of AE and not of Lorain MAS.

The results from the VARBRUL statistical analysis are also surprising in another way.

Because the Puerto Ricans outnumber the Mexicans in Lorain, one might expect the

124 second group to converge in their language usage with the Puerto Ricans. However, the

Mexicans maintain their usage of [r] in both speech styles, instead of converging with the

Lorain Puerto Rican [r] usage. Lateralization is a characteristic of LPRS and as far as this variable is concerned, MAS are avoiding it to signal their different identity. Earlier tensions between the two groups affect the Mexicans in the community and prevent them from converging with LPRS. Thus, LPRS and MAS in Lorain are diverging in their usage of [r] in an effort to signal their different identities as Mexicans and Puerto Ricans through language. Although I did not collect any evaluations about LPRS from the

Mexican speakers, Puerto Ricans in Lorain perceive Mexican Spanish as a different variety, and both the speakers and their Spanish as a variety that is undesirable to be acquired (see Chapter 3). Comments such as the Mexicans being “wetbacks” or as speaking slang, that the Mexicans have forgotten how to speak Spanish or that they are not able to speak “proper Spanish” (see Chapter 3) lead me to believe that the Puerto

Ricans perceived themselves as different from the Mexicans with regard to language use and so do not share the same patterns of use for [r] as the MAS speakers.

Lorain Puerto Ricans follow the pattern of (r) usage in IPRS, using [l] almost as much as

(r). The Mexicans in Lorain, instead of converging with the usage of (r) by an ethnic group that speaks their same language and outnumbers them in Lorain, follow a different pattern of [r] usage, using [r] in the vast majority of cases, diverging in this way (Otheguy and Zentella 2004) from the Lorain Puerto Rican speakers.

125

In conclusion, Lorain Puerto Ricans have maintained the same patterns of usage for (r) as the Island Puerto Rican speakers, that is, both groups use [l] in informal contexts and also use deletion as an alternative for (r). The gross distribution with regard to percentages, style and phonetic environment are essentially the same –the numbers are not identical, but the trend is. On the other hand, Lorain Puerto Ricans are making use of a retroflex [] when they are asked to read in Spanish. A likely reason for this is AE influence because

LPRS speakers are more likely to read in AE and have received most all of their education in this language. The Mexican Americans in Lorain, on the other hand, differ from the Puerto Ricans in that they use [r] in formal and informal speech styles. MAS in

Lorain also uses lateralization, but without the same frequency of usage as the Puerto

Ricans. Moreover, Lorain MAS does not have the retroflex found in other varieties of

MAS in the Southwest. Thus, the Mexican American and the Puerto Ricans in Lorain do not show a clear pattern of convergence as far as the variants for /r/ are concerned. The overall situation with (r) in LPRS and MAS is different with regard to [l], retroflex [] and deletion, i.e., MAS and LPRS differ in the pattern of variation observed for variable

(r).

Table 6.20 presents a summary of percentages and a comparison of probabilities of [l] for

IPRS, LPRS and MAS. The probabilities show that LPRS favors the use of [l] more

(0.434) than IPRS (0.412). MAS has a very low input probability weight (0.079) because it favors [r] instead of [l]. The results on Table 6.20 show that [l] serves as a social diagnostic of the differences between PRS and MAS in Lorain.

126

Community IPRS LPRS MAS Total % 47 54 6 Input Probability 0.412 0.434 0.079 Factors Phonetic Environment Syllable Final 0.397 0.382 0.432 Word Final 0.597 0.611 0.583 Style Reading 0.329 0.416 0.437 Talking 0.794 0.695 0.566 Gender *** Male 0.607 0.520 0.533 Female 0.391 0.482 0.467 Generation First 0.421 Second 0.542 Third 0.597 Age * *** Older 0.481 0.490 Younger 0.518 0.508

* = Not chosen as significant in IPRS by VARBRUL ** = Not chosen as significant in LPRS by VARBRUL *** = Not chosen as significant in MAS in Lorain by VARBRUL

Table 6.20 Summary of percentages and comparison of input probabilities and weight values of [l] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican

American Spanish

127

A separate VARBRUL run was done for the ten members of the first generation who were more exposed to AE to determine if they were the ones producing the retroflex [] more than the other individuals in their generation who had little or not exposure to AE.

Table 6.21 presents these results.

Dialect First Generation Lorain Puerto Rican Spanish Variable [r] [l] deletion [] Total % Factor groups Phonetic Environment Word final N 196 220 0 17 433 49 % 45 51 0 4 Syllable final N 253 122 67 5 447 51 % 54 29 15 2 Total N 449 342 67 22 880 % 51 39 8 3 Style Reading N 255 114 16 22 406 71 % 63 28 4 5 Talking N 22 118 29 0 169 29 % 13 70 17 0 Total N 449 342 67 22 880 % 51 39 8 3 Gender Male N 66 126 25 2 218 38 % 30 58 11 0 Female N 211 106 20 20 357 62 % 59 30 6 6 Total N 449 342 67 22 880 % 51 39 8 3 Total (all factors) N 449 342 67 22 880 % 51 39 8 3

Table 6.21 VARBRUL percentages for [r], [l], deletion and [] in the ten speakers of the first generation of LPRS with exposure to AE

128

The results of this separate run demonstrate that those who had more exposure to AE, because they use it more at work and with their relatives, produced the retroflex [] in 22 cases or 3%. If this number of cases is compared with the total of number of retroflex []

(3%, N=52 cases) produced by the third generation (see Table 6.15), it is obvious that first generation speakers with more exposure to AE did not produce the retroflex [] more than the other speakers who have little or no contact with this language.

Another separate run was done for the seven speakers of the third generation of LPRS who were Spanish dominant to determine if their great exposure to Spanish had any effect in the use of the (r) variables in their conversations. Table 6.22 shows the descriptive statistics.

129

Dialect Third Generation Lorain Puerto Rican Spanish Variable [r] [l] deletion [] Total % Factor groups Phonetic Environment Word final N 9 81 0 0 90 58 % 10 90 0 0 Syllable final N 19 33 14 0 66 42 % 29 50 21 0 Total N 28 114 14 0 156 % 18 73 9 0 Gender Male N 5 12 3 0 20 13 % 25 60 15 0 Female N 23 102 11 0 136 87 % 17 75 8 0 Total N 28 114 14 0 156 % 18 73 9 0 Total (all factors) N 28 114 14 0 156 % 18 73 9 0

Table 6.22 VARBRUL percentages for [r], [l], deletion and [] in the conversations of the seven speakers of the third generation of LPRS who were Spanish dominant

The descriptive statistics on Table 6.22 shows that the seven third generation speakers who were Spanish dominant used [l] more than [r] in their conversations following the same pattern found in the first generation, as shown already on Table 6.15. These speakers did not use retroflex [] in their conversations, retroflex [] was found in the reading tasks of the three different generations.

For the first generation, then, the use of the retroflex [] is not related to their exposure to

AE through their contacts in the community and in their families. But, for the second and

130

third generations, it seems that AE influence is behind their use of the retroflex. The

third generation speakers who had more exposure to Spanish in the community and who

used the language in their conversations showed the same pattern of use for the [r]

variables found in IPRS, that is, they prefer [l] over [r]. With the results for the third

generation, I conclude that as far as (r) is concerned, the community will maintain their

preference for [l] in conversations given the results from the speakers of the third

generation who are still using LPRS and who still prefer to use [l], just as the first generation of Puerto Ricans who first arrived to Lorain.

6.3.2 Introduction to variable (b)

For this analysis, the variable (b) consists of three variants [b], [β] and [v]. The first two are the Standard PRS /b/ allophones [b] and [β]. As a bilabial stop, /b/ occurs word

initially or after a consonant /m/, /n/ or /l/. The bilabial fricative occurs intervocalically.

In this analysis, [b] and [β] are treated as a single variant. Another variant that is found in

the Spanish phonology of the Puerto Rican community in Lorain was the labiodental [v].

This variant is used in LPRS, as well as in varieties of Central and South American

Spanish, especially in reading. The following are examples:

131

SPRS LPRS Spanish Gloss "lal] volar "to flyا.lal] [voا.bo] "sio] vacío "emptyا.sio] [vaا.ba] "ban.ko] banco "bankا] [ban.koا] "ni.to] bonito "beautifulا.ni.to] [boا.bo] "mien.to] movimiento "movementا.mien.to] [mo.viا.mo.βi] do.ra] lavadora "washingا.do.ra] [la.vaا.la.βa] machine" "βe.ha] abeja "beeا.βe.ha] [aا.a] "ho] trabajo "workا.ho] [tra.βaا.tra.βa] "er.no] invierno "winterا.er.no] [in.viا.im.bi] "vi.dia] envidia "envyا.bi.dia] [emا.em] "en.te] ambiente "ambienceا.en.te] [am.biا.am.bi] "om.bre] hombre "manا] [om.breا]

Paraguayan Spanish has a [v] articulation in all positions, even word-initially (Lipski

1990); Cassano (1972) attributes this to influence from Guaraní36 although without providing much further evidence.

Espinosa (1930) describes the Spanish of New Mexico as not having a [v] sound. In cases where English words are borrowed into Spanish and these are written with , the letter is pronounced not as a labiodental fricative, but as labial stop or as a labial fricative depending on their surrounding phonological environment.37 More recently,

Torres-Cacoullos & Ferreira (2000) studied the relation between lexical frequency and the production of the voiced labiodental in New Mexican Spanish. In their study, these researchers found that there was a distinction between what they called “archaic” [v] and

36 Contrary to other Spanish dialects, the Spanish of Paraguay realizes /b/ as a [B] bilabial fricative instead of a [b] bilabial stop in initial position under the influence of Guaraní. (Malberg 1947-1948 cited in Cassano 1972)

37 “La labiodental fricativa v es desconocida en nmej., tanto que en los anglicismos ese sonido está remplazado por B o b, según los casos.” (Espinosa 1930: 135)

132

language-contact [v]. When a labiodental was found in high frequency words, these

words were retaining an old language feature carried on from Old Spanish transferred to

New Mexico and maintained in Colonial Mexican Spanish (e.g. recibir "to receive",

caballo "horse", nube "cloud"). On the other hand, there were words that follow the

English patterns of distribution of [b] and [v], particularly English cognates, and these were low in frequency (e.g. valor "value").

6.3.2.1 Results of the VARBRUL analyses regarding variable (b) in IPRS, LPRS and

MAS

The realization of (b) depends on its phonetic environment; thus, I tested for word initial

position, intervocalic position and occurrence after a consonant (/m/, /n/ or /l/) as the

phonetic environments where (b) could occur as [b], [B] or [v]. Age was considered for

the IPRS and MAS speakers. LPRS speakers were divided in three different generations

according to the criteria discussed in Chapter 5. Tables 6.23 – 6.32 present the results of

the VARBRUL analyses drawn from three separate runs.

In Tables 6.23 - 6.24, I present the findings for the distribution of all the possible variants

of (b) in IPRS. In the data, I found that there were 106 instances or 10% of a labiodental

fricative [v] among the Puerto Rican Islanders. The descriptive statistics show that

Standard (b/B) are the preferred variables in rural IPRS.

The Standard variants of (b) occurred in word initial (92%, N=513), intervocalic (87%,

N=369) or after a consonant (90%, N=83) as their percentages show. The variants of (b)

133 were preferred over labiodental [v] in reading (86%, N=566) and conversation (97%,

N=399). IPRS males (92%, N=502) and females (88%, N=463) preferred [b]/[B] over labiodental [v]. Furthermore, [b] and [B] were used by older (92%, N=489) and younger

(88%, N=476 IPRS speakers.

134

Dialect Island Puerto Rican Spanish Variable [b]/[B] [v] Total % Factor groups Phonetic Environment Word Initial N 513 43 556 52 % 92 8 Intervocalic N 369 54 423 39 % 87 13 After a Consonant N 83 9 92 9 % 90 10 Total N 965 106 1071 % 90 10 Style Reading N 566 93 659 62 % 86 14 Talking N 399 13 % 97 3 Total N 965 106 1071 % 90 10 Gender Male N 502 44 546 51 % 92 8 Female N 463 62 % 88 12 Total N 965 106 1071 % 90 10 Age Older N 489 41 530 49 % 92 8 Younger N 476 65 541 51 % 88 12 Total N 965 106 1071 % 90 10 Total (all factors) N 965 106 1071 % 90 10

Table 6.23 VARBRUL percentages for (b) in IPRS

135

Factor Group Factor Weight App/Total Input & Weight Phonetic Word initial 0.540 0.92 0.93 Environment Intervocalic 0.438 0.87 0.90 After a Consonant 0.544 0.90 0.93 Style Reading 0.358 0.86 0.87 Talking 0.717 0.97 0.97 Gender Male 0.537 0.92 0.93 Female 0.462 0.88 0.91 Age Older 0.547 0.92 0.93 Younger 0.454 0.88 0.91 Input: 0.922

Table 6.24 VARBRUL weights, application values and input and weights for [b]/[B]

Island Puerto Rican Spanish

Tables 6.25 - 6.26 show the distribution of the variants of (b) in LPRS. In the data, I found that, out of 2368 (b) tokens, there were 454 instances or 19% of labiodental [v] in

LPRS. [b]/[B] were found more often at the beginning of a word (83%, N=1028), or intervocalically (81%, N=734), than after a consonant (68%, N=152), where labiodental

[v] was also found (32%, N=73). The labiodental [v] was also found at the beginning of a word (17%, N=210) and intervocalically (19%, N=171).

The labiodental [v] and [b]/[B] were found when LPRS speakers were reading (26%,

N=404; 74%, N=1170, respectively). Labiodental [v] was not found as often in conversations (6%, N=50), but [b]/[B] occurred more frequently (94%, N=744).

136

Dialect Lorain Puerto Rican Spanish Variable [b]/[B] [v] Total % Factor groups Phonetic Environment Word Initial N 1028 210 1238 52 % 83 17 Intervocalic N 734 171 905 38 % 81 19 After a Consonant N 152 73 225 10 % 68 32 Total N 1914 454 2368 % 81 19 Style Reading N 1170 404 1574 66 % 74 26 Talking N 744 50 794 34 % 94 6 Total N 1914 454 2368 % 81 19 Gender Male N 964 184 1148 48 % 84 16 Female N 950 270 1220 52 % 78 22 Total N 1914 454 2368 % 81 19 Generation First N 991 97 1088 46 % 91 9 Second N 699 244 943 40 % 74 26 Third N 224 113 337 14 % 66 34 Total N 1914 454 2368 % 81 19 Total (all factors) N 1914 454 2368 % 81 19

Table 6.25 VARBRUL percentages for (b/B) in LPRS

137

Males and females preferred the standard variants, with the males (84%, N=964) having a

slightly higher frequency than females (78%, N=950), who had 270 cases of [v], or 22%.

Perhaps the most interesting results come from the generation group, which shows a

decline in [b]/[B] use across the three generations, with the first generation using [b]/[B]

in 991 cases or 91%; the second generation in 699 cases, or 74%; and the third generation

using [b]/[B] in 224 cases, or 66%. The third generation uses a labiodental [v] more

(34%, N=113) than the first (9%, N=97) or second generation (26%, N=244).

Factor Group Factor Weight App/Total Input & Weight Phonetic Word Initial 0.503 0.83 0.85 Environment Intervocalic 0.527 0.81 0.87 After a Consonant 0.378 0.68 0.78 Style Reading 0.374 0.74 0.78 Talking 0.736 0.94 0.94 Gender Male 0.527 0.84 0.87 Female 0.475 0.78 0.84 Generation First 0.670 0.91 0.92 Second 0.371 0.74 0.77 Third 0.307 0.66 0.72 Input: 0.853

Table 6.26 VARBRUL weights, application values and input and weights for (b) in

Lorain Puerto Rican Spanish

138

The distribution of the (b) variants in MAS is presented in Tables 6.27 - 6.28. Of a total of 1328 tokens, MAS in Lorain used [b]/[B] more often than labiodental [v], with 76% vs.

24%, respectively. The Mexicans in Lorain preferred the standard variants in word initial

(81%, N=638), and intervocalic positions (73%, N=329), but were evenly divided when

[b]/[B] was after a consonant.

139

Dialect Mexican American Spanish Variable [b]/[B] [v] Total % Factor groups Phonetic Environment Word Initial N 638 145 783 59 % 81 19 Intervocalic N 329 122 451 34 % 73 27 After a Consonant N 47 47 94 7 % 50 50 Total N 1014 314 1328 % 76 24 Style Reading N 410 293 703 53 % 58 42 Talking N 604 21 625 47 % 97 3 Total N 1014 314 1328 % 76 24 Gender Male N 500 50 550 41 % 91 9 Female N 514 264 778 59 % 66 34 Total N 1014 314 1328 % 76 24 Age Older N 580 165 745 56 % 78 22 Younger N 434 149 583 44 % 74 26 Total N 1014 314 1328 % 76 24 Total (all factors) N 1014 314 1328 % 76 24

Table 6.27 VARBRUL percentages for (b/B) in MAS

140

The standard variants were also used more in conversation (97%, N=604), than when reading, where [b]/[B] and labiodental [v] had a slight difference in frequency (58%,

N=410; 42%, N=293, respectively).

Mexican males preferred to use [b]/[B] more (91%, N=500) than the Mexican women

(66%, N=514), who also used labiodental [v] (34%, N=264).

When considering age, there was not a great difference between the older and the

younger group, since both used [b]/[B] in 78% and 74%, respectively, while using a

labiodental [v], the older group less (22%, N=165) than the younger group (26%,

N=149).

141

Factor Group Factor Weight App/Total Input & Weight Phonetic Word Initial 0.558 0.81 0.91 Environment Intervocalic 0.461 0.73 0.88 After a Consonant 0.234 0.50 0.72 Style Reading 0.134 0.58 0.57 Talking 0.891 0.97 0.99 Gender Male 0.854 0.91 0.98 Female 0.223 0.66 0.71 Age Older 0.529 0.78 0.90 Younger 0.463 0.74 0.88 Input: 0.895

Table 6.28 VARBRUL weights, application values and input and weights for [b/B] in

Mexican American Spanish in Lorain

Table 6.29 presents a summary comparing overall percentages, input probability, and input and weight values for [b]/[B] in IPRS, LPRS and MAS in Lorain obtained in three separate runs. The weights show that [b]/[B] is most often preferred in IPRS than in

MAS in Lorain and that LPRS patterns more closely to IPRS.

There is no clear similarity in usage of [b]/[B] by phonetic environment between IPRS,

LPRS and MAS. Each ethnic group has a different weight value for each phonetic

environment considered in the VARBRUL analyses.

142

[b]/[B] were found more frequently in conversations, while less in reading. The MAS speakers in Lorain had a particular lower probability (0.133) for [b]/[B] in reading than all the other ethnic groups. IPRS and LPRS are similar in this speech style when it comes to the use of [b]/[B] with probability weights of 0.349 and 0.374, respectively.

Community IPRS LPRS MAS Total % 90 81 76 Input Probability 0.922 0.853 0.895 Factors Phonetic Environment * Word Initial 0.540 0.503 0.558 Intervocalic 0.438 0.527 0.461 After a Consonant 0.544 0.378 0.234 Style Reading 0.349 0.374 0.133 Talking 0.730 0.736 0.891 Gender * ** Male 0.537 0.57 0.852 Female 0.462 0.475 0.225 Generation First 0.670 Second 0.371 Third 0.307 Age * *** Older 0.547 0.529 Younger 0.454 0.463

* = Not chosen as significant in IPRS by VARBRUL ** = Not chosen as significant in LPRS by VARBRUL *** = Not chosen as significant in MAS by VARBRUL

Table 6.29 Summary of percentages and comparison of input probabilities and weight values of [b]/[B] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and

Mexican American Spanish

143

Gender was selected as significant only for MAS in Lorain. It is interesting to mention that females in all three ethnic groups have a lower probability for [b]/[B], than males.

Mexican females, however, showed the lowest probability of all the females (0.225).

Although age was not chosen as a significant factor group for MAS and IPRS, it is relevant to look at the probability weights for IPRS and MAS. As opposed to the older generations of IPRS and MAS speakers, young IPRS and MAS speakers have a lower probability for [b]/[B] with 0.454 and 0.463, respectively.

Generation as a factor group was chosen as significant by VARBRUL. LPRS shows a decline in the use of [b]/[B] across its three generations, with the older speakers using

[b]/[B] with higher probability (0.670), as opposed to the second (0.371) and third (0.307) generations. Then, the third generation of LPRS speakers are converging in their use of

[b]/[B] with the younger speakers of IPRS and MAS, while differing from the first generation of LPRS speakers as well as from the older generation of speakers in Puerto

Rico.

The variant [v], because of its low occurrence in the data collected in Puerto Rico, is not part of the IPRS phonology, although its speakers are able to produce it. Thus, I propose that [v] did not exist in the Spanish spoken by the first generation of Puerto Ricans who first arrived in Lorain, but acquired it later on in Lorain.

144

It is therefore relevant at this point to look at the production of [v] particularly among the

Lorain Puerto Ricans. Although [v] is not the preferred variant among Lorain Puerto

Ricans and Mexicans, it is nonetheless the variant found in the reading tasks of the two

ethnic groups that are in contact with AE. From these results, it follows, then, that

schooling in AE plays a role when Spanish and AE bilinguals read. AE use serves as an

additional reinforcement. Both Mexicans and Puerto Ricans in Lorain identify

orthographic with AE orthography and phonology, producing instead of a (b/B), a

[v].

Another factor that plays a role in choosing [v] over (b/B) in reading tasks is the degree of formality that this speech style triggers. Speakers tend to hypercorrect when reading in

Spanish using AE [v] for orthographic , instead of the corresponding Spanish allophones.

Table 6.30 presents a summary of the percentages and a comparison of the probabilities

of [v] for IPRS, LPRS and MAS.

145

Community IPRS LPRS MAS Total % 10 19 24 Input Probability 0.097 0.192 0.236 Factors Phonetic Environment * Word Initial 0.460 0.497 0.442 Intervocalic 0.562 0.473 0.539 After a Consonant 0.456 0.622 0.766 Style Reading 0.642 0.627 0.866 Talking 0.283 0.264 0.109 Gender * ** Male 0.463 0.473 0.146 Female 0.538 0.525 0.777 Generation First 0.330 Second 0.629 Third 0.693 Age * *** Older 0.453 0.471 Younger 0.546 0.537

* = Not chosen as significant in IPRS by VARBRUL ** = Not chosen as significant in LPRS by VARBRUL *** = Not chosen as significant in MAS by VARBRUL

Table 6.30 Summary of percentages and comparison of input probabilities and weight values of [v] for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican

American Spanish

The input probabilities on Table 6.30 show that MAS speakers (0.236) made a more frequent use of [v] than the LPRS (0.192) and IPRS (0.097) speakers. In fact, LPRS stands in the middle of the two communities as far as input probabilities for [v] are concerned. The two communities with the highest exposure to AE used [v] more frequently than the community with less exposure to this language.

146

A separate run was done for the ten speakers of the first generation who have more exposure to AE to determine if AE had any effect. Table 6.31 presents the descriptive statistics.

Dialect First Generation Lorain Puerto Rican Spanish Variable [b]/[B] [v] Total % Factor groups Phonetic Environment Word Initial N 282 25 307 56 % 92 8 Intervocalic N 182 22 204 37 % 89 11 After a Consonant N 36 6 42 8 % 86 14 Total N 500 53 553 % 90 10 Style Reading N 299 52 351 63 % 85 15 Talking N 201 1 202 37 % 100 0 Total N 500 53 553 % 90 10 Gender Male N 286 3 289 48 % 99 1 Female N 214 50 264 48 % 81 19 Total N 500 53 553 % 90 10 Total (all factors) N 500 53 553 % 90 10

Table 6.31 VARBRUL percentages for [b] and [v] in the ten speakers of the first generation of LPRS with exposure to AE

147

The descriptive statistics for the first generation speakers with more exposure to AE reveal that these individuals are not necessarily the only ones who are producing [v] in their generation. Of all the 97 cases of [v] in the first generation (see Table 6.25), 53 were found among those who said they used AE with relatives and friends.

Dialect Third Generation Lorain Puerto Rican Spanish Variable [b]/[B] [v] Total % Factor groups Phonetic Environment Word Initial N 53 12 65 70 % 82 18 Intervocalic N 182 22 23 25 % 91 9 After a Consonant N 4 1 5 5 % 80 20 Total N 78 15 93 % 84 16 Gender Male N 19 4 23 25 % 83 17 Female N 59 11 70 75 % 84 16 Total N 78 15 93 % 84 16 Total (all factors) N 78 15 93 % 84 16

Table 6.32 VARBRUL percentages for [b]/[B] and [v] in the conversations of the seven speakers of the third generation of LPRS who were Spanish dominant

148

The results presented in Table 6.32 show that the seven third generation speakers who

use Spanish more used [v] in their conversations even more than first generation

speakers.

For the first generation, then, the use of the labiodental [v] is not related to their exposure

to AE. Third generation speakers who use Spanish in the community and who use this

language in their conversations showed the same pattern of use of [v] as the first generation.

As far as the (b) variants are concerned, the community will maintain [b] and [B] in conversations given the results from the speakers of the third generation.

6.3.3. Introduction to Variable (dΩ)38

In LPRS, the (dΩ) variants are the affricate [dΩ], the semiconsonant [j] and the vowel [i].

The variable (dΩ) represents the initial and medial consonants in words like ya "already", lleno "full", and mayo "May", allá "there", respectively in IPRS.

38 Before starting the section on variable (dΩ), I feel it is important to clarify the symbols used in this section. In the Spanish linguistics literature, /y/ is used for IPA’s /J/, or voiced palatal fricative; however, for this study, I have used /J/. It is important to note at this point that for the voiced palatal fricative, Díaz- Campos and Morgan (2002) have identified five different symbols used in the literature on Spanish. On the other hand, to refer to the IPRS palatal nonsibilant affricate variant, there are at least six different symbols (Díaz-Campos and Morgan: 2002) that have been used by different linguists. I, however, prefer to use IPA /dΩ/ for the sake of consistency and because it is the best symbol that captures the articulatory nature of the affricate’s production, i.e., closure followed by friction.

149

The affricate (dΩ) in IPRS is a result of internal developments in the language. In Old

Spanish, there were two sounds, a voiced lateral approximant /Ò/ and a voiced palatal

fricative /J/ (Penny, 1991). These sounds are preserved in very few Spanish dialects,

known as lleísta dialects. These Spanish varieties use for the initial sound in words like

ya "already" a voiced palatal fricative /J/, and for the medial consonant in words like allá

"there" a voiced lateral approximant /Ò/. However, PRS and some dialects of Mexican

Spanish have lost this distinction in favor of /J/, which later becamse [dΩ] because it is no longer productive, as it was only used to distinguish few words such as maya "Mayan"

and malla "net". These varieties of Spanish are described as yeísta39 dialects.

Navarro-Tomás (1948) and Álvarez-Nazario (1990) described earlier IPRS as having a

voiced palatal fricative, produced with a narrow articulation with a soft and weak

frication. Álvarez-Nazario (1990) points out that as early as 1521, IPRS already

presented examples of being a yeísta dialect. According to Álvarez-Nazario (1991,

1990), Navarro-Tomás (1948), Saciuk (1980), IPRS completely lost the distinction of /J/ and /Ò/ at the end of the 17th century.

39 Carreter (1990: 415) defines yeísmo as “Fenómeno que consiste en pronunciar la ll como y, por tendencia a hacer central una articulación lateral. Son yeístas Madrid, Toledo, Ciudad Real, Extremadura, Andalucía y Canarias. En Hispanoamérica hay distinción entre ll y y en varias provincias de Argentina, Chile, Perú, Colombia y Ecuador."

150

Navarro-Tomás (1948) and Álvarez-Nazario (1990) claim that both orthographic

and orthographic are pronounced as [dΩ] (which they represent as "/y/"). They

observed that speakers tended to narrow the opening so to produce an affricate. This so-

called “strengthening” is found, according to the researchers, in stressed initial position.

Navarro-Tomás (1948) and Álvarez-Nazario (1990) also point out some cases of

vocalization and deletion particularly between the vowels /u/ and /i/, e.g. sambuír for

zambullir ("to dive"). This sound is not “weakened” in words like palmillo "little palm",

patilla "sideburn", peinilla "comb". Navarro-Tomás (1948) also noted cases of

hypercorrection (e.g. lucilla for lucía "to show").

Álvarez-Nazario (1990) argues that the loss of the voiced palatal fricative [J] in the rural

areas of Puerto Rico is due to the increased contact with the urban communities of IPRS speakers, education, modernization (e.g., electricity and as consequence, access to radio stations outside of the rural areas, access to TV, as well as the construction of new roads that accelerated the access to and from the metropolitan areas) and the loss of the social scenarios where contacts in the rural areas were fostered (e.g., ventorrillos, bars and local

convenience stores).

151

Like IPRS, most varieties of Mexican Spanish are "yeísta" dialects. Lipski (1994) describes the Spanish of Mexico as being "yeísta". Boyd-Bowman (1952) argues that

there is no convincing evidence for other allophones to have existed in Mexico for more

than two centuries.

152

Unlike IPRS and most varieties of Mexican Spanish, MAS dialects are not “yeístas”

dialects. According to Espinosa (1930) and Ross (1980), the MAS dialects of New

Mexico and Colorado Spanish have the sound /J/ where IPRS and Mexican Spanish use

/dΩ/. Espinosa (1930) describes this /J/40 sound as “similar” to the Spanish, but more

open, lax and weaker, resembling more the English semiconsonant. When trying to

produce the Spanish word /maJo/ "May", Espinosa (1930) explains, speakers pronounce

instead [majo]41. The realization is common when /J/ is next to /e/ or /i/ or when in

intervocalic position. Word-initially /J/ is rarely altered, except in cases of fast speech

(e.g., [la ierba] [la jerba], la hierba "the grass"). Ross (1980) also adds that /J/ vocalization is found as an alternative to /J/ in New Mexico and Colorado Spanish.

As in PRS, [J] has become [dΩ] in MS. The variants of (dΩ) in PRS are illustrated in cases such as:

LPRS LPRS SPRS Gloss "dΩe.no] "fullا] [ie.noا] [je.noا] [jo] [io] [dΩo] "I" "dΩe.gua] "mareا] [ie.guaا] [je.guaا] "(.ke.dΩa] "that one over there (fem.singا.ke.ia] [aا.ke.ja] [aا.a]

40 In the cases below, whenever the description for /y/ was not provided by the linguists cited, I maintained the use of the phonetic symbols provided by them.

41 “Esta fricativa palatal sonora es un sonido semejante al del español, pero un poco más abierto, débil y relajado, aproximándose más que la y española a la semiconsonante y del inglés you. El nuevomejicano necesita un esfuerzo especial para decir ma-yo con la y a la castellana; dice algo como májo. Es el sonido nuevomejicano de la y inicial de sílaba o de la ll que equivale a y en las mismas condiciones.» (Espinosa 1930: 141)

153

The realization of /J/ as [j] or [i] is also found in the Central American Spanish of Costa

Rica, Honduras (Lipski 1987), Guatemala (Henríquez-Ureña 1938), Nicaragua (Lacayo

1954), El Salvador (Canfield 1960), parts of Mexico (Henríquez-Ureña 1938), Yucatán

(Suárez 1945) and in some South American Spanish-speaking countries (Canfield 1981,

Lipski 1994), parts of Argentina (Vidal de Battini 1949), the south of Chile (Oroz 1966)

and Ecuador (Toscano Mateus 1953), for instance.

In general, the change of /J/ to [j] is a phenomenon found in Central American Spanish

and other Spanish-speaking countries. Island Puerto Rican Spanish, however, has /dΩ/

with an affricate pronunciation in onset position and there is no “weakening”

intervocalically as in the other Spanish dialects already mentioned or “vocalization” of /J/ to [j] or [i] (Lipski 1994, Saciuk 1980). In his 1952 study of LPRS, Decker (1952) did not find /dΩ/ realized as [j], but found it to be realized either as an affricate or as aspiration, or even as a latero-palatal. Neither of the latter two occurred in my data.

6.3.3.1 Results of the VARBRUL analyses regarding variable (dΩ) in IPRS, LPRS and

MAS

Because the realization of (dΩ) as [dΩ], [j] or [i] depends on its phonetic environment, I tested for word initial position, occurrence after a front vowel and after a back vowel as possible phonetic environments. The social factors were the same as for the variables discussed above. Tables 6.33 – 6.41 present the results of the VARBRUL analyses drawn from three separate runs.

154

I present the descriptive statistics for [dΩ] in IPRS in Table 6.33. In Table 6.34, I present

the VARBRUL weights, application totals, input and weights for [dΩ] in IPRS. In IPRS,

the preferred variant of /dΩ/ was [dΩ]. Regardless of the phonetic environment, beginning of a word, after a front or a back vowel, IPRS speakers used [dΩ].

The VARBRUL percentages in Table 6.33 show that the use of [dΩ] among young and old female and male Puerto Ricans was near-categorical in all speech styles.

155

Dialect Island Puerto Rican Spanish Variable [dΩ] [j] [i] Total % Factor groups Phonetic Environment Word Initial N 946 10 1 957 49 % 99 1 0 After a Front Vowel N 396 8 0 404 21 % 98 2 0 After a Back Vowel N 575 12 2 589 30 % 98 2 0 Total N 1917 30 3 1950 % 98 2 0 Style Reading N 1407 29 3 1439 74 % 98 2 0 Talking N 510 1 0 511 26 % 100 0 0 Total N 1917 30 3 1950 % 98 2 0 Gender Male N 879 24 3 906 46 % 97 3 0 Female N 1038 6 0 1044 54 % 99 1 0 Total N 1917 30 3 1950 % 98 2 0 Age Older N 970 13 1 984 50 % 99 1 0 Younger N 947 17 2 966 50 % 98 2 0 Total N 1917 30 3 1950 % 98 2 0 Total (all factors) N 1917 30 3 1950 % 98 2 0

Table 6.33 VARBRUL percentages for (dΩ) in IPRS

156

Factor Group Factor Weight App/Total Input & Weight Phonetic Word Initial 0.582 0.99 0.99 Environment After a Front Vowel 0.409 0.98 0.99 After a Back Vowel 0.429 0.98 0.99 Style Reading 0.355 0.98 0.98 Talking 0.843 1.00 1.00 Gender Male 0.300 0.97 0.98 Female 0.676 0.99 1.00 Age Older 0.536 0.99 0.99 Younger 0.464 0.98 0.99 Input: 0.992

Table 6.34 VARBRUL weights, application values and input and weights for [dΩ] in

Island Puerto Rican Spanish

The descriptive statistics for (dΩ) in LPRS are presented in Table 6.35. Table 6.36

presents the VARBRUL weights, application totals, input and weights for [dΩ] in LPRS.

The preferred variant of /dΩ/ was [dΩ]. In LPRS, [dΩ] is found at the beginning of a

word (92%, N=1877), after a front vowel 83%, N=680) or a back vowel (83%, N=917).

The same preference that the IPRS speakers showed for the usage of [dΩ] is maintained among LPRS speakers when speaking (93%, or N=980) and reading (86%, or N=2994), with a very slight decrease in frequency because LPRS also make use of [j] when reading

(9%, N=254) and talking (4%, N=39).

157

LPRS has a similar pattern of IPRS [dΩ] usage when gender is considered, with LPRS

males (90%, N=1659) and females (86%, N=1815) preferring [dΩ]. LPRS males and

females also used [j] (6%, N=110; 9%, N=183, respectively) and [i] (4%, N=74, 6%,

N=121, respectively).

First and second generation LPRS speakers maintain the pattern of [dΩ] usage found in the IPRS brought from Puerto Rico to Lorain by the first generation, i.e., the first generation (95%, N=1846) and the second generation (84%, N=1414) still use [dΩ] and not [j]. Speakers of the third generation, however, seem to be changing their preference to that of [j] (20%, N=65) and [i] (15%, N=51), while still maintaining [dΩ].

158

Dialect Lorain Puerto Rican Spanish Variable [dΩ] [j] [i] Total % Factor groups Phonetic Environment Word Initial N 1877 93 76 2046 52 % 92 5 4 After a Front Vowel N 680 86 51 817 21 % 83 11 6 After a Back Vowel N 917 114 68 1099 28 % 83 10 6 Total N 3474 293 195 3962 % 88 7 5 Style Reading N 2494 254 165 2913 74 % 86 9 6 Talking N 980 39 30 1049 26 % 93 4 3 Total N 3474 293 195 3962 % 88 7 5 Gender Male N 1659 110 74 1843 47 % 90 6 4 Female N 1815 183 121 2119 53 % 86 9 6 Total N 3474 293 195 3962 % 88 7 5 Generation First N 1846 71 25 1942 49 % 95 4 1 Second N 1414 157 119 % 84 9 7 Third N 214 65 51 330 8 % 65 20 15 Total N 3474 293 195 3962 % 88 7 5 Total (all factors) N 3474 293 195 3962 % 88 7 5

Table 6.35 VARBRUL percentages for (dΩ) in LPRS

159

Factor Group Factor Weight App/Total Input & Weight Phonetic Word Initial 0.588 0.92 0.94 Environment After a Front Vowel 0.389 0.83 0.88 After a Back Vowel 0.419 0.83 0.89 Style Reading 0.399 0.86 0.88 Talking 0.757 0.93 0.97 Gender Male 0.532 0.90 0.93 Female 0.473 0.86 0.91 Generation First 0.696 0.95 0.96 Second 0.371 0.84 0.87 Third 0.103 0.65 0.56 Input: 0.917

Table 6.36 VARBRUL weights, application values and input and weights for [dΩ] in

Lorain Puerto Rican Spanish

Table 6.37 presents the descriptive statistics for (dΩ) in Lorain MAS. The VARBRUL

weights, application totals, input and weights for [dΩ] in MAS are shown in Table 6.38.

Lorain MAS prefers [j] (46%, N=620) over [dΩ] (33%, N=449) and even produces [i]

(21%, N=281). In all phonetic environments, except after a back vowel (36%, N=140),

[j] is used more by the Lorain Mexicans.

MAS speakers make a frequent use of [dΩ] when reading (44%, N=338), but not when

talking (19%, or N=111), using instead either [j] or [i] when talking (77%, N=454; 4%,

N=22, respectively). [j] was also used when reading (22%, or N=166) as well as [i]

(34%, N=259).

160

Moreover, MAS males and females do not follow the same pattern of the LPRS speakers.

They prefer [j], males make use of [j] (30%, N=115) or [i] (64%, N=245) and the females used [j] in 52% of the cases, with [dΩ] used by them 44%, with 4% of [i].

The younger MAS speakers preferred [j] (43%, or N=475), over [dΩ] (31%, or N=344).

They also used [i] (26%, N=281). Although, the older MAS speakers still prefer [j]

(58%, or N=145) over [dΩ] (42%, or N=105), the difference is very slight. The older

MAS speakers did not use the variant [i], which is found among the young MAS speakers, although both groups still diverge from the LPRS, the largest group of Spanish speakers in the community.

161

Dialect Mexican American Spanish Variable [dΩ] [j] [i] Total % Factor groups Phonetic Environment Word Initial N 230 354 128 712 53 % 32 50 18 After a Front Vowel N 66 126 56 248 18 % 27 51 23 After a Back Vowel N 153 140 97 390 29 % 39 36 25 Total N 449 620 281 1350 % 33 46 21 Style Reading N 338 166 259 763 57 % 44 22 34 Talking N 111 454 22 587 43 % 19 77 4 Total N 449 620 281 1350 % 33 46 21 Gender Male N 20 115 245 380 28 % 5 30 64 Female N 429 505 36 970 72 % 44 52 4 Total N 449 620 281 1350 % 33 46 21 Age Older N 105 145 0 250 19 % 42 58 0 Younger N 344 475 281 1100 81 % 31 43 26 Total N 449 620 281 1350 % 33 46 21 Total (all factors) N 449 620 281 1350 % 33 46 21

Table 6.37 VARBRUL percentages for (dΩ) in MAS

162

Factor Group Factor Weight App/Total Input & Weight Phonetic Word Initial 0.486 0.32 0.18 Environment After a Front Vowel 0.501 0.27 0.19 After a Back Vowel 0.524 0.39 0.20 Style Reading 0.880 0.44 0.63 Talking 0.070 0.19 0.02 Gender Male 0.065 0.05 0.02 Female 0.740 0.44 0.40 Age Older 0.936 0.42 0.77 Younger 0.352 0.31 0.11 Input: 0.188

Table 6.38 VARBRUL weights, application values and input and weights for [dΩ] in

Mexican American Spanish in Lorain

In Table 6.39, I present a summary comparing overall percentages, input probability, input and weight values for [dΩ] in IPRS, LPRS and MAS in Lorain obtained in three separate runs. There is a considerable difference between the two dialects of Puerto

Rican Spanish and MAS, with the Mexicans in Lorain having an input probability of

0.188 for [dΩ].

163

Community IPRS LPRS MAS Total % 98 88 33 Input Probability 0.992 0.917 0.188 Factors Phonetic Environment * *** Word Initial 0.582 0.588 0.486 After a Front Vowel 0.409 0.389 0.501 After a Back Vowel 0.429 0.419 0.524 Style Reading 0.353 0.399 0.881 Talking 0.846 0.757 0.069 Gender Male 0.301 0.532 0.066 Female 0.675 0.473 0.739 Generation First 0.696 Second 0.371 Third 0.103 Age * Older 0.536 0.937 Younger 0.464 0.351

* = Not chosen as significant in IPRS by VARBRUL ** = Not chosen as significant in LPRS by VARBRUL *** = Not chosen as significant in MAS by VARBRUL

Table 6.39 Summary of percentages and comparison of input probabilities and weight values of (dΩ) for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and

Mexican American Spanish

164

Phonetic environment was not selected as significant for IPRS and MAS. In LPRS, [dΩ] is found in initial position (0.588), after a front vowel (0.389) and after a back vowel

(0.419). The results show that IPRS speakers prefer to use [dΩ] in all phonetic environments and all speech styles, demonstrating that IPRS still maintains today the same variant [dΩ] that Álvarez-Nazario (1991, 1990) and Navarro-Tomás (1948) once described in the earlier studies of IPRS; [dΩ] is not “weakened” intervocalically and it is not vocalized.

For [dΩ] in MAS, is shown to be used in all phonetic environments. In MAS, [dΩ] is most frequent in reading (0.881) and less when speaking (0.069), while IPRS and LPRS show similar weights for reading (0.353 and 0.399, respectively) and speaking (0.846 and

0.757, respectively).

Female Puerto Ricans used [dΩ] (0.675) more than the Puerto Rican males (0.301). A different pattern in the production of [dΩ] by gender in LPRS was found with [dΩ] used by Lorain Puerto Rican females less (0.473) than the males (0.532). MAS in Lorain, when gender is considered, seems to be similar to IPRS, with the males using [dΩ] less

(0.066) than the females (0.739).

Old and young Puerto Ricans used [dΩ] as the preferred variant with input probability weights of 0.536 and 0.464, respectively. However, older MAS speakers used [dΩ] more

(0.937) than the younger (0.351) MAS speakers in Lorain. Both the younger and the

165

older generation of MAS are using the variants [j] and [i] more, instead of

accommodating to LPRS [dΩ].

In LPRS, a decline in the use of [dΩ] is observed across three generations. The first

generation has a higher probability (0.696), followed by the second generation (0.371) and the third generation (0.103) speakers. Third generation speakers tend to favor [j],

while first and second generation speakers favor [dΩ], or the variable preferred in IPRS

as reported above. Also, third generation speakers are showing change in their usage of

[dΩ]. They use [j] (20%) and [i] (15%) as an additional variable indicating their

accommodation to MAS.

The results presented above show that the old MAS speakers have a very close frequency

of usage of both [dΩ] and [j], but [j] is still the preferred variant, as opposed to the Puerto

Ricans. Younger MAS speakers are using [j] and [i]. The younger speakers of MAS

have a more marked difference in preference as demonstrated by the results. Their usage

of [j] is preferred over the IPRS variant and, thus, MAS in Lorain is maintaining the

pattern of usage found in MAS and other Spanish dialects, while diverging from the

Spanish dialect with which it is in continuous contact in Lorain.

166

Although Decker (1952) did not find the production of /dΩ/ as [j] or [i] in LPRS,

currently, third generation LPRS speakers are also using [j] or [i], while first and second

generations are still maintaining the same [dΩ] that was transmitted to them from IPRS.

Young speakers of MAS in Lorain have both [j] and [i] as the preferred variants. The third generation of LPRS speakers are also using [j] and [i] just like the MAS youngest generation. In the use of the variants for (dΩ), the younger speakers of LPRS are in the middle, between the MAS older speakers and the IPRS speakers. Third generation LPRS and younger MAS speakers are converging in their use of [j] and [i], while both groups are diverging from [dΩ], the variable preferred by older IPRS and first and second generation LPRS speakers. MAS and AE are the sources for the use of [j] and [i] in

LPRS. Both reinforce the use of these (dΩ) alternatives in the PRS dialect in Lorain.

Table 6.40 presents a summary of percentages and a comparison of percentages of (j) for

IPRS, LPRS and MAS.

167

Community IPRS LPRS MAS Total % 2 7 46 Input Probability 0.015 0.074 0.459 Factors Phonetic Environment * *** Word Initial 0.418 0.403 0.528 After a Front Vowel 0.614 0.621 0.525 After a Back Vowel 0.556 0.590 0.433 Style Reading 0.641 0.590 0.253 Talking 0.163 0.268 0.803 Gender ** Male 0.687 0.469 0.508 Female 0.336 0.527 0.497 Generation First 0.349 Second 0.594 Third 0.848 Age * Older 0.469 0.155 Younger 0.531 0.595

* = Not chosen as significant in IPRS by VARBRUL ** = Not chosen as significant in LPRS by VARBRUL *** = Not chosen as significant in MAS by VARBRUL

Table 6.40 Summary of percentages and comparison of input probabilities and weight values of (j) for Island Puerto Rican Spanish, Lorain Puerto Rican Spanish and Mexican

American Spanish

168

The input probability for MAS (0.459) demonstrates that the use of [j] is considerable higher in this dialect than for LPRS (0.074) and IPRS (0.015). Although LPRS has a low input probability, it is nonetheless higher than IPRS and, thus, it stands in the middle of the other two community dialects.

Table 6.41 presents the descriptive statistics for the seven third generation speakers who are Spanish dominant, use Spanish in their conversations and are bilingual.

Dialect Third Generation Lorain Puerto Rican Spanish Variable [dΩ] [j] [i] Total % Factor groups Phonetic Environment Word Initial N 78 5 3 86 59 % 91 6 3 After a Front Vowel N 26 8 7 41 28 % 63 20 17 After a Back Vowel N 16 2 1 19 13 % 84 11 5 Total N 120 15 11 146 % 82 10 8 Gender Male N 24 0 0 24 16 % 100 0 0 Female N 96 15 11 122 84 % 79 12 9 Total N 120 15 11 146 % 82 10 8 Total (all factors) N 120 15 11 146 % 82 10 8

Table 6.41 VARBRUL percentages for (dΩ) in the conversations of the seven speakers of the third generation of LPRS who were Spanish dominant

169

The results from the VARBRUL run of the conversations of the third generation speakers

who were Spanish dominant reveal that these speakers do not use [j] or [i] more, but that

they prefer [dΩ], maintaining the use of the pattern brought from Puerto Rico. Third

generation non-Spanish dominant speakers are similar to MAS in their patterns for [j]

use. However, due to their limited Spanish proficiency, their behavior is not reliable and

the data does not give any evidence to the LPRS of the community. Considering the data

obtained from those speakers who are Spanish dominant in the third generation, I propose then that the use of [dΩ] will be maintained.

In the community of Puerto Ricans in Lorain, different patterns of phonological variation

and linguistic influence have been observed. AE has had an influence and many

members of the third generation prefer this language. However, those who chose to use

Spanish maintain for the most part the same patterns of variation observed in IPRS. AE

has influenced the Spanish of the second and third generations in the use of a labiodental

[v] as an alternative for [b]/[B] and a retroflex [] as an alternative for [r]. AE also has an

indirect influence on first generation speakers who are not proficient in AE through

contact with those speakers of the first generation who have more exposure to AE. Thus,

first generation speakers, just as second and third generation speakers in Lorain, use the

alternatives for (r) and (b).

170

As far as (e) and (o) raising is concerned, LPRS has maintained the same pattern of

variation found in IPRS, that is, (e) and (o) raising is almost nonexistent in IPRS and also

in LPRS.

In conclusion, LPRS has maintained most of the patterns of variation still found in IPRS.

Considering the data collected from the two communities, I conclude that LPRS is still a continuity of IPRS. Perhaps, in the future, because of the continuing influence of AE and

MAS, this pattern will change and LPRS will become a dialect different from IPRS, if it survives at all.

171

CHAPTER 7

OUTCOMES OF LANGUAGE AND DIALECT CONTACT: EXPLANATIONS.

At the earlier stages of this study, I ventured to argue that the variety of Spanish spoken by the Mexicans and Puerto Ricans was a new Spanish dialect and that we were witnessing the emergence of a new Spanish variety. However, before arguing that a new language or dialect is emerging, one first has to explore the changes taking place within the language, while considering who the agents of these changes are.

In Chapter 1, I formulated the question of whether or not a limited number of

phonological features are enough when determining the emergence of a new language

variety. The answer is most likely not. I would say rather that more aspects of the

language need to be explored when deciding if a new dialect is emerging. Different

domains of the language need to be considered before establishing that a given variety is

different from another. An analysis limited to phonology, syntax or morphology is not

enough. All these areas need to be considered while trying to decide if a new language

variety is emerging. Similarly, one domain is not enough to decide when there is a new

dialect emerging, but rather, other areas need to be considered. There may not be any

changes taking place in one area of a dialect (e.g. phonology), while there are changes

172

occurring in others (e.g. morphology, syntax). However, in the case of LPRS, I chose to

focus on only five phonological variables not deliberately, but rather because these were

the only ones that were salient to me as an IPRS speaker (see Chapter 1). No other areas

of the language, i.e., morphology and/or syntax, had any variation or change that was

salient to me.

Despite their limited occurrence in LPRS, (e) and (o) raising are, nevertheless, very

salient and noticeable to an IPRS speaker when they do occur. At this point, I can only

speculate about alternative explanations had the five variables I looked at for this project

turned out to be radically different from IPRS. Had this been the case, I would have

consider LPRS a different dialect from IPRS, just as Southern AE is considered different

from Northern AE because these two varieties have variation that does not impede

communication, but nonetheless is sufficient to make one noticeably different from the

other.

As far as LPRS is concerned, it should not be considered a different dialect, and therefore

not a "new" one. As has been shown in the previous chapter, Puerto Rican Spanish in

Lorain has maintained many phonological features transmitted from IPRS, while still having some variation. LPRS is a dialect that is simply a continuation of IPRS that

exhibits some variation in its phonology using more alternatives that are available for (r)

and (b) due to the influence of AE. There is no change in LPRS; rather, Spanish in

Lorain is reacting to the pressures of AE and the lack of an appropriate Spanish school

program for heritage learners. Furthermore, MAS in Lorain and LPRS are merging in

173

certain variables (e.g. (j)), but the mergers vary within each age and generation. On the

other hand, both groups are diverging in variables such as (r), presenting different

degrees of divergence depending on generation and age group.

Puerto Rican Spanish in Lorain is experiencing different pressures from MAS and AE.

As a result, different outcomes are observed as reactions to the different linguistic pressures within the community. At the outset, a reasonable, and eminently testable, hypothesis was that LPRS and MAS were converging in their phonology. This proposal has been tested and shown to be unsupportable as it has been observed that some of the phonological variables studied for this project have not converged with variables from

MAS. However, AE is exerting pressure on LPRS phonology as far as reading tasks are concerned. The data also show that the pattern that was thought to be maintained from

IPRS, i.e., the raising of [e] and [o], was not observed in Lorain, despite the fact that the first generation of Puerto Ricans in Lorain came from the regions in the island that have been documented to have this variation.

The main results of the study are revealing of the pressures of MAS and AE on LPRS and the outcomes of dialect and language contact. Instances of the retroflex [] were collected in the recordings of the Lorain Puerto Ricans, although these were very few.

However, those few instances of the retroflex were found among the third generation of

Puerto Ricans. No retroflex was found among the Mexicans, contrary to what is described in the Spanish linguistic literature. As discussed in Chapter 6, the evidence for a retroflex in MAS comes from the variety of this dialect spoken in the Southwest. Thus,

174

I had earlier proposed that, if a similar sound was found in LPRS, it was due to MAS

influence. However, inasmuch as a retroflex has been shown not to exist in MAS in

Lorain, I thus argue now that the retroflex in LPRS is due to AE influence. Furthermore,

the retroflex is found when the LPRS speakers are reading, not when they are speaking.

Since the LPRS speakers are exposed more to English in the classroom than Spanish, and, thus, are more proficient in reading in this language, I propose that the presence of the retroflex in their reading tasks is due to AE influence. On the other hand, first generation speakers are also producing a retroflex. Since this generation is not in direct contact with AE speakers because of their lack of proficiency in this language, but on the other hand know of the negative prestige that lateralizing their [r]s carry in IPRS, it seems that they avoid the use of [l] in reading tasks and use other alternatives instead, such as deletion and a retroflex [], instead of [l]. I can only speculate about the reasons as to why a first generation speaker will use a retroflex [] as an alternative, but perhaps they are imitating the younger speakers of the community. An alternative explanation for the

retroflex [] in the first generation can be attributed to indirect influence from AE through

contact with speakers of the same generation in contact with AE and whose LPRS has

been influenced directly by AE. Eventually, the LPRS of these speakers in particular

affect the LPRS of those speakers of the first generation who are not in contact with AE.

The Mexicans are also differentiating their phonology from that of the Puerto Ricans in

Lorain, that is, instead of using an [l] in informal settings, they prefer the use of [r] for

both reading and speaking. Although it is a very low percentage, it is important to note

that the MAS speakers have an overall total use of [l] of 16%. I will not speculate at this

175

time about the presence of the [l] in MAS since the main focus of this project is LPRS,

but several explanations (parallel to those proposed for variation in LPRS) can be

proposed such as the possibility of MAS converging with LPRS in this feature. This,

however, will require further research into the phonology of MAS in Lorain.

Moreover, similar to the presence of the retroflex in the readings tasks for the Lorain

Puerto Ricans, I propose that literacy and orthographic issues influence the production of

[v]. For example, speakers of the second and third generations produce [v] whenever

they saw a written because of their schooling in English. Second and third

generation speakers have more exposure to texts written in English, and thus, produce [v]

for written whenever they read a text in Spanish. The lack of literacy in Spanish also

plays a role in that Hispanics do not necessarily know that written in Spanish does

not stand for the labiodental fricative.

LPRS third generation speakers are replacing [dΩ] by either [j] or [i], while first and second generations are still maintaining the same [dΩ] that was transmitted to them from

IPRS. MAS young speakers have both [j] and [i] as the preferred variants. Thus, third generation LPRS and younger MAS speakers are converging, at the same time that they are diverging from the variable preferred by first and second generation LPRS speakers.

This community of Puerto Ricans is experiencing a rapid shift to AE among the three different generations. The shift is due to their perceived identity, acculturation, and/or communicative needs for Spanish vs. AE. The Lorain community of Puerto Ricans is not

176

isolated; it is a dense community that interacts with Mexican Americans and AE

speakers. The community also engages in continuous interaction with Spanish speakers

from New York, Cleveland, as well as other U.S. cities, (including those in the

Southwest), from Puerto Rico, and from Mexico. Regardless of its physical distance

from Mexico and the Caribbean, a constant flow of newcomers has been maintained through the years, and Latinos currently living in Lorain also travel to their ancestral countries and to other Hispanic communities in the United States. The continuous contact and travel to the ancestral countries make it hard to say what will happen to

Lorain Spanish, but one can predict that Spanish will be kept alive with their input.

Similar to many Mexican-border communities (Hudson-Edwards and Bills 1982), the first generation maintains the use of Spanish at home. Thus, first and second generations prefer to use Spanish with them, encouraging LPRS maintenance in the community

(Gonzales & Wherrit 1990). In the Spanish-speaking community in Fortuna, California

(Rivera-Mills 2001) by contrast, less than half of the first generation speakers use

Spanish at home and consequently there is a strong tendency to exclusive AE usage in the same locality while latter generations are completely losing Spanish and experiencing a dramatic shift to AE. The Lorain Puerto Rican community, however, has maintained its

Spanish variety due in part to the first generation usage (Gonzales & Wherrit 1990).

Issues related to the need to use Spanish, particularly among speakers who deal with customers and students, have been expressed among the members of the PRS Lorain community. Many participants express a concern that Spanish is becoming used more

177

frequently all over the United States and that more often they are meeting with customers

who only speak Spanish as well as students who come directly from Puerto Rico and

other Spanish-speaking countries who do not know AE. For communicative purposes,

they resort to Spanish. Those who do not speak the language tend to re-enroll in Spanish classes at the local community college or other teaching institutions. However, in these institutions of higher learning, the language taught is not LPRS, but Standard Spanish.

Thus, the language that these individuals think they are reviving is not their ancestral

language, but another variety, a situation very similar to that found among the youngest

generation of Russian Old Believers in Erie, Pennsylvania (Holdeman 2002) who speak

Suvalki Old Believer Russian, but who learned Standard Russian in college. Similarly to

the case of the Russian Old Believer community in Erie, Pennsylvania, the Puerto Ricans

in Lorain look down upon their language and apologize for their “incorrect” Spanish (see

Chapter 3). The question arises then as to which variety, if any, will be transmitted by

these individuals in particular in Lorain to the future generations. The transmission of

this variety, however, will not be the natural transmission that Nicole and Phillip (see

Chapter 3) and many others in the community experience. These latter individuals are

the ones who are the natural actors in the transmission of LPRS, helping to preserve their

Spanish in the community and passing it along to the following generation42.

The community of Hispanics in Lorain also has access to news outlets in Spanish. There

are newspapers written in Spanish that discuss issues of interest to Mexican Americans,

Puerto Ricans and Latinos in general. These newspapers, however, do not publish

42 Nicole is a young single mother and, along with her second generation mother, is raising two children. Nicole and her mother were firm in admitting that they use LPRS with the children.

178 headline news. There are also two radio stations in Spanish that play Mexican, Mexican

American and Puerto Rican music. Cable television has also made available to the

Latinos in Lorain several TV stations that broadcast programs in Spanish. The linguistic varieties they are exposed to range from Peninsular Spanish to Standard Latin American

Spanish to Spanglish. Literature and music are also available in Spanish in Lorain in many of the local stores and markets.

Also, the role of the different religious congregations is important as they encourage

Spanish usage in their services and people who are conducting, or who are leaders, are required to have knowledge of Spanish in order to read, speak and comprehend the language. Many times, there are either newcomers to their churches or people who are from Mexico or Puerto Rico who are visiting so that the use of Spanish is needed.

Similarly, the schools are constantly serving new students, primarily from Puerto Rico, and their teachers have to make use of their Spanish to foster communication. Similarly, employers at banks, hospitals and stores need Spanish-speaking employees to help their customers and the need is encouraging the Lorain Latino youth to learn Spanish.

However, similar to the Spanish-speaking community in Fortuna, California (Rivera-

Mills 2001), the second generation shows an increase in AE language use across different localities. Thus, although the loyalty of the second generation is undeniably towards

Spanish, the fact that AE is preferred in specific locales such as the bank, stores, post office and the like, their need to survive in the mainstream community requires AE

(Rivera-Mills 2001). The conscious efforts of the second generation to still maintain the

179 use of Spanish outside the home environment encourage members of their community, particularly third generation speakers and Mexican Americans to keep Spanish alive in

Lorain.

Thus, several factors come into play in the Lorain Puerto Rican community encouraging its members to maintain and/or learn Spanish. Although the first generation is still using

Spanish on a daily basis and the second generation works to promote its maintenance, it is really the third generation who will decide whether Spanish will survive the AE influence, or if it will succumb to its pressures as many other languages in the United

States and around the world have. The need to use Spanish and the positive attitudes towards the language are strong in the community and the constant input of new Spanish speakers may encourage the survival of Spanish. However, if the need to speak Spanish is lost, if attitudes towards its usage change and if the newcomers acculturate to the mainstream society, Spanish, just as many other endangered languages, will experience death in the community in the next generations, a somewhat strange situation since the status of this language worldwide is rather robust.

180

APPENDIX A

WORD LIST IN SPANISH

181

Instructions/Instrucciones

Before reading the tables Please, read each word twice.

Antes de la lectura de las tablas Por favor, lea las siguientes palabras dos veces.

182

1. 2. 3. 4. 5. ocho Mago lago asco anda envidia Quemazón guerrero ñeque terrero perro Leche cierra risa hierba televisor Sentir querer mar radio lluvia Pero lleno llovizna comida Yiya Llanto tayote cocuyo nene bola esa yerba bélico vela tayote cabes Volar cabe estima bola Juan Memorias tren patín lento año Ratón tu yerba niño vivir

183

6. 7. 8. 9. 10. caldo tu yuca amenaza ya campo horno labio caro botar sillón cocuyo caldo triste oscuro letrina biberón Irma bacalao aquella abecedario ave leyes hierva horno cayó leyenda volar tu yate pedir funciona televisor martes betún yegua tirijala usted gallina terrero octavo cabo maya cano carrera pollo pote vaca según gato hierba un yugo

184

11. 12. 13. lesión pelo tirijala caballo baca ponga amenaza El Yunque mejora rico rosa Pérez ayer pasto urna votar borroso seguro yo betún quesos pera ruso regreso singular tu yegua normal tosco funciona mejora

185

APPENDIX B

READING PASSAGE IN SPANISH

186

Instructions/Instrucciones

Before the passage reading Please, read the following passage at your regular reading speed.

Antes de la lectura del pasaje Por favor, lea el siguiente pasaje al ritmo que regularmente usa para leer.

187

En un campo del Yunque... Michelle F. Ramos-Pellicia

El martes, a las ocho, en un campo del Yunque, Juan, el Mago, le dijo a Irma: “Esa yerba que está al lado del camino del tren es buena comida para las vacas.” Juanito, el nieto de

Irma Pérez, vino un poco después llorando con Yiya, la nena de Irma y mamá de Juanito.

El nene necesitaba la bola que cayó cerca del caballo, el trencito con el que el perro jugaba, y, también, quería volar la chiringa, pero las gallinas la habían picoteado. Nadie se había dado cuenta que el nene lloraba porque estaban hablando entre ellos, escuchando la música que tocaba la radio, pero el llanto los hizo despertar. Juanito tenía todavía puesto el patín que no funcionaba. El nene estaba triste, aunque jugaba con el gato.

Tenía el biberón en la boca porque no quería comer el pollo al horno que habían hecho los mayores. Prefería tomar la leche de pote que su mamá siempre le hierve todos los días. Estaba lleno de leche y no quería más comida. Juan, el Mago, cogió al nene en brazos. Le dio un apretón bien fuerte, le hizo muchas cucas monas y le dio muchas chucherías, entre ellas un carro precioso y radiante que lo corrió sin parar por todo aquello. Dio una carrera tremenda alrededor de los mayores. Así, el nene cambió el llanto desesperado por una sonrisa que valía un millón y se olvidó de la bola que cayó cerca del caballo, del trencito con el que el perro jugaba y de la chiringa que las gallinas habían picoteado. Su risa jugueteaba en el aire y hacía recordar y regresar a los adultos a la niñez.

188

APPENDIX C

QUESTIONS FOR THE INTERVIEW IN SPANISH

189

PREGUNTAS PARA LA ENTREVISTA

¿Tiene valor conocer y hablar inglés en Lorain? ¿Es importante? ¿Por qué?

¿Tiene valor conocer y hablar español en Lorain? ¿Es importante? ¿Por qué?

¿Prefiere usted que la gente hable inglés en Lorain? ¿Por qué?

¿Prefiere usted que la gente hable español en Lorain? ¿Por qué?

¿Algún miembro de su familia habla más inglés que los demás? ¿Hablan más inglés las personas mayores? ¿Cuándo? ¿Con quién? ¿Hablan más inglés las personas jóvenes? ¿Cuándo? ¿Con quién? ¿Usa usted inglés con los miembros mayores o menores de la familia? ¿Con cuánta frecuencia?

¿Algún miembro de su familia habla más español que los demás? ¿Hablan más español las personas mayores? ¿Cuándo? ¿Con quién? ¿Hablan más español las personas jóvenes? ¿Cuándo? ¿Con quién? ¿Usa usted español con los miembros mayores o menores de la familia? ¿Con cuánta frecuencia?

¿Hay puertorriqueños/as en Lorain que nunca usan español? ¿Puertorriqueños/as mayores? ¿Puertorriqueños/as jóvenes? ¿Puertorriqueños/as con más dinero? ¿Puertorriqueños/as pobres? ¿Puertorriqueños/as con más escuela? ¿Puertorriqueños/as con menos escuela? ¿Hombres puertorriqueños? ¿Mujeres puertorriqueñas?

¿Hay puertorriqueños/as en Lorain que nunca usan inglés? ¿Puertorriqueños/as mayores? ¿Puertorriqueños/as jóvenes? ¿Puertorriqueños/as con más dinero? ¿Puertorriqueños/as pobres? ¿Puertorriqueños/as con más escuela? ¿Puertorriqueños/as con menos escuela? ¿Hombres puertorriqueños? ¿Mujeres puertorriqueñas?

190

¿Le sugiere a usted algo el uso del inglés sobre el carácter del puertorriqueño? ¿Qué tipo de puertorriqueño/a es el que usa inglés? Positivo Nada Negativo

¿Le sugiere a usted algo el uso del español sobre el carácter del puertorriqueño? ¿Qué tipo de puertorriqueño/a es el que usa español? Positivo Nada Negativo

¿Hay algún sitio donde esté bien o mal usar inglés en Lorain? ¿Cuándo usaría usted inglés en Lorain? Hablar con un/a amigo/a puertorriqueño/a sobre un programa de televisión Hablar con un/a amigo/a americano/a sobre un programa de televisión Escribir una carta a un familiar Enseñar en la escuela Hablar con el/la supervisor/a Contestar el teléfono Escribir una carta a un periódico En una entrevista de trabajo

¿Hay algún sitio donde esté bien o mal usar español en Lorain? ¿Cuándo usaría usted español en Lorain? Hablar con un/a amigo/a puertorriqueño/a sobre un programa de televisión Hablar con un/a amigo/a americano/a sobre un programa de televisión Escribir una carta a un familiar Enseñar en la escuela Hablar con el/la supervisor/a Contestar el teléfono Escribir una carta a un periódico En una entrevista de trabajo

¿Sería aceptable para usted si un/a amigo/a puertorriqueño/a le describe un juego de baloncesto en inglés?

¿Sería aceptable para usted si un/a amigo/a puertorriqueño/a le describe un juego de baloncesto en español?

¿Sería aceptable para usted si un/a puertorriqueño/a le preguntara por direcciones en inglés?

¿Sería aceptable para usted si un/a puertorriqueño/a le preguntara por direcciones en español?

191

¿Sería aceptable para usted si un/a puertorriqueño/a hiciera un chiste en inglés?

¿Sería aceptable para usted si un/a puertorriqueño/a hiciera un chiste en español?

¿Sería aceptable para usted si un/a puertorriqueño/a rezara en inglés?

¿Sería aceptable para usted si un/a puertorriqueño/a rezara en español?

¿Sería aceptable para usted si un/a puertorriqueño/a maldijera en inglés?

¿Sería aceptable para usted si un/a puertorriqueño/a maldijera en español?

¿Sería aceptable para usted si un/a puertorriqueño/a contara cuentos para niños en inglés?

¿Sería aceptable para usted si un/a puertorriqueño/a contara cuentos para niños en español?

¿Mezcla usted los dos idiomas cuando habla? ¿Por qué?

En su opinión, ¿es bueno usar ambos idiomas cuando habla?

En su opinión, ¿por qué es bueno usar ambos idiomas cuando habla?

En su opinión, ¿no es bueno usar ambos idiomas cuando habla?

En su opinión, ¿por qué no es bueno usar ambos idiomas cuando habla?

¿Cuáles son las ventajas de usar ambos idiomas en una conversación?

¿Cuáles son las desventajas de usar ambos idiomas en una conversación?

¿Cuál es su reacción cuando otras personas usan ambos idiomas en una conversación con usted o con otros?

¿Cuándo fue su última conversación en inglés?

¿Cuándo fue su última conversación en español?

192

APPENDIX D

QUESTIONNAIRE IN SPANISH

193

Cuestionario

Instrucciones: Para contestar cada pregunta, marque con una X en la línea que corresponda o complete la información.

I. Información General

1. Edad: _____ 2. Sexo: ______

3. Lugar de nacimiento: ______

4. Lugar donde creció: ______

5. Lugar donde ha vivido la mayor parte de su vida: ______

6. Lugar donde vive actualmente: ______

7. ¿De qué tipo de escuela se graduó?/¿En qué tipo de escuela estudia? _____ Pública _____ Privada

II. Contextos donde se usan los idiomas

8. Especifique si las clases que no eran de lenguas se impartían en inglés o en español. Escoja una respuesta: _____ a. Clases en inglés desde tercer grado o antes _____ b. Clases en inglés en la escuela superior _____ c. Clases en español _____ ch. Clases en ambas lenguas, pues era un programa bilingüe

9. Si ha hecho/hace parte de sus estudios en los Estados Unidos o alguna base militar norteamericana, especifique: a. Nivel o niveles que estudió/estudia (marque todos los que apliquen) _____ 1. Antes de la escuela elemental _____ 2. Escuela elemental _____ 3. Escuela superior _____ 4. Universidad b. El tiempo que estudió en los Estados Unidos _____ 1. Un año o menos de un año _____ 2. Entre uno y dos años _____ 3. Más de tres años

10. ¿Cuál idioma aprendió primero? _____ Español _____ Inglés _____ Ambas

194

11. ¿ Cuál idioma se usaba en su hogar cuando era un/a niño/a? _____ Inglés _____ Español _____ Ambas

12. ¿Qué edad tenía cuando aprendió español? _____ inglés? _____

13. ¿Cuál idioma usa con Español Inglés Ambas a. sus padres? ______b. sus hermanos/as? ______c. sus amigos/as en la escuela y en el trabajo? ______ch. sus amigos/as en fiestas y reuniones? ______d. su/s hijo/s y o hija/s ______e. su/s mascota/s? ______f. Dios? ______g. con usted mismo/a? ______

14. Mencione los nombres de tres amistades en Lorain a quienes usted ve con frecuencia durante la semana. Nombre Idioma Frecuencia con la que hablan durante la semana a. ______b. ______c. ______

195

15. Mencione los nombres de tres amistades en Lorain que sean (1) hablantes nativos del inglés, (2) que usted ve con frecuencia cada semana y (3) con los que usted hable mayormente inglés. ¿Cuál es la relación entre ustedes? (Por ejemplo, maestro/a, amigo/a, compañero/a de la escuela, vecino/a, chofer de la guagua escolar, etc.). ¿Cuánto tiempo comparten cada semana? Nombre Relación Frecuencia con la que hablan durante la semana a. ______b. ______c. ______

16. Mencione los nombres de tres amistades en Lorain que sean (1) mexicanos o mexicoamericanos, (2) a quienes usted ve cada semana. ¿Cuál es la relación entre ustedes? (Por ejemplo, maestro/a, amigo/a, compañero/a de la escuela, vecino/a, chofer de la guagua escolar, etc.). ¿Cuánto tiempo comparten cada semana? Nombre Relación Idioma Frecuencia con la que hablan durante la semana a. ______b. ______c. ______

17. ¿Con cuánta frecuencia usted Todos 2 ó 3 Una Una o Nunca los veces por vez por dos veces días semana semana al mes a. habla inglés en casa? ______b. habla inglés fuera de casa? ______c. habla inglés con: 1. su esposo/a ______2. su/s hijo/s y/o hija/s ______3. sus familiares ______4. sus amistades ______5. sus vecinos/as ______6. sus compañeros/as de escuela ______7. sus compañeros/as de trabajo ______8. un/a empleado/a de tiendas o bancos ______9. su/s mascota/s ______10. con usted mismo? ______11. Dios? ______

196

Todos 2 ó 3 Una Una o Nunca los veces por vez por dos veces días semana semana al mes 12. otros ______ch. habla español en su casa? ______d. habla español fuera de casa? ______e. habla español con: 1. su esposo/a ______2. su/s hijo/s y/o hija/s ______3. sus familiares ______4. sus amistades ______5. sus vecinos/as ______6. sus compañeros/as de escuela ______7. sus compañeros/as de trabajo ______8. un/a empleado/a de tiendas o bancos ______9. su/s mascota/s ______10. con usted mismo? ______11. Dios? ______12. otros ______

18. ¿Cuál idioma prefiere cuando ve televisión? _____ Español _____ Inglés _____ Ambas

19. ¿Cuál idioma prefiere para sus lecturas? _____ Inglés _____ Español _____ Ambas

III. Dominio del idioma

20. ¿Puede leer en: _____ inglés? _____ español? _____ ambas?

21. ¿Puede escribir en: _____ español? _____ inglés? _____ ambas?

22. ¿Puede hablar en: _____ inglés? _____ español? _____ ambas?

197

23. ¿Cuál idioma usted domina mejor cuando usted: Español Inglés Ambas a. lee? ______b. escribe? ______c. conversa? ______

24. ¿Cómo evaluaría su dominio del español en la lectura, escritura y conversación? Escoja el número que mejor describa su dominio del idioma: 1 significa “no es bueno”, 5 significa “regular” y 10 significa “perfecto”. a. lectura 1 2 3 4 5 6 7 8 9 10 b. escritura 1 2 3 4 5 6 7 8 9 10 c. conversación 1 2 3 4 5 6 7 8 9 10

25. ¿Cómo evaluaría su dominio del inglés en la lectura, escritura y conversación? Escoja el número que mejor describa su dominio del idioma: 1 significa “no es bueno”, 5 significa “regular” y 10 significa “perfecto”. a. lectura 1 2 3 4 5 6 7 8 9 10 b. escritura 1 2 3 4 5 6 7 8 9 10 c. conversación 1 2 3 4 5 6 7 8 9 10

26. Cuando habla español, ¿lo entienden los hablantes de este idioma? _____ Sí _____ No

27. Cuando habla inglés, ¿lo entienden los hablantes de este idioma? _____ Sí _____ No

IV. Actitudes hacia el idioma

28. ¿Cree usted que para ser considerado/a puertorriqueño/a se tiene que entender y hablar español? _____ Sí _____ No

29. ¿Cree usted que lo/as puertorriqueño/as que no entienden o hablan el español dividen la comunidad? _____ Sí _____ No

30. ¿Cree usted que lo/as puertorriqueño/as de esta comunidad han conservado su cultura? _____ Sí _____ No

31. ¿Entienden y hablan español los puertorriqueños de esta comunidad? _____ Sí _____ No

32. ¿Entienden y hablan inglés los puertorriqueños de esta comunidad? _____ Sí _____ No

198

33. ¿Es importante para usted hablar español? _____ Sí _____ No

34. ¿Es importante para usted hablar inglés? _____ Sí _____ No

35. ¿Quieren hablar español los/as jóvenes puertorriqueños/as de esta comunidad? _____ Sí _____ No

36. ¿Quieren hablar inglés los/as jóvenes puertorriqueños/as de esta comunidad? _____ Sí _____ No

37. ¿Entienden español los/as jóvenes puertorriqueños/as de esta comunidad? _____ Sí _____ No

38. ¿Entienden inglés los/as jóvenes puertorriqueños/as de esta comunidad? _____ Sí _____ No

V. Información Personal

39. De los siguientes niveles, ¿cuál se acerca más a la profesión o el empleo de su familia? _____ a. Obrero no diestro (el trabajo no requiere entrenamiento, ni destrezas especiales: trabajos en construcción, en fábricas, etc.). _____ b. Obrero diestro (el trabajo requiere entrenamiento y destrezas, pero no requiere estudios universitarios: perito electricista, ebanista, cosmetología, etc.) _____ c. Trabajo de oficina o ventas (secretarial, administrativo, negocio propio, etc.) _____ ch. Técnico (el trabajo es especializado y requiere algunos estudios universitarios: técnico de ingeniería, de laboratorio, etc.) _____ d. Profesional (el trabajo requiere, generalmente, cuatro o más años de estudios universitarios: trabajo social, contabilidad, etc.) _____ e. Pensionado _____ f. Otro _____ g. Desempleado

199

40. ¿Cuánto es el ingreso mensual (anual) con que cuenta su familia para el sostenimiento del hogar? Incluya todos los ingresos tales como salarios, beneficios de negocios, pensiones, alquileres, ayuda de familiares, etc. _____ a. Menos de $250 mensuales (menos de $3,000 por año) _____ b. Entre $251 y $400 mensuales (entre $3,001 y $4,800 por año) _____ c. Entre $401 y $500 mensuales (entre $4,801 y $6,000 por año) _____ ch. Entre $501 y $750 mensuales (entre $6,001 y $9,000 por año) _____ d. Entre $751 y $1,083 mensuales (entre $9,001 y $13,000 por año) _____ e. Entre $1,084 y $1,666 mensuales (entre $13,001 y $20,000 por año) _____ f. Entre $1,667 y $2,500 mensuales (entre $20,001 y $30,000 por año) _____ g. Más de $2,500 mensuales (más de $30,000 por año)

41. Indique su nivel de educación más avanzado: _____ a. No asistí a la escuela. _____ b. Asistí a la escuela, por favor, indique el grado más avanzado que completó: ______c. Cursos técnicos, vocacionales, etc. _____ ch. Uno o más años de universidad _____ d. Grado Asociado _____ e. Bachillerato _____ f. Maestría o Ph.D.

42. Indique el nivel de educación más avanzado de su madre o encargada: _____ a. No asistió a la escuela. _____ b. Asistió a la escuela, por favor, indique el grado más avanzado que completó: ______c. Cursos técnicos, vocacionales, etc. _____ ch. Uno o más años de universidad _____ d. Grado Asociado _____ e. Bachillerato _____ f. Maestría o Ph.D.

43. Indique el nivel de educación más avanzado de su padre o encargado: _____ a. No asistió a la escuela. _____ b. Asistió a la escuela, por favor, indique el grado más avanzado que completó: ______c. Cursos técnicos, vocacionales, etc. _____ ch. Uno o más años de universidad _____ d. Grado Asociado _____ e. Bachillerato _____ f. Maestría o Ph.D.

200

44. Incluyéndole/la a usted, ¿cuántas personas dependen del ingreso familiar informado? _____ a. Una _____ b. Dos _____ c. Tres _____ ch. Cuatro _____ d. Cinco _____ e. Seis _____ f. Siete

45. Por favor, mencione las edades de sus hermanos/as. a. Hermanos: ______b. Hermanas: ______

201

APPENDIX E

WORD LIST IN ENGLISH

202

Instructions

Before reading the tables Please, read each word twice.

203

1. 2. 3. 4. 5. robot cash mere color television therapy debilitate Hummer Vegas ballot vulgar hung share mayor answers judge terminator vermillion Marsha chemical questions jail bang desk mobile

204

6. 7. 8. 9. 10. aback robotic tar yard development far chair shy vote James hoarse bonus vanity calorie arm steering John purchase memorial believe tan car king buyer yellow

205

11. 12. 13. 14. Yale therapeutic renovate kin trash labial hour shoe vision chime choose ban chemistry glamour jeer evacuate mobility caloric pencils penitentiary

206

15. 16. 17. 18. morning number Jello adventure binocular spin cheese pass shine vanish flower march year jar bulletin rang ran unless renovation memory

207

19. 20. 21. Jeep Hon’ lavatory desks terminator banish penitent boat Humvee spring jarred father horse aback Tang

208

APPENDIX F

READING PASSAGE IN ENGLISH

209

Instructions

Before the passage reading Please read the following passage at your regular reading speed.

210

A trip to Vegas Michelle F. Ramos-Pellicia

Marsha and John purchased a Jeep and went to Vegas. Marsha took her mobile to stay in touch with James, their son who was in Yale. They always joked that it was better that

James was in Yale and not in jail. Marsha and John loved the mobility that having a college son provided them now.

Marsha suddenly started remembering when James was a child. After the renovation, in their renovated yard, John’s father had an old yellow Humvee. When James was a little boy, he was very shy, but --when he would hop in the yellow Humvee-- he became a complete different child. James pretended he was a Terminator, a robot with robotic impulses, inside the destructive car, or would pretend he was a king on top of a horse. He loved to drink Tang and eat cheese. He always had many questions and expected Marsha and John to have answers to all of them. At night, James will go to his room and pretend he was sleeping, but, when John will go into the room to check on the child, he would find James at his desk reading. Sometimes, James will even fall asleep in his chair, with some of his pencils still in his hand. John will have to bring him to his bed to sleep. She admired her clear memory and wish that some day she will create a memorial for John’s father too. All this, Marsha was thinking about, when John put his arm around her and admire the vermillion cashmere sweater that she was wearing.

211

APPENDIX G

QUESTIONS FOR THE INTERVIEW IN ENGLISH

212

QUESTIONS FOR THE INTERVIEW

Is it valuable to know and speak English in Lorain? Is it important in Lorain? Why?

Is it valuable to know and speak Spanish in Lorain? Is it important in Lorain? Why?

Would you prefer that people speak English in Lorain? Why?

Would you prefer that people speak Spanish in Lorain? Why?

Are any members of your family more likely to speak English than others? Do older people use more English? At what times? With whom? Do younger people use more English? At what times? With whom? Do you use English with older/younger family members? At what times?

Are any members of your family more likely to speak Spanish than others? Do older people use more Spanish? At what times? With whom? Do younger people use more Spanish? At what times? With whom? Do you use Spanish with older/younger family members? At what times?

Are there any Puerto Ricans in Lorain who never use Spanish? Older Puerto Ricans? Younger Puerto Ricans? Puerto Ricans with more money? Puerto Ricans with less money? Puerto Ricans with more schooling? Puerto Ricans with less schooling? Puerto Rican males? Puerto Rican females?

Are there any Puerto Ricans in Lorain who never use English? Older Puerto Ricans? Younger Puerto Ricans? Puerto Ricans with more money? Puerto Ricans with less money? Puerto Ricans with more schooling? Puerto Ricans with less schooling? Puerto Rican males? Puerto Rican females?

Does the use of English suggest anything to you about a Puerto Rican’s character? What kind of a Puerto Rican (she) he is if (she) he uses English? Positive Nothing Negative

213

Does the use of Spanish suggest anything to you about a Puerto Rican’s character? What kind of a Puerto Rican (she) he is if (she) he uses Spanish? Positive Nothing Negative

Are there any wrong or right places to use English in Lorain? When would you use English in Lorain? Talk with a Puerto Rican friend about a TV program Talk with an American friend about a TV program Write a letter to a relative Teach in school Talk to a supervisor Answer the telephone Write a letter to the newspaper Conduct a job interview

Are there any wrong or right places to use Spanish in Lorain? When would you use Spanish in Lorain? Talk with a Puerto Rican friend about a TV program Talk with an American friend about a TV program Write a letter to a relative Teach in school Talk to a supervisor Answer the telephone Write a letter to the newspaper Conduct a job interview

Would it be acceptable to you if a Puerto Rican friend described a basketball game in English?

Would it be acceptable to you if a Puerto Rican friend described a basketball game in Spanish?

Would it be acceptable to you if a Puerto Rican asked for directions in English?

Would it be acceptable to you if a Puerto Rican asked for directions in Spanish?

Would it be acceptable to you if a Puerto Rican told a joke in English?

Would it be acceptable to you if a Puerto Rican told a joke in Spanish?

Would it be acceptable to you if a Puerto Rican prayed in English?

Would it be acceptable to you if a Puerto Rican prayed in Spanish?

214

Would it be acceptable to you if a Puerto Rican cursed in English?

Would it be acceptable to you if a Puerto Rican cursed in Spanish?

Would it be acceptable to you if a Puerto Rican told children’s stories in English?

Would it be acceptable to you if a Puerto Rican told children’s stories in Spanish?

Do you mix both languages when you speak? Why?

In your opinion, is it good to use both languages when you speak?

In your opinion, why is it good to use both languages when you speak?

In your opinion, is it not good to use both languages when you speak?

In your opinion, why is it not good to use both languages when you speak?

What are the advantages of using both languages in a conversation?

What are the disadvantages of using both languages in a conversation?

What is your reaction when other people use both languages in a conversation with you or with others?

How long has it been since your last conversation in English?

How long has it been since your last conversation in Spanish?

215

APPENDIX H

QUESTIONNAIRE IN ENGLISH

216

Questionnaire

Directions: In order to answer each question, please make a check mark in the corresponding line or complete the information.

I. General Information

1. Age: _____ 2. Sex: ______

3. Place of birth: ______

4. Place where you grew up: ______

5. Place where you have lived most of your life: ______

6. Place where you currently live: ______

7. Which one will best describe the school where you graduated from/where you study at? _____ Public _____ Private

II. Language Domains

8. Specify if the classes, other than the language classes, were conducted in English or Spanish. Choose one: _____ a. All classes in English since or before third grade _____ b. Classes in English in high school _____ c. Classes in Spanish _____ d. Classes in both languages, it was a bilingual progam

9. If you have studied in the United States, or at a military base, specify: a. Level or levels that you have studied (choose all that you have studied) _____ 1. Before primary school _____ 2. Primary school _____ 3. High school _____ 4. College b. Period of study in the United States _____ 1. A year or less than a year _____ 2. Between one and two years _____ 3. More than three years

10. Which language did you learn first? _____ Spanish _____ English _____ Both

217

11. Which language was used at home when you were a child? _____ English _____ Spanish _____ Both

12. How old were you when you learned _____ Spanish? _____ English?

13. Which language did you use with Spanish English Both a. your parents? ______b. your sibblings? ______c. your friends at school and work? ______d. your friends when you meet in social activities? ______e. your offspring? ______f. your pets? ______g. God? ______h. with yourself? ______

14. Mention the names of three friends in Lorain who you see the most every week. Name of friend Language used Amount of time each week a. ______b. ______c. ______

15. Mention the names of three friends in Lorain who are (1) American English-speakers, (2) you see every week and (3) speak English with the most. What is the relationship to you? (For example: teacher, friend, classmate, neighbor, bus driver, etc.). How much time do you spend with them each week? Name of friend Relationship Amount of time each week a. ______b. ______c. ______

16. Mention the names of three friends in Lorain who are (1) Mexicans or Mexican Americans, (2) you see every week. What is the relationship to you? (For example: teacher, friend, classmate, neighbor, bus driver, etc.). How much time do you spend with them each week? Name of friend Relationship Language Used Amount of time each week a. ______b. ______c. ______

218

17. How often do you Every 2 or 3 Once a Once or Never day times a week twice week a month a. speak English in your home? ______b. speak English outside your home? ______c. speak English to: 1. your husband/ wife? ______2. your children? ______3. your relatives? ______4. your friends? ______5. your neighbors? ______6. classmates? ______7. workmates? ______8. shop/bank employees? ______9. your pet/s? ______10. yourself? ______11. God? ______12. others? ______d. speak Spanish in your home? ______e. speak Spanish outside your home? ______f. speak Spanish to: 1. your husband/ wife? ______2. your children? ______3. your relatives? ______4. your friends? ______5. your neighbors? ______6. classmates? ______7. workmates? ______8. shop/bank employees? ______9. your pet/s? ______10. yourself? ______11. God? ______12. others? ______

18. Which language do you prefer when watching TV? _____ Spanish _____ English _____ Both

19. Which language do you prefer for recreational reading? _____ English _____ Spanish _____ Both

219

III. Language Proficiency

20. Can you read in: _____ English? _____ Spanish? _____ Both?

21. Can you write in: _____ Spanish? _____ English? _____ Both?

22. Can you speak in: _____ English? _____ Spanish? _____ Both?

23. Which language are you best at: Spanish English Both a. in reading? ______b. in writing? ______c. in speaking? ______

24. How would you evaluate your proficiency in Spanish in reading, writing and speaking? Circle the number that best describes your language proficiency: 1 means “not good”, 5 means “regular”, and 10 means “perfect”. a. reading 1 2 3 4 5 6 7 8 9 10 b. writing 1 2 3 4 5 6 7 8 9 10 c. speaking 1 2 3 4 5 6 7 8 9 10

25. How would you evaluate your proficiency in English in reading, writing and speaking? Circle the number that best describes your language proficiency: 1 means “not good”, 5 means “regular”, and 10 means “perfect”. a. reading 1 2 3 4 5 6 7 8 9 10 b. writing 1 2 3 4 5 6 7 8 9 10 c. speaking 1 2 3 4 5 6 7 8 9 10

26. When you speak Spanish, do native speakers of the language understand you? _____ Yes _____ No

27. When you speak English, do native speakers of the language understand you? _____ Yes _____ No

IV. Language Attitudes

28. Do you think that in order to be considered a Puerto Rican one has to understand and speak Spanish? _____ Yes _____ No

29. Do you think that Puerto Ricans who do not understand or speak Spanish divide the community? _____ Yes _____ No

220

30. Do you think that the Puerto Ricans in this community have preserved their culture? _____ Yes _____ No

31. Do Puerto Ricans in this community understand and speak Spanish? _____ Yes _____ No

32. Do Puerto Ricans in this community understand and speak English? _____ Yes _____ No

33. Is it important for you to speak Spanish? _____ Yes _____ No

34. Is it important for you to speak English? _____ Yes _____ No

35. Do the young Puerto Ricans in the community want to speak Spanish? _____ Yes _____ No

36. Do the young Puerto Ricans in the community want to speak English? _____ Yes _____ No

37. Do the young Puerto Ricans in the community understand Spanish? _____ Yes _____ No

38. Do the young Puerto Ricans in the community understand English? _____ Yes _____ No

V. Background Information

39. From the following, which one best describes the employment level of your household? _____ a. Non-skilled worker (the job does not require training or special skills: work in construction, factory, etc.) _____ b. Skilled worker (the job requires training and skills, but it does not require a college degree: electrician, factory, etc.) _____ c. Clerical work (office assistant, administrative, own business, etc.) _____ d. Technician (the job is specialized and requires some college preparation) _____ e. Professional (the job requires four or more years of college: social work, accountant, etc.) _____ f. Pensioner _____ g. Other _____ h. Unemployed

221

40. What is the total monthly (yearly) income of your household? Please include every salary, business benefits, pensions, family gifts, etc. _____ a. Less than $250 per month (less than $3,000 per year) _____ b. Between $251 and $400 per month (between $3,001 and $4,800 per year) _____ c. Between $401 and $500 per month (between $4,801 and $6,000 per year) _____ d. Between $501 and $750 per month (between $6,001 and $9,000 per year) _____ e. Between $751 and $1,083 per month (between $9,001 and $13,000 per year) _____ f. Between $1,084 and $1,666 per month (between $13,001 and $20,000 per year) _____ g. Between $1,667 and $2,500 per month (between $20,001 and $30,000 per year) _____ h. More than $2,500 per month (more than $30,000 per year)

41. Specify your highest level of education. _____ a. I did not go to school. _____ b. Kinder to 12, please specify highest level completed: ______c. Technical courses, vocational, etc. _____ d. One or two more years of college _____ e. Associate degree _____ f. Bachelors degree _____ g. Masters or Ph.D.

42. Specify your mother’s highest level of education. _____ a. She did not go to school. _____ b. Kinder to 12, please specify highest level completed: ______c. Technical courses, vocational, etc. _____ d. One or two more years of college _____ e. Associate degree _____ f. Bachelors degree _____ g. Masters or Ph.D.

222

43. Specify your father’s highest level of education. _____ a. He did not go to school. _____ b. Kinder to 12, please specify highest level completed: ______c. Technical courses, vocational, etc. _____ d. One or two more years of college _____ e. Associate degree _____ f. Bachelors degree _____ g. Masters or Ph.D.

44. Including yourself, how many persons live in your household? _____ a. one _____ b. two _____ c. three _____ d. four _____ e. five _____ f. six _____ g. seven or more

45. Please mention the ages of your siblings. a. Brothers: ______b. Sisters: ______

223

BIBLIOGRAPHY

Agüero Chaves, Arturo. 1960. El español en Costa Rica. San José: Universidad de Costa Rica.

Agüero Chaves, Arturo. 1962. El Español de América y Costa Rica. San José: A. Lehmann.

Alonso, Amado. 1930. Estudios sobre el español de Nuevo Méjico. Buenos Aires: Universidad de Buenos Aires.

Alers-Montalvo, Manuel. 1982. The Puerto Rican migrants of New York City: a study of anomie. New York: AMS Press.

Alvar, Manuel. 1959. El español hablado en Tenerife. Madrid.

Álvarez-Nazario, Manuel. 1991. Historia de la lengua española en Puerto Rico. San Juan: Academia Puertorriqueña de la Lengua Española.

Álvarez Nazario, Manuel. 1990. El habla campesina del país. Río Piedras, PR: Editorial Universitaria.

Álvarez-Nazario, Manuel. 1972. La herencia lingüística de Canarias en Puerto Rico. San Juan: Instituto de Cultura Puertorriqueña.

Amastae, Jon & David Satcher. 1993. Linguistic assimilation in two variables. Language variation and change. 5: 77-90.

Anderson, Keith O. and Williard M. Martin. 1976. Language loyalty among the Pennsylvania Germans: A status report on Old Order Mennonites in

224

Pennsylvania and Ontario. Germanica Americana, ed. by Erich A. Albrecht and J. Anthony Burzle, 73-80. Lawrence: University of Kansas.

Attinasi, John J. 1982. Hispanic attitudes in Northwest Indiana and New York. Spanish language use and public life in the United States, ed. by: Lucía Elías- Olivares, Elizabeth A. Leone, René Cisneros and John R. Gutiérrez, 27-58. New York: Mouton Publishers.

Bailey, Charles-James and K. Maroldt. 1977. The French lineage of English. Langues en contact – Pidgins – Creoles –Languages in contact, ed. by: Jürgen M. Meisel, 21-53. Tübingen: TBL Verlag, G. Narr.

Barrutia, Richard and Armin Schwegler. 1994. Fonética y fonología españolas. New York: John Wiley & Sons, Inc.

Bavin, Edith L. 1989. Some lexical and morphological changes in Walpiri. Investigating obsolescence: Studies in language contraction and death, ed. by Nancy Dorian. Cambridge: Cambridge University Press.

Blom Jan-Petter and John J. Gumperz. 1972. Social meaning in linguistic structures: Code switching in Norway. In: Directions in sociolinguistics, ed. by John J. Gumperz and Dell Hymes, 407-434. New York: Holt, Rinehart and Winston.

Bongaerts,Theo. 1999. Ultimate attainment in L2 pronunciation: The case of very advanced late L2 learners. Second language acquisition and the Critical Period Hypothesis, ed. by David Birdsong, 133-160. Mahwah, N.J.: Erlbaum.

Boissevain, Jeremy. 1974. Friends of friends: networks, manipulators and coalitions. Oxford: Blackwell.

Boissevain, Jeremy and J. Clyde Mitchell (eds.). 1973. Network analysis: studies in human interaction. The Hague: Mouton.

Bortoni-Ricardo, Stella Maris. 1985. The urbanization of rural dialect speakers. A sociolinguistic study in Brazil. Cambridge: Cambridge University Press.

225

Bourdieu, Pierre. 1977. Outline of a theory of practice. Cambridge: Cambridge University Press.

Bourdieu, Pierre. 1982. Ce que parler veut dire : l'économie des échanges linguistiques. Paris: Fayard.

Boyd-Bowman, Peter. 1952. Sobre restos de lleísmo en México. Nueva Revista de Filología Hispánica. 6: 69-74.

Brody, Jill. 1995. Lending the ‘unborrowable’: Spanish discourse markers in Indigenous American languages. Spanish in four continents. Spanish in language contact and bilingualism, ed. by: Carmen Silva-Corvalán. Washington: Georgetown University Press.

Brown, H. Douglas. 1993. Principles of language learning and teaching. Englewood Cliffs, N.J.: Prentice Hall Regents.

Canfield, D. Lincoln. 1981. Spanish pronunciation in the Americas. Chicago: University of Chicago Press.

Canfield, D. Lincoln. 1960. Observaciones sobre el español salvadoreño. Filología. 6: 29-76.

Carreter, Fernando Lázaro. 1990. Diccionario de términos filológicos. Madrid: Editorial Gredos.

Casiano Montañez, L. 1975. La pronunciación de los puertorriqueños en Nueva York. Bogotá: Ediciones Tercer Mundo.

Cassano, Paul Vincent. 1972. La [B] del español del Paraguay, en posición inicial. Revue Romane. Tome VII. Fascicule 2. 186-188.

Cassano, Paul Vincent. 1973. The influence of American English on the phonology of American Spanish. Orbis. 22:1. 201-214.

226

Cassano, Paul Vincent. 1977. Problems in language borrowing and lending exemplified by American Spanish phonology. Orbis. 36:1. 149-163.

Cedergren, Henrietta and David Sankoff. 1974. Variable rules: Performance as a statistical reflection of competence. Language. 50: 333-355.

Cerrón-Palomino, Rodolfo. 1988. Aspectos sociolingüísticos y pedagógicos de la motosidad en el Perú. Pueblos Indígenas y educación 5. marzo: 55-83.

Chambers, Jack K. 1995. Sociolinguistic theory. Massachusetts: Blackwell Publishers.

Chambers, Jack K. 1992. Dialect acquisition. Language. 68: 673-705.

Cheshire, Jenny. 1978. Present tense verbs in Reading, English. In: Sociolinguistic patterns in British English, ed. by: Peter Trudgill. London : Edward Arnold.

Dave, Radhekant. 1977. Retroflex and dental consonants in Gujarati: a palatographic and acoustical study. Annual report of the Institute of Phonetics, University of Copenhagen. 11: 27-156.

Decker, Bob D. 1952. Phonology of the Puerto Rican Spanish of Lorain, Ohio. A study in the environmental displacement of a dialect. Columbus, OH: The Ohio State University M.A. Thesis.

Díaz-Campos, Manuel and Terrell Morgan. 2002. On the production of Spanish [j], [J], [Ô]: an acoustic phonetic study with implications for phonology, dialectology and pedagogy.

Dietz, James L. 1986. Economic history of Puerto Rico: institutional change and capitalist development. New Jersey: Princeton University Press.

Dittmar, Norbert, Peter Schlobinski and Inge Wachs. 1988. The social significance of the Berlin urban vernacular. The sociolinguistics of urban vernaculars. Case studies and their evaluation. Berlin: de Gruyter. 19-43.

227

Domingue, Nicole Z. 1975. Another creole: Middle English. Paper presented at the 1975 International Conference on Pidgins and Creoles, Honolulu.

Dressler, Wolfang and Ruth Wodak-Leodolter. 1977. Language preservation and language death in Brittany. International Journal of the Sociology of Language. 12:33-44.

Eckert, Penelope. 1989. Jocks and burnouts. Social categories and identity in the High school. New York: Teacher’s College Press.

Eckert, Penelope. 1997. Age as a sociolinguistic variable. The Handbook of sociolinguistics, ed. by Florian Coulmas. Oxford, UK ; Cambridge, Massachusetts: Blackwell Publishers.

Eckert, Penelope. 1998. Gender and sociolinguistic variation. Language and gender. A reader, ed. by Jennifer Coates. Oxford: Blackwell.

Eckert, Penelope. 2000. Linguistic variation as social practice: the linguistic construction of identity in Belten High. Massachusetts: Blackwell Publishers.

Edwards, Walter. 1992. Sociolinguistic behavior in a Detroit inner-city black neighborhood. Language in society. 21: 93-115.

Elías-Olivares, Lucía, Elizabeth A. Leone, René Cisneros and John Gutiérrez. (eds.) 1985. Spanish language use and public life in the United States. Berlin: Mouton Publishers.

Espinosa, Aurelio. 1930. Estudios sobre el español de Nuevo Méjico. Parte I Buenos Aires: Biblioteca de Dialectología Hispanoamericana.

Eubank, Lynn and Kevin R. Gregg. 1999. Critical periods and (second) language acquisition: divide et impera. Second language acquisition and the Critical Period Hypothesis, ed. by David Birdsong, 65-100. Mahwah, N.J.: Erlbaum.

Fant, Gunnar. 1968. Analysis and synthesis of speech processes. Manual of phonetics, ed. by Bertil Malmberg, 171-172. North Holland, Amsterdam.

228

Figueroa, Neysa and Kristi Hislope. 1998. A study of syllable final /r/ neutralization in Puerto Rican Spanish. Romance Language Annual. 563-569.

Flege, James, E. 1999. Age of learning and second language speech. Second language acquisition and the Critical Period Hypothesis, ed. by David Birdsong, 101-132. Mahwah, N.J.: Erlbaum.

Gal, Susan. 1998a. Cultural bases of language-use among German-speakers in Hungary. The sociolinguistics reader. Volume 1: Multilingualism and variation, ed. by Peter Trudgill and Jenny Cheshire. London: Arnold.

Gal, Susan. 1998b. Peasant men can’t get wives: Language change and sex roles in a bilingual community. Language and gender. A reader, ed. by Jennifer Coates. Oxford: Blackwell.

Giles, Howard and Philip Smith. 1979. Accomodation theory: optimal levels of convergence. Language and social psychology. Oxford: Basil Blackwell. 45-65.

Giles, Howard, M. Donald Taylor, and Richard Boorhis. 1973. Towards a theory of interpersonal accommodation through speech: some Canadian data. Language in society. 2: 177-192.

Gonzales, Nora and Irene Wherrit. 1990. Spanish language use in West Liberty, Iowa. Spanish in the United States: sociolinguistic issues, ed. by: John Bergen, 67-78. Washington, D.C.: Georgetown University Press.

Gregg, Catherine. 1954. Ethnic membership in community organizations of an Ohio steel town: a study of the rate of social assimilation. New York, New York: Columbia University dissertation.

Greenough, William T. 1986. What’s special about development? Thoughts on the bases experience sensitive synaptic plasticity. Developmental neuropsychobiology, ed. by William Greenough and Janice M. Juraska, 387-407. New York: Academic Press.

229

Hagerty, Timothy. 1996. The influence of English on the Spanish language of Belize. Belize: Selected proceedings from the Second Interdisciplinary Conference, ed. by: Michael D. Phillips, 131-142. Lanham, MD: University Press of America.

Henríquez-Ureña, Pedro. 1938. El español en Méjico, los Estados Unidos y la América Central. Buenos Aires: Biblioteca de Dialectología Hispanoamericana. 4.

Hock, Hans and Brian D. Joseph. 1996. Language history, language change and language relationship. An introduction to historical and comparative linguistics. Berlin, New York: Mouton de Gruyter.

Højrup, Thomas. 1983. The concept of life-mode: a form-specifying mode of analysis applied to contemporary western Europe. Ethnologia Scandinavica. 1-50.

Holdeman, Jeffrey David. 2002. Language maintenance and shift among the Russian Old Believers of Erie, Pennsylvania. Columbus, OH: The Ohio State University Dissertation.

Holmes, John. 2000. An introduction to Pidgins and Creoles. Cambridge: Cambridge University Press.

Horvarth, Barbara. 1998. Finding a place in Sydney: migrants and language and change. The sociolinguistics reader. Volume 1: Multilingualism and variation, ed. by Peter Trudgill and Jenny Cheshire. London: Arnold.

http://factfinder.census.gov/

http://mapquest.com

http://www.census.gov/

http://www.Ci.Lorain.Oh.Us/nfLorain/largeemploy.htm

230

http://www.lonelyplanet.com

Hudson-Edwards, Alan and Bills Garland. 1982. Intergenerational language shift in an Alburquerque barrio. Spanish in the United States: sociolinguistic aspects, ed. by John Amastae and Lucía Elías-Olivares, 135-153. Cambridge: Cambridge University Press.

Institute for Puerto Rican Policy. 1992. Puerto Ricans and other Latinos in New York City today: a statistical profile. New York: Author.

Joseph, Brian D. 1992. The Balkan languages. Oxford International Encyclopedia of Linguistics, ed. by W. Bright, Vol.1: 153-155. Oxford: Oxford University Press.

Keiser, Steven Hartman. 2001. Language change across speech islands: the emergence of a Midwestern dialect of Pennsylvania German. Columbus, OH: The Ohio State University Dissertation.

Kerswill, Paul. 1984. Social and linguistic aspects of Durham (e:). Journal of the International Phonetic Association. 14: 13-34.

Kerswill, Paul. 1987. Levels of linguistic variation in Durham. Journal of Linguistics. 23: 25-49.

Kerswill, Paul. 1993. Rural dialect speakers in an urban speech community: the roles of dialect contact in defining a sociolinguistic concept. International Journal of Applied Linguistics. 3/1: 33-56.

Kerswill, Paul. 1994. Dialects converging. Rural speech in urban Norway. Oxford: Clarendon Press.

Kerswill, Paul. 1996. Milton Keynes and dialect levelling in south-eastern British English. English: history, diversity, and change, ed. by David Graddol, Dick Leith, and Joan Swann. London: Routledge.

231

Kerswill, Paul. 1996b. Divergence and convergence of sociolinguistic structures in Norway and England. Sociolinguistica. 10: 90-104.

Kerswill, Paul. 2000. Linguistics in the UK and Germany: differing approaches to a discipline at the Sciences/Humanities interface. The humanities in the new millennium, ed. by S. Peters, M. Biddiss and I. Roe, 93-103. Tübingen: A. Francke.

Kerswill, Paul and Ann Williams. 1994. A new dialect in a new city: children’s and adults’ speech in Milton Keynes. Final Report.

Kim, Eun Joo. 1997. The sensitive period for second-language acquisition: a reaction- time study of maturational effects on the acquisition of L2 lexico-semantic and syntactic systems. Urbana-Champaign, Illinois: University of Illinois at Urbana- Champaign Dissertation.

King, R. 1989. On the social meaning of linguistic variability in language death situations: variation in Newfoundland French. Investigating obsolescence: studies in language contraction and death, ed. by Nancy Dorian. Cambridge: Cambridge University Press.

Labov, William. 1966. The social stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics.

Labov, William. 1972. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.

Labov, William. 1994. Principles of linguistic change, I: internal factors. Massachusetts: Blackwell.

Labov, William. 2001. Principles of linguistic change, II: social factors. Massachusetts: Blackwell.

Labov, William, Paul Cohen, Clarence Robins, and John Lewis. 1968. A study of the Non-Standard English of Negro and Puerto Rican speakers in New York City. Report on Co-operative Research Project 3288. Washington, D.C.: Office of Education.

232

Labov, William and Wendell Harris. 1986. De facto segregation of Black and White Vernaculars. Diversity and diachrony, ed. by: David Sankoff. Amsterdam: John Benjamins.

Lacayo, Heberto. 1962. Cómo pronuncian el español en Nicaragua. Mexico: Universidad Iberoamericana.

Lacayo, Heberto. 1954. Apuntes sobre la pronunciación del español en Nicaragua. Hispania. 37: 267-268.

Ladefoged, Peter. 1993. A course in phonetics. Forth Worth: Harcourt Brace Jovanovich College Pulishers.

Ladefoged, Peter and Ian Maddieson. 1996. The sounds of the world’s languages. Oxford, Massachussetts: Blackwell.

Lastra de Suárez, Yolanda. 1975. El habla y la educación de los niños de origen mexicano en Los Ángeles. El lenguaje de los chicanos: regional and social characteristics of language used by Mexican Americans, ed. by Eduardo Hernández-Chávez, Andrew Cohen and Anthony Beltramo, 61-69. Arlington: Center for Applied Linguistics.

Lenneberg, Eric H. 1967. Biological foundations of language. New York: Wiley.

Lipski, John. 1987. Fonética y fonología del español de Honduras. Tegucigalpa: Editorial Guaymuras.

Lipksi, John. 1990. Elision of Spanish intervocalic /y/: toward a theoretical account. Hispania. 73: 797-804.

Lipski, John M. 1994. Latin American Spanish. New York: Longman Linguistics Library.

233

Lope Blanch, Juan M. 1990. El español hablado en el suroeste de los Estados Unidos. Materiales para su estudio. México: Universidad Nacional Autónoma de México.

Malberg, Bertil. 1947-1948. L’espagnol dans le nouveau monde. Problèmes de linguistique générale. Tirage à part de Studia Linguistica. I, II. Lund.

Maldonado-Dennis, Manuel. 1980. The emigration dialectic. Puerto Rico and the U.S.A. International Publishers.

Martinez, N. 2000. Leveling and koineization in Tijuana. 134. Arts (Spanish): SDSU.

Matluck, Joseph. 1963. La e trabada en la ciudad de México. Anuario de Letras. 3: 5- 34.

Mendieta, Eva and Molina, Isabel. 2000. Caracterización léxica del español hablado en el noroeste de Indiana. Southwest Journal of Linguistics. Vol.19.N.2. pp.63-72.

Milroy James and Lesley Milroy. 1999. Authority in language: investigating Standard English. London: Routledge.

Milroy, Lesley. 1987. Language and social networks. 2nd ed. Oxford: Blackwell.

Milroy, Lesley. 1992a. Linguistic variation and change. Oxford: Blackwell.

Milroy, Lesley. 1992b. Social network and prestige arguments in sociolinguistics. Sociolinguistics today, ed. by Kingsley Bolton and Helen Kwok, 146-162. Cambridge: Cambridge University Press.

Milroy, Lesley. 1995. Observing and analysing natural language. Massachusetts: Blackwell Publishers.

234

Milroy, Lesley and James Milroy. 1985. Linguistic change, social network and speaker innovation. Journal of Linguistics. 21. 339-384.

Milroy, Lesley and James Milroy. 1992. Social network and social class: toward an integrated sociolinguistic model. Language in society. 21. 1-26.

Mitchell, J. Clyde. 1969. Social networks in urban situations. Manchester: Manchester: Manchester University Press.

Navarro Tomás, Tomás. 1948. El español en Puerto Rico. Río Piedras: Editorial de la Universidad de Puerto Rico.

O’Brien, Robert W. 1954. A survey of the Puerto Ricans in Lorain, Ohio. Lorain: Neighborhood House Association of Lorain.

Odlin, Terence. 1997. Hiberno-English: Pidgin, Creole, or neither? CLCS Occasional Paper No. 49. Centre for Language and Communication Studies, Trinity College Dublin.

Ornstein, Jacob. 1974. Mexican American sociolinguistics: a well-kept scholarly and public secret. Sociolinguistics in the Southwest, ed. by Bates Hoffer and Jacob Ornstein, 91-121. San Antonio: Trinity University.

Oroz, Rodolfo. 1966. La lengua castellana en Chile. Santiago: Universidad de Chile.

Otheguy, Richard and Ana Celia Zentella. 2004. Contact and leveling in pronominal usage in six dialects of Spanish in New York City. Paper presented at NWAVE 33. Ann Arbor, Michigan.

Oyama, S. 1976. A sensitive period for the acquisition of a nonnative phonological system. Journal of Psycholinguistic Research. 5: 261-285.

Payne, Arvilla C. 1980. Factors controlling the acquisition of the Philadelphia dialect by out-of-state children. Locating language in time and space, ed. by William Labov. New York: Academic Press.

235

Pedraza, Pedro. 1985. Language maintenance among New York Puerto Ricans. Spanish language use and public life in the United States, ed. by Lucía Elías- Olivares, Elizabeth A. Leone, René Cisneros and John Gutiérrez, 59-72. Berlin: Mouton Publishers.

Penny, Ralph. 1995. A history of the Spanish language. New York: Cambridge University Press.

Peñalosa, Fernando. 1980. Chicano sociolinguistics: a brief introduction. MA: Newbury House Publishers.

Pfaff, Carol W. 1979. Constraints on language mixing: intrasentential code-switching and borrowing in Spanish/English. Language. 55: 291-318.

Poplack, Shana. 1980a. Deletion and disambiguation in Puerto Rican Spanish. Language. 56: 371-385.

Poplack, Shana. 1980b. Sometimes I’ll start a sentence in Spanish y termino en español: toward a typology of code-switching. Linguistics. 18: 581-618.

Poplack, Shana, David Sankoff and Christopher Miller. 1988. The social correlates and linguistic processes of lexical borrowing and assimilation. Linguistics. 26: 47- 104.

Post, Anita. 1934. Southern Arizona Spanish phonology. University of Arizona Bulletin. Humanities Bulletin. Vol.V. No.1.

Quilis, Antonio and María Vaquero. 1973. Realizaciones de /č/ en el area metropolitana de San Juan de Puerto Rico. Revista de Filología Española. 56: 1-52.

Ramos-Pellicia, Michelle F. 2000. A Puerto Rican Spanish retroflex approximant: some socio- and ethnolinguistic explanations to the issues raised by its use, or how in the world a retroflex got here?. Paper presented at Minority Languages in the Americas: The 29th Annual Meeting of the Linguistic Association of the Southwest. Benemérita Universidad Autónoma de Puebla.

236

Ramos-Pellicia, Michelle F. 2002. The phonology of Lorain Spanish: dialect convergence or dialect leveling? Paper presented at XIX Conferencia del español en los EEUU y el español en contacto con otras lenguas en el mundo Iberoamericano. Universidad de Puerto Rico, Río Piedras, Puerto Rico.

Rand, David and David Sankoff. 1990. GoldVarb version 2. A variable rule application for the Macintosh.

Ringer Uber, Diane. 1998. Narrative strategies and social identity in Puerto Rican Spanish. LACUS XXIV, ed by Sheila Embleton, 375-385.

Rivera-Mills, Susana V. 2001. Acculturation and communicative need: language shift in an ethnically diverse Hispanic community. Southwest Journal of Linguistics. Volume 20. Number 2. 211-223.

Ross, L. Ronald. 1980. La supresión de /y/ en el español chicano. Hispania. 63: 552- 554.

Saciuk, Bohdan. 1980. Estudio comparativo de las realizaciones fonéticas de /y/ en dos dialectos del Caribe hispánico. Dialectología hispanoamericana, estudios actuales, ed. by G. Scavnicky, 16-31 Washington: Georgetown University Press.

Sánchez, Rosaura. 1973 (1971) Nuestra circunstancia lingüística. Voices. Readings from El Grito. A Journal of Mexican American thought 1967-1973. Berkeley: Quinto Sol Publications. 420-449.

Sandfeld, Kristian. 1930. Linguistique Balkanique: problèmes et résultats. Paris: Librairie Ancienne Honoré Champion.

Sankoff, Gillian and Henrietta Cedergren. 1971. Some results of a sociolinguistic study of Montréal French. Linguistic diversity in Canadian society, ed. by: R. Darnell. Edmonton: Linguistic Research Inc.

Schaller, Helmut W. 1975. Die Balkansprachen: eine einführung in die Balkanphilologie. Heidelberg: Winter.

237

Schiffrin, Deborah. 1994. Discourse markers. Cambridge: Cambridge University Press.

Schreier, Daniel. 2003. Convergence and language shift in New Zealand: consonant cluster reduction in 19th Century Maori English. Journal of Sociolinguistics. 7/3: 378-391.

Selinger, H. 1978. Implications of a multiple critical periods hypothesis for second language learning. Second language acquisition research, ed. by W.C. Ritchie. New York: Academic Press.

Shuy, Roger, Walt Wolfram and William K. Riley. 1968. Field techniques in an urban language study. Washington, D.C.: Center for Applied Linguistics.

Silva-Corvalán, Carmen. 1989. Sociolingüística. Teoría y análisis. Madrid: Editorial Alhambra, S.A.

Silva-Corvalán, Carmen. 1994. Language contact and change: Spanish in Los Angeles. Oxford: Clarendon Press.

Silvestrini, Blanca G. and María Dolores Luque de Sánchez. 1988. Historia de Puerto Rico: trayectoria de un pueblo. San Juan, Puerto Rico: Editorial de la Biblioteca, Inc.

Steinsholt, Anders. 1962. Målbryting i Hedrum. Oslo: Universitetsforlaget.

Stevens, Kenneth and Sheila Blumstein. 1975. Quantal aspects of consonant production and perception: a study of retroflex consonants. Journal of Phonetics. 3: 215- 233.

Suárez, Víctor. 1945. El español que se habla en Yucatán. Mérida: Díaz Massa.

Thomason, Sarah G. 2001. Language contact. an introduction. Washington, D.C.: Georgetown University Press.

238

Thomason, Sarah G. and Terrence Kaufman. 1991. Language contact, creolization and genetic linguistics. Berkeley: University of California Press.

Torgersen, Eivind and Paul Kerswill. 2004. The processes of accent levelling and chain shifting in Southeastern British English.

Torres-Cacoullos, Rena and Fernanda Ferreira. 2000. Lexical frequency and voiced labiodental-bilabial variation in New Mexican Spanish. Southwest Journal of Linguistics. Vol.19. N.2. 1-17.

Toscano Mateus, Humberto. 1953. El español del Ecuador. Madrid: Consejo Superior de Investigación Científica.

Trudgill, Peter. 1974. The social differentiation of English in Norwich. Cambridge: Cambridge University Press.

Trudgill, Peter. 1983. Sociolinguistics: an introduction to language and society. England: Penguin Books.

Trudgill, Peter. 1986. Dialects in contact. New York: Blackwell.

Tuten, Donald N. 2003. Koineization in Medieval Spanish. Berlin, New York: Mouton de Gruyter.

Valdés, Dennis Nodín. 1991. Al norte: agricultural workers inthe Great Lakes Region, 1917-1970. Austin: University of Texas.

Vidal de Battini, Berta. 1949. El habla rural de San Luis. Buenos Aires: Instituto de Filología.

Watt, Dominic and Lesley Milroy. 1999. Patterns of variation and change in three Newcastle vowels: is this dialect levelling? Urban voices. Accent studies in the British Isles, ed. by Paul Foulkes and Gerard J. Docherty, 25-46. London: Arnold.

239

Williams, Ann and Paul Kerswill. 1999. Dialect levelling: change and continuity in Milton Keynes, Reading and Hull. Urban voices. Accent studies in the British Isles, ed. by Paul Foulkes and Gerard J. Docherty, 141-162. London: Arnold.

Winford, Donald. 2003. An introduction to contact linguistics. Malden, MA: Blackwell Publishers.

Wolfram, Walt. 1969. A sociolinguistic description of Detroit Negro Speech. Washington, D.C.: Center for Applied Linguistics.

Wolfram, Walt. 1973. Sociolinguistic aspects of assimilation. Arlington: Center for Applied Linguistics.

Western Interstate Commission for Higher Education. 2004. State of America 2004: Latino perspectives on the American agenda. Washington, D.C.: National Council of La Raza.

Young, Richard and Robert Bayley. 1996. VARBRUL analysis for second language acquisition research. Second language acquisition and linguistic variation, ed. by Robert Bayley and Dennis Preston, 253-306. Philadelphia: John Benjamins.

Zentella, Ana Celia. 1997. Growing up bilingual. Puerto Rican children in New York. Massachusetts: Blackwell Publishers.

240