Colombian Varieties of Spanish 9783954870196

Table of contents :
TABLE OF CONTENTS
PREFACE
COLOMBIAN SPANISH AT THE TURN OF THE 21ST CENTURY
THE “NEW” PALENQUERO: REVITALIZATION AND RE-CREOLIZATION
EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA: DOS RASGOS FONÉTICOS DE PRESUNTO SUSTRATO AFRICANO
LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION
SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI, COLOMBIA: ONE VARIABLE OR TWO?
ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’) STRUCTURE IN THE SPANISH OF BUCARAMANGA
A LABORATORY APPROACH TO S-LENITION IN THE SPANISH OF BARRANQUILLA, COLOMBIA
LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL DE LOS COLOMBIANOS DE BOGOTÁ Y MIAMI
SECOND PERSON SINGULAR PRONOUN USE DURING CLOSED ROLE PLAYS BY COLOMBIANS IN THE NORTH CAROLINA PIEDMONT TRIAD
THE EXPRESSION OF NOMINAL POSSESSION IN THE SPANISH OF COLOMBIANS IN NEW YORK CITY
REFERENCES
EDITORS
AUTHORS

Citation preview

MURIEL/OROZCO

31/5/12

14:24

Página 1

R i ch ard J . F i l e-M u ri el R afael Oro zco (ed s . )

Colombia n Va r ie tie s of Spa nish

MURIEL/OROZCO

31/5/12

14:24

Página 2

L INGÜÍS T IC A

IB ER OAM ER IC ANA Vo l . 5 0

DIRECTORES: MARIO BARRA JOVER, Université Paris VIII IGNACIO BOSQUE MUÑOZ, Universidad Complutense de Madrid ANTONIO BRIZ GÓMEZ, Universitat de València GUIOMAR CIAPUSCIO, Universidad de Buenos Aires CONCEPCIÓN COMPANY COMPANY, Universidad Nacional Autónoma de México STEVEN DWORKIN, University of Michigan ROLF EBERENZ, Université de Lausanne MARÍA TERESA FUENTES MORÁN, Universidad de Salamanca DANIEL JACOB, Universität Freiburg JOHANNES KABATEK, Eberhard-Karls-Universität Tübingen EMMA MARTINELL, Universitat de Barcelona JOSÉ G. MORENO DE ALBA, Universidad Nacional Autónoma de México RALPH PENNY, University of London REINHOLD WERNER, Universität Augsburg

MURIEL/OROZCO

31/5/12

14:24

Página 3

R i c h a r d J. F ile- Mur iel R a f ael Or ozco ( eds.)

Colombian Varieties of Spanish

Iberoamericana

•

Ve r v u e r t

•

2012

MURIEL/OROZCO

31/5/12

14:24

Página 4

Library of Congress Cataloging-in-Publication Data Colombian varieties of Spanish / Richard J. File-Muriel, Rafael Orozco (eds.). p. cm. – (Lingüística iberoamericana; 50.) Includes bibliographical references and index. ISBN 978-1-936353-07-1 (pbk.) – ISBN 978-3-86527-685-8 – ISBN 978-1-936353-07-1 1. Spanish language–Variation–Latin America. 2. Spanish language–Spoken Spanish–Latin America. I. File-Muriel, Richard J. II. Orozco, Rafael, 1959 PC4074.7.C65 2012 467’.9861–dc23 2012009679

Reservados todos los derechos © Iberoamericana, 2012 Amor de Dios, 1 – E-28014 Madrid Tel.: +34 91 429 35 22 Fax: +34 91 429 53 97 [email protected] www.ibero-americana.net © Vervuert, 2012 Elisabethenstr. 3-9 – D-60594 Frankfurt am Main Tel.: +49 69 597 46 17 Fax: +49 69 597 87 43 [email protected] www.ibero-americana.net ISBN 978-84-8489-631-9 (Iberoamericana) ISBN 978-3-86527-685-8 (Vervuert) ISBN 978-1-936353-07-1 (Iberoamericana Vervuert Publishing Corp.) e-ISBN 978-3-95487-019-6 Depósito Legal: Diseño de la cubierta: Carlos Zamora Impreso en España Este libro está impreso integramente en papel ecológico blanqueado sin cloro

MURIEL/OROZCO

31/5/12

14:24

Página 5

TA B L E O F C O N T E N T S

Armin Schwegler PREFACE .......................................................................................................

7

Rafael Orozco/Richard J. File-Muriel COLOMBIAN SPANISH AT THE TURN OF THE 21ST CENTURY ...........................

11

John M. Lipski THE “NEW” PALENQUERO: REVITALIZATION AND RE-CREOLIZATION ................

21

José Alejandro Correa EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA: DOS RASGOS FONÉTICOS DE PRESUNTO SUSTRATO AFRICANO .........................................................

43

Catherine E. Travis/Timothy Jowan Curnow LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION .....................

67

Earl K. Brown/Esther L. Brown SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI, COLOMBIA: ONE VARIABLE OR TWO? ...............................................................................

89

Catalina Méndez Vallejo ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’) STRUCTURE IN THE SPANISH OF BUCARAMANGA .......................................................................................

107

Richard J. File-Muriel A LABORATORY APPROACH TO S-LENITION IN THE SPANISH OF BARRANQUILLA, COLOMBIA ....................................................................................................

127

Luz Marcela Hurtado Cubillos LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL DE LOS COLOMBIANOS DE BOGOTÁ Y MIAMI .............................................................

141

Scott Lamanna SECOND PERSON SINGULAR PRONOUN USE DURING CLOSED ROLE PLAYS BY COLOMBIANS IN THE NORTH CAROLINA PIEDMONT TRIAD .............................

169

MURIEL/OROZCO

31/5/12

14:24

Página 6

Rafael Orozco THE EXPRESSION OF NOMINAL POSSESSION IN THE SPANISH OF COLOMBIANS IN NEW YORK CITY ......................................................................................

205

REFERENCES .................................................................................................

235

AUTHORS ......................................................................................................

261

MURIEL/OROZCO

31/5/12

14:24

Página 7

P R E FA C E ARMIN SCHWEGLER UC Irvine

The collection of fascinating articles assembled in this volume is a refreshing attempt to give renewed impetus to the study of a speech variety – Colombian Spanish – that in years past had been one of the most intensely studied in all of Latin America. Towards the end of the latter half of the 20th century, that successful tradition, it seemed, had become a weighty legacy from which most Colombianistas seemed unable to free themselves. As a result, linguistic studies in Colombia became overly conventional, thereby gradually losing the opportunity to profit from more theoretically oriented trends in modern linguistics. To be sure, shortly after the middle of the 20th century, prominent Colombian linguists (e.g., José Joaquín Montes Giraldo) had carried out extensive and wellinformed dialectological work that culminated in the ambitious Atlas lingüísticoetnográfico de Colombia (1981-1983). At the same time, the Instituto Caro y Cuervo was still actively promoting and publishing informative linguistic inquiries in a mostly philology-based tradition. And, with the able assistance of European colleagues, large-scale lexicological investigations led to ambitious publications that produced the much-celebrated Nuevo diccionario de colombianismos (Haensch/Werner 1993). Missing, however, were investigations that would, for instance, have taken advantage of the latest trends in Labovian-style approaches, or of the emerging field of pidgin and creole studies, which in the 1980s and 1990s increasingly attracted the attention of linguists in many parts of the world (Schwegler 2010). Within Colombian academia, other, more recent new directions in linguistics went equally unnoticed. Despite its unusually rich multilingualism, during the last quarter of the 20th century, Colombia had, for instance, been noticeably absent from theory-oriented discussions on languagecontact phenomena, studied so profitably by Thomason & Kaufman’s trailblazing Language Contact, Creolization, and Genetic Linguistics (1988). With few exceptions, the impetus for new and fresh approaches to the study of Colombian speech varieties thus seemed to come from scholars located outside of Colombia. This explains, for instance, why, some 40 years ago, Germán de Granda (Spain) and Derek Bickerton (USA) rather than a Colombian national were first to correctly identify Palenquero as a creole rather than simply as a “Spanish dialect”. By the same token, this also clarifies why the vast majority of publica-

MURIEL/OROZCO

8

31/5/12

14:24

Página 8

ARMIN SCHWEGLER

tions on the same creole have been from the pens of European or North American scholars. Since the beginning of this millennium, forceful attempts have been made to reconnect the study of Colombian Spanish and related speech varieties with contemporary linguistic theory. Key Colombian institutions are participating in this effort, including the Instituto Caro y Cuervo, whose forward-looking Director (Dr. Genoveva Iriarte Esguera) has expressed genuine interest in restoring the Institute’s former leadership role. Within this same post-2000 period, fascinating studies have delved into Colombian varieties from an array of different areas of linguistics, including dialectology, sociolinguistics, contact linguistics, syntax, phonology, morphology, and typology, often from an interdisciplinary approach. The fact that this volume has brought together twelve specialists (with widely differing theoretical orientations and preferences) on Colombian Spanish and/or Palenquero is firm and welcome evidence that the rejuvenation of investigations into Colombia’s speech varieties is well under way, and bearing fruits at a time when young(er) generations of scholars in and outside of Latin America are in need of inspiration and guidance. As readers will undoubtedly note, Colombian Varieties of Spanish has several features worthy of mention. First, and foremost, the volume is unique in that it brings together studies on Colombian Spanish that employ current theoretical approaches to linguistics, while at the same time addressing topics and varieties of Colombian Spanish that remain unexplored or understudied. As regards its overall conceptualization, the editors have aimed high by including articles that either extend far beyond the national borders of Colombia (see Hurtado Cubillos’s study of Colombian Spanish in Miami, or Orozco’s examination of nominal possession in the Spanish of Colombians in New York City), or transcend the narrow(er) confines of Spanish. The latter is the case with Lipski’s latest contribution on Palenquero, whose revitalization is occurring far more swiftly than anyone had imagined possible a decade or two ago. Some of the contributions in this volume are narrow in scope, but are nonetheless very useful for understanding regional variation from a synchronic as well as diachronic perspective. Commendable in this regard are Méndez Vallejo’s article on Focalizing Ser ‘to be’ in the Spanish of Bucaramanga, and Travis & Curnow’s study of the locational adverbs (aquí, allí, ahí, acá, allá) in the conversational Spanish of Cali. As Travis & Curnow demonstrate, the use of these frequent adverbs differs markedly from that traditionally described in the literature. This “deviant” behavior may help to explain in part why Palenquero adopted and subsequently changed these elements into all-purpose prepositions (cp. AÍ kasa = ‘IN/AT/BY/ON the house’). As such, the study – like all the others in this volume – invites further investigations into Colombian Spanish and/or

MURIEL/OROZCO

31/5/12

14:24

Página 9

PREFACE

9

Palenquero, and serves as excellent point of departure for theoretical considerations into language contact phenomena. In their introductory article, the editors (File-Muriel/Orozco) correctly point out that the study of Colombian Spanish in the United States is in its infancy. It would be wrong to conclude, however, that certain varieties of Spanish spoken within the national confines of Colombia are not similarly underexplored or unknown. One of the true merits of this book is that it informs readers not only of what is being done in the field of Colombian Spanish, but also of what is not being done. In this regard, immediate attention must be drawn to the Pacific Lowlands of Colombia, where slavery during the 18th century and the concomitant fluvial gold mining brought Spanish into contact with thousands of speakers of African languages, and probably also with an Afro-Hispanic pidgin similar to that once spoken in Cartagena (Schwegler, forthcoming). Correa’s detailed investigation in this volume concentrates opportunely on these Pacific Lowlands by examining two phonetic features of presumed African origin. As mentioned earlier, throughout most of the 20th century, sociolinguistic and/or variationist studies were largely ignored in Colombia. Moreover, partly because of Bogotá’s location in the Highlands, the Cachaco macro-dialect – which includes the Spanish spoken in the interior and in the Andean Highlands of Colombia – had received far greater attention from linguists than the remaining speech varieties. This led to a situation in which empirical variationist information on certain Lowland dialects was virtually non-existent until well after the year 2000. Brown & Brown’s study (this volume) on syllable-final and syllableinitial /s/ reduction in Cali constitutes a laudable attempt to redress this situation. Fortunately, this volume also supplements the aforementioned contribution with a very informative second study of s-lenition (by File-Muriel) that concentrates on the Spanish of coastal Barranquilla.

A. Schwegler & Carlos Patiño Rosselli (Bogotá, June 2010) PHOTO BY ALEJANDRO CORREA

This volume could not have been assembled without the enthusiastic collaboration of three generations of linguists from inside and outside of Colombia. All

MURIEL/OROZCO

10

31/5/12

14:24

Página 10

ARMIN SCHWEGLER

of us who have enjoyed studying the speech of this fascinating South American country have in one way or another benefited from the pioneering fieldwork and publications of Carlos Patiño Rosselli. An Honorary member of the Academia Colombiana de la Lengua and Emeritus professor of the Universidad Nacional de Colombia (Bogotá), he was the first to write a grammar in the Palenquero Creole (Patiño Rosselli 1984) while also tirelessly promoting the study of the indigenous languages of Colombia. Sadly, on June 15 of 2010, he passed away just a few days after he cheerfully attended my three-day seminar on Afro-Hispanic creoles (held at the Instituto Caro y Cuervo, Bogotá). It is to this beloved colleague that the editors, the contributing authors, and I jointly dedicate this volume.

MURIEL/OROZCO

31/5/12

14:24

Página 11

C O L O M B I A N S P A N I S H AT T H E T U R N O F T H E 2 1 S T C E N T U RY RAFAEL OROZCO/RICHARD J. FILE-MURIEL

Colombia, with a population of 46,300,000 has the second largest concentration of Spanish speakers in the world and the first in South America. According to Patiño Rosselli (1991: 148) Colombia’s linguistic situation is one of multilingualism, made up of three components: Colombian Spanish acting as the national language, approximately sixty indigenous languages that still survive, and two creole languages: Palenquero o Lengua, spoken in the village of Palenque1 (approximately 60 km Southeast of Cartagena) and Isleño spoken in the archipelago of San Andrés y Providencia. Today, there seems to be no consensus as to the name of the language spoken by most inhabitants of Colombia. To that effect, Villa Mejía (2001: 17) indicates that, 45% of respondents to a survey reported to speak Spanish, 40% Castilian, 10% American, and 5% Colombian. Colombian Spanish is often characterized in popular discussion as the “best spoken Spanish in Latin America” (Arango Cano 1994: 40), with many laypersons making (perhaps) unsubstantiated claims regarding the “neutral value” of the “most comprehensible” Spanish in Latin America. The sociolinguistic prestige of Bogotá is quite strong, and its speech patterns have traditionally been considered a model for educated Colombians to follow. In turn, this has helped maintain the prestige of Bogotano speech (Villa Mejía 2001: 29), which is spoken by the Colombian elite since Bogotá is home to Colombia’s highest status individuals. As with all other varieties of Latin American Spanish, Colombian Spanish shows indigenous as well as African influence (Zamora/Guitart 1982). The African influence, especially strong in the coastal regions, is attested in Palenquero, in the Caribbean region, and the Spanish of Chocó, on the Pacific

1

There is some variation regarding the actual name of this village: Schwegler, Morton, Escalante, Friedemann, and others consistently refer to it as “El Palenque de San Basilio,” whereas Del Castillo, Montes Giraldo, Navarrete, and Lipski refer to it as “San Basilio de Palenque.” Native Palenqueros represent it as both on their website (http://palenque desanbasilio.com) and have this to say about it: “Los abuelos no aceptan que se diga Palenque de San Basilio argumentan que el pueblo no es del santo sino el santo del pueblo, por eso se le de llamar San Basilio de Palenque y no Palenque de San Basilio, pero políticamente tiene mayor reconocimiento llamarle Palenque de San Basilio y así lo han hecho los académicos.”

MURIEL/OROZCO

12

31/5/12

14:24

Página 12

RAFAEL OROZCO/RICHARD J. FILE-MURIEL

coast. Perhaps these diverse influences inspired the publication of León Rey’s (1983) El breve diccionario de colombianismos, which compiles a sizable list of lexemes and expressions that are specific to Colombian Spanish. At the phonetic level, varieties of Colombian Spanish run from the Andean highland dialects, which are very conservative in terms of their pronunciation, to the coastal varieties, which feature drastic consonantal sound variations. The Colombian dialect regions mainly follow geographical criteria (Flórez 1961, Montes Giraldo 1982). Concurring with Henríquez Ureña (1921), Montes Giraldo (1982: 12) divides Colombia into two macro-dialect areas, a classification that also incorporates the main distinctions popularly made by most Colombians. One of these macro-dialects is Costeño, the word used to refer to the inhabitants of the Colombian coastal regions. The other has been called Cachaco (Orozco 2004: 30, 2009a: 97), after the word used by Costeños to refer to those from the Colombian interior including the Andean highlands. The two Colombian macro-dialects are mainly differentiated by a series of phonological distinctions. The occurrence of coda /s/ in the Cachaco dialect and its weakening, aspiration, or deletion in the Costeño dialect are seen as the main distinctions between them (Flórez 1961, Lipski 1994, Montes Giraldo 1982, Quesada Pacheco 2000, inter alia). Another major difference between the Colombian macro-dialects is the pronunciation of the other Spanish coda consonants [d, n, l, r]. Costeño, being Caribbean Spanish, shares the main phonological features found throughout the Caribbean (Canfield 1988, Zamora/Guitart 1982). For instance, besides coda /s/ being consistently aspirated or deleted, /n/ is velarized. The Cachaco macro-dialect is characterized by phonological conservatism and a predominantly Spanish-derived lexicon as well as the ustedeo, a preference for the second person singular pronoun usted (Quesada Pacheco 2000: 89). According to Lipski (1994: 213), the choice of familiar pronouns and their corresponding verb morphology is the most outstanding morphosyntactic variable in Colombian Spanish. Montes Giraldo (1982) points out that the use of the second person singular tú as the familiar pronoun is the form of choice on the Caribbean coast. On the Pacific coast the use of vos is widespread although tú is also heard. The rest of the country prefers some combination of usted and vos, but usted predominates and speakers of the Cachaco macro-dialect preferentially use usted in most contexts. Nevertheless, in recent decades tú has started to gain popularity among speakers of this variety. Additionally, Colombian Spanish is spoken in diasporic settings and is found in contact with English in North America. The first Colombian community in the US was formed in New York City almost a century ago (Orozco 2007a: 312). According to US Census figures, today Colombians constitute the largest segment of the population of South American origin in the United States, where the

MURIEL/OROZCO

31/5/12

14:24

Página 13

COLOMBIAN SPANISH AT THE TURN OF THE 21ST CENTURY

13

largest concentrations of Colombians are found in New York City, Miami, and Chicago, respectively. There are also sizeable numbers of people of Colombian origin in virtually every major North American city. This direct contact situation of rather recent inception provides a unique opportunity for short-term diachronic observation. Studies of this sort, as Weinreich proposes, “may make it possible to clarify basic problems involving longer time spans as well” (1967: 104). The contact with English and with many other varieties of Spanish make the Colombian Spanish spoken in the US fertile ground for the expansion of studies on Spanish in North America. The Colombian communities in the United States provide ideal opportunities for empirical exploration of the simultaneous effects of language contact and dialect leveling on an immigrant population. The status of Colombian Spanish as a minority variety within larger minority language communities makes it even more interesting since, as Lamanna (this volume) indicates, Colombians do not constitute the largest Hispanic group in any of the major North American Spanish-speaking conglomerates. Colombian Spanish has been studied by linguists for over a century. Rufino José Cuervo’s Apuntaciones críticas sobre el lenguaje bogotano published in 1872 constitutes the first linguistic study of any variety of Colombian Spanish. However, not much was done until the founding of the Instituto Caro y Cuervo (ICC) in 1942 served to restart linguistic studies in Colombia. The impact of this institution was so important that most of the work on Colombian Spanish completed during the 20th century was produced by members of the ICC dialectology department (Montes Giraldo 1995: 137). As a result of the research carried out there, Colombia is one of the most thoroughly studied Latin American countries in terms of dialectology (Lipski 1994: 204). In fact, the bulk of 20th century linguistic research on Colombian Spanish was mainly devoted to contributions to the Atlas lingüistíco-etnográfico de Colombia (ALEC), directed by Luis Flórez and published in six volumes between 1981 and 1983. The ALEC was the first such project completed in a Latin American nation (Montes Giraldo 1995: 78). The ICC has also played a pivotal role as the main outlet for research on Colombian Spanish through the publication of its journal, Thesaurus, and the dozens of volumes published as part of the series entitled Biblioteca de Publicaciones del Instituto Caro y Cuervo. Perhaps one of the most fruitful areas of investigation in the Colombian territory has been in the area of pidgin and creole studies (i.e. creolistics). As already noted, Colombia has two documented “creole” languages, Palenquero or Lengua, spoken in the village of Palenque de San Basilio and Isleño spoken in the archipelago of San Andrés y Providencia. Patiño Rosselli (2002) and Schwegler (2011b) provide a thorough treatment of the historical and social conditions that gave birth to these languages, namely, the multiethnic language contact that

MURIEL/OROZCO

14

31/5/12

14:24

Página 14

RAFAEL OROZCO/RICHARD J. FILE-MURIEL

took place during the forced transplantation of hundreds of thousands of peoples of African origin. The majority of research on the Colombian creole languages has focused on the former (Palenquero), perhaps due to its unique status of being only one of two Spanish-based creoles in all of Latin America (cf. McWhorter 1995). Indeed, Palenquero has attracted the attention of numerous scholars, including Ochoa Franco (1945), Escalante (1954, 1988), Montes Giraldo (1962), Granda (1968, 1978), Bickerton/Escalante (1970), Del Castillo (1982, 1984), Friedemann/Patiño Rosselli (1983), Patiño Rosselli (1983, 1999), Megenney (1986), Schwegler (1989, 1990, 2010), Davis (1993, 2000), Piñeros (2003), Morton (2005), and Moñino (2002, 2003, 2005), Moñino/Schwegler (2002) among others. These scholars have examined a diverse array of issues from a variety of different perspectives, including the language bioprogram hypothesis (i.e. that creole languages are formed when children use their innate language endowment to transform a highly-unstructured language into a fully-fledged language), monogenesis (i.e. that creole languages are all derived from a once wide-spread Portuguesebased pidgin), and the substratist hypothesis, which sees the similarities among creole languages as attributable to the African substrate languages. With reference to African influence, Hualde/Schwegler (2008) delve into one of the least understood aspects of Palenquero: its intonational system. They argue, convincingly, that at some point in the history of Palenquero, its prosodic system was interpreted as involving lexical tone. This finding contradicts a strong reading of McWhorter’s (2005) claim regarding lack of tone in the creole prototype. Equally interesting research has been carried out on the language contact situation in the islands of San Andrés and Providencia. Patiño Rosselli (1986) and Bartens (2002, 2003, 2009) have conducted several comparative studies looking at several lesser-known varieties of Western Caribbean Creole English, including Isleño. Bartens (2002) provides a clear picture of the Spanish-Isleño contact situation. Although English-lexified Isleño is in contact with Spanish as the prestige language, there is no stable diglossia of any kind on the Colombian islands, where a shift to Spanish is under way. She notes that most of the population is bilingual, code-switching is minimal, and most creole influence on the local Spanish is eliminated during schooling. Furthermore, she reports massive Spanish calquing in the creole. Unlike other varieties of creole English, which are in contact with Standard English (Jamaica, Belize, and Costa Rica) due to the demographic makeup of the tourists, Isleño sustains heavy contact with continental Colombian Spanish. Bartens (2003) provides an interesting overview of specific structures of Isleño grammar followed by descriptions of corresponding structures in Caribbean Standard English (CSE) and Spanish. She looks at each word class in turn, with attention to their sentential role and phrasal function. In

MURIEL/OROZCO

31/5/12

14:24

Página 15

COLOMBIAN SPANISH AT THE TURN OF THE 21ST CENTURY

15

addition to three future markers and one past marker, Isleño verbs may occur without any temporal marking in unambiguous contexts. In the appendix, Bartens provides a list of purported lexemes of African origin with their suggested etymologies. In addition to the valuable contributions to Colombian dialectology provided by Flórez (1950, 1951, 1961) and Montes Giraldo (1962, 1974, 1982, 1985), there have been a number of studies looking at different phonological processes in Colombian varieties of Spanish. Several researchers provide new insights at perhaps the most contemplated phonological problem in Hispanic Linguistics: the weakening (i.e. aspiration and deletion) of the implosive /s/. Lafford (1986) looks at the role of socioeconomic status, age, and gender in Cartagena; FileMuriel (2007, 2009) looks at the role of lexical frequency in Barranquilla, and Brown (2009) also looks at the role of lexical frequency in four different s-weakening dialects, including that of Cali. In their 2010 study, File-Muriel/Brown depart from the traditional transcription approach and quantify s-lenition in terms of three acoustic measurements, including duration, centroid, and voicelessness. Besides syllable-final s-reduction, Correa Ramirez (1990) examines the reduction of stops in consonant clusters in a rural Spanish variety spoken in Antioquia. In addition to the fascinating work on spoken Spanish varieties in Colombia, there have also been a handful of studies carried out in the area of sign language. These studies are particularly interesting because they shed light on a speech community that has received very little scholarly attention. Colombian Sign Language (LSC) is used by an increasing number of deaf people in Colombia due to its implementation in bilingual/bicultural education programs throughout the nation (cf. Tovar 2006). Gómez (1999) looks at the LSC community from a phonological perspective, providing a description of the movement-hold model for the phonetic and phonemic representation of sign structure. She provides a useful overview of the basic sign structures and major phonological processes found in LSC. Despite extensive research on regional variation in Colombian Spanish, which has been largely devoted to the Cachaco macro-dialect, empirical variationist studies are a relatively recent development (Orozco 2009: 96). Consequently, the assertion that “there is very little data available for coastal subvarieties” (Placencia 2007: 86) applies to studies on politeness as well as to research on other areas of linguistics such as language attitudes (cf. Castellanos 1980). By the same token, variationist studies of Colombian Spanish continue to be scarce (Orozco 2010a: 196). The chapters by File-Muriel, Lamanna, Hurtado and Orozco, respectively, in this volume add to recent variationist work by Orozco (2004, 2005, 2009, 2010a), File-Muriel (2007, 2009, 2010), and Orozco/Guy (2008) that has started to fill the existing void.

MURIEL/OROZCO

16

31/5/12

14:24

Página 16

RAFAEL OROZCO/RICHARD J. FILE-MURIEL

The study of Colombian Spanish in the US is also in its infancy. This volume aims to contribute to the emerging body of sociolinguistic literature on Colombian Spanish in general, and specifically on Colombian Spanish in the US. Previous research is limited to communities in Florida and New York State. The earliest studies include Hurtado’s (2001) analysis of variable pronominal usage among Colombians and Colombian Americans in Miami-Dade County, Florida, Ramírez’s (2003) study of pronominal expression in impersonal sentences among Colombians in New York State, and Orozco’s (2004) preliminary analysis of the future and the possessive in the Colombian community in New York City. Those studies have been followed by research further exploring those issues (Hurtado 2005a, 2005b; Orozco 2007a, 2007b; Ramírez 2007a). Additionally, Orozco (2010b) explores subject personal pronoun usage by Colombians in New York City, and Montoya (2010) analyzes the expression of possession by Colombians and Latinos of various other backgrounds in New York State. The studies in this volume by Hurtado, Lamanna, and Orozco, respectively, constitute valuable additions to the study of Colombian Spanish in the US. The immense research possibilities that Colombian Spanish offers have given rise to an increasing number of studies that extends beyond the national borders and that has been shared by linguists from Colombia and abroad, mainly the United States. This volume touches on the great linguistic diversity found in Colombia, bringing together a cadre of “Colombianistas.” Although our goal is to foster the study of Spanish varieties that have not received sufficient attention, we have not intended to produce an all-inclusive volume, which would cover every aspect of Colombian Spanish. Indeed, to suggest this as a possibility would be to underestimate the great diversity that exists within the Colombian territory. The contributors to this volume are only a few of the scholars currently conducting exciting research on Colombian Spanish. In fact, within the past ten years alone, there have been a score of fascinating studies looking at Colombian varieties from an array of different areas of linguistics, including dialectology, sociolinguistics, contact linguistics, syntax, phonology, morphology, and typology, often from an interdisciplinary approach. It is our hope that the reader of this volume will not only learn from the work that is being done, but also from what is not being done. We truly hope that this volume will inspire linguists, whether they have worked on Colombian Spanish or not, to carry out further research in this still under investigated nation. The idea of producing a volume devoted exclusively to Colombian varieties of Spanish is not completely novel; in fact, the ICC has published two monographs authored by José Joaquín Montes Giraldo, which address Colombian Spanish dialectology from an ethnographic perspective: Estudios sobre el español de Colombia (1985) and Otros estudios sobre el español de Colombia

MURIEL/OROZCO

31/5/12

14:24

Página 17

COLOMBIAN SPANISH AT THE TURN OF THE 21ST CENTURY

17

(2000). Additionally, the ICC has published monographs by Luis Flórez and several others. However, this volume strives to take a fresh look at Colombian Spanish at the turn of the 21st century from perspectives not previously used. This volume is unique in that it represents a compilation of studies of Colombian Spanish that employ current theoretical approaches to linguistics, as well as addressing both topics and varieties of Colombian Spanish that remain unexplored or understudied. As the Table of Contents shows, the remainder of this volume is devoted to three main themes. The next two chapters deal with contact varieties in Colombia. Chapters 4-7 deal with mainland regional varieties. Chapters 8-10 focus on the third main theme in this volume: Colombian Spanish in the United States. The first main theme of the remainder of this volume looks at language contact varieties in Colombia. The first chapter focuses on Palenquero, which is one of the few remaining Spanish-based creole languages in the world. John Lipski examines the unique language contact and acquisition situation of Palenquero, which has been caused by recent language revitalization efforts to teach Palenquero in Palenque’s schools. Lipski notes that most young speakers acquire Palenquero as a second language and that many of them are starting to speak it outside of the school environment as an affirmation of ethnic pride. Thus, this new generation’s version of Palenquero represents a unique opportunity to test models of post-creole continua. Young Palenquero speakers are in effect acquiring a language that is morphosyntactically a proper subset of Spanish (lacking nominal and verbal inflection), and in doing so are creating a variety of Palenquero, which, like creole and semi creole languages, is the result of incomplete trans-generational transmission. Lipski’s study analyzes both morphosyntactic carryovers from Spanish (e.g. lingering feminine gender concord) and innovations (e.g. the transformation of ma from a plural marker to a definiteness marker). He also discusses a partial lexical split, between words cognate with Spanish (which show a greater tendency to exhibit some morphological inflection) and Palenquero words with little or no resemblance to Spanish, which are more resistant to inflection. By explaining both the “undoing” of Ibero-Romance inflection systems and the emergence of new determiner configurations, Lipski sheds additional light on the role of functional categories in second language acquisition and creolization. Continuing with the theme of African influence in Colombian varieties of Spanish, José Alejandro Correa explores two features of African origin in the Spanish spoken in the Pacific Lowlands, which until now had been presumed to be attributable to African influence (cf. Montes 1974, Granda 1988, Lipski 2007). The intervocalic phones [ð
] and [ɾ] and the realization of the Spanish voiceless velar stop [k] (realized as a glottal stop [ʔ]) in particular had been tied

MURIEL/OROZCO

18

31/5/12

14:24

Página 18

RAFAEL OROZCO/RICHARD J. FILE-MURIEL

to substratal influence. This article provides a new and refreshing look at these earlier proposals. Correa provides phonetic analyses of spontaneous speech to enrich and deepen his discussion. The author presents compelling, synchronic as well as diachronic, evidence which suggests that (1) the variation between [ð
] and [ɾ] also involves the coronal consonants [r] and [l], and (2) the variable articulatory phenomena in question may well have been diffused by African languages once spoken in the Pacific Lowlands. Furthermore, the author presents strong evidence that the development of the glottal stop represents a regular phonetic change, and, contrary to the [ð
~ ɾ ~ r ~ l] variation, it is not substrate driven. The next set of articles is devoted to mainland regional varieties. Chapter Four, by Catherine Travis and Timothy Curnow, is the first paper in this part. This article constitutes an analysis of the ‘locational’ adverbs aquí, acá, ahí, allí and allá, as used in a corpus of conversational Caleño Spanish. The authors address both spatial and non-spatial uses of these adverbs, focusing their discussion of the non-spatial uses primarily on ahí, as close to 40 percent of its tokens in the corpus are non-spatial, including what they call ‘situational’, ‘temporal’, ‘approximative’ and ‘emotive’ uses. Travis and Curnow draw meaningful comparisons with the use of these adverbs in other dialects and genres based on Sedano’s work on spoken Venezuelan Spanish (1994, 1996, 1999), and Richardson’s work on novels by Spanish authors (1996). The results reveal that traditional analyses of locational adverbs fail to fully account for their use, in some cases because of dialect specific features and in others because of the nature of interactive conversation. Chapter Five provides an analysis of the aspiration of syllable- and word-initial /s/ in Cali. The authors, Earl K. Brown and Esther L. Brown, investigate the aspiration of syllable- and word-initial /s/ (for example, la señora > la [h]eñora) in the spontaneous speech of 25 speakers from Cali. The authors compare their results with those of previous studies of syllable- and word-initial /s/ aspiration in other dialects of Spanish (such as Brown, Esther 2005), as well to the results from the study of syllable- and word-final /s/ reduction (aspiration and deletion) in these same data from Cali (Brown, Earl 2008). This analysis contributes to the ongoing discussion in the literature about the possibility that syllable- and wordinitial /s/ aspiration is merely a continuation of syllable- and word-final /s/ reduction as suggested by Lipski (1999) and Brown, Esther (2004). In Chapter Six Catalina Méndez Vallejo explores the syntax of the Focalizing Ser structure in the Spanish of Bucaramanga. This structure has been attested in only a few varieties of Spanish (Venezuelan, Colombian, Ecuadorian, Dominican, and Panamanian), and is not stigmatized, despite being dialectally marked. Although the FS shows some resemblance to the pseudo-cleft construction, Méndez Vallejo’s syntactic analysis reveals that the two forms are syntactically dif-

MURIEL/OROZCO

31/5/12

14:24

Página 19

COLOMBIAN SPANISH AT THE TURN OF THE 21ST CENTURY

19

ferent. The author presents compelling evidence suggesting that the FS is a functional projection generated below TP and above vP, and that the FS is present in all varieties of Spanish. In Chapter Seven, Richard J. File-Muriel takes us back to the Caribbean region. He examines the production of the sound “s” in the Spanish of Barranquilla. With very few exceptions (cf. File-Muriel/Brown 2010, 2011; Erker 2010), previous studies looking at s-weakening have relied almost exclusively on impressionistic categorizations of “s” as [s], [h], etc. In the present study, the author employs CPU-assisted measurements (developed in File-Muriel/Brown 2010) in order to capture three acoustic properties of the sound “s:” duration, central tendency of the spectrum, and voicing. Linear regressions reveal that sproduction is significantly conditioned by speaking rate, surrounding sounds, lexical frequency, gender and the socioeconomic class of the speaker. FileMuriel discusses the advantages of adopting CPU-assisted measurements in lieu of symbolic representation, as temporal and gradient acoustic details about the sound are concealed when tokens are represented symbolically. The third main theme explored in this volume takes us beyond the territory of Colombia. It consists of three variationist papers on Colombian Spanish in the United States. Chapter Eight, by Luz Marcela Hurtado deals with impersonal pronouns in the Spanish of Bogotá and that of the Colombian community in Miami-Dade County, Florida. Hurtado’s analysis reveals that the most frequently occurring impersonal pronoun is uno. She also shows that the use of impersonal pronouns is simultaneously conditioned by a multiplicity of factors rather than just one specific social constraint. As occurs in the Colombian Community in New York (Orozco, this volume) the Spanish of Colombians in Miami-Dade County appears to reflect dialect leveling. Hurtado’s study indicates the importance of the context of interaction as well as the influence of language and dialect contact. It also constitutes an important foundation for the study of impersonal pronouns in other Hispanic communities. In the second chapter on Colombian Spanish in the US, Scott Lamanna, takes us to one of the newest Colombian communities in North America. He examines the use of second person singular pronominal address forms (tú and usted) in Bogotá as it compares to that of Bogotanos who reside in the, Mexican-dominated, Hispanic community of the North Carolina Piedmont Triad. Lamanna’s findings indicate North Carolina Colombians use tú less frequently than both those in Bogotá and North Carolina Mexicans. This study also reveals that dialect contact can influence linguistic phenomena below the individual speaker’s level of consciousness without influencing other behavior. Lamanna’s findings further tell us about the effects of dialect contact on the choice of second person singular pronoun and open up interesting possibilities for future research.

MURIEL/OROZCO

20

31/5/12

14:24

Página 20

RAFAEL OROZCO/RICHARD J. FILE-MURIEL

The concluding chapter takes us to the Colombian enclave in New York City, the community most geographically remote from Colombia studied in this volume. Rafael Orozco uses data from the Corpus del Español Colombiano en Nueva York to study the virtually unexplored linguistic variable used to express nominal possession in Spanish. This study reveals that possessive periphrases occur more frequently in New York than in Barranquilla. It also reveals that the expression of possession is conditioned by a number of linguistic and social constraints including semantic category and speaker’s sex. Interestingly, the linguistic conditioning on the expression of possession is largely the same as that found in Colombia (Orozco 2010a), notwithstanding the effects of direct contact with English. In general terms, this study shows how language and dialect contact simultaneously affect the Spanish of New York Colombians. Orozco’s results help increase our understanding of variation in contemporary Spanish and of how the sociolinguistic forces constraining language variation in Colombian Spanish conform to or depart from established sociolinguistic theory. Despite recent research, including the papers in this volume, and despite constituting the largest segment of the population of South American origin in North America, the linguistic situation of expatriate Colombian communities continues to be under investigated. As Orozco (2004: 60) affirms, the future of Spanish in Colombia depends largely on the country’s social and demographic conditions. As population continues to increase in the southern and Amazonian regions, one inevitable outcome would be the genesis and evolution of new dialects, perhaps influenced by the surviving indigenous languages. Concurrently, if the internal migrations that Colombia experienced during the latter part of the 20th century are an indication of increased mobility, they stand to further impact the instances of ongoing variation in Colombian Spanish. Additionally, the continued emigration of Colombians to other Latin American nations, as well as to the United States, Canada and Europe will most likely result in the imminent formation of other Colombian communities abroad. All of these factors will undoubtedly contribute to opening exciting lines of research for scholars in linguistics and other disciplines. We wish to thank our colleagues who provided us with valuable feedback on various aspects of this volume, including: Dary Marcela Ángel, Iris Bachman, Hugh Buckingham, Alicia Cipria, J. Clancy Clements, Concepción DeGodev, Manuel Díaz-Campos, Jeremy King, Edwin Lamboy, Tom Morton, Rafael Nuñez-Cedeño, Richard Ogden, Daniel Olson, Alberto Pastor, Carmen Ruiz Sanchez, Agnes Ragone, Lotfi Sayahi, Armin Schwegler, Sandro Sessarego, and Erik Wills. Their insightful comments were valuable to the contributors as well as to the editors in enhancing the quality of this volume. They are absolved, of course, of responsibility for any shortcomings, which we fully assume ourselves.

MURIEL/OROZCO

31/5/12

14:24

Página 21

T H E “ N E W ” PA L E N Q U E R O : R E V I T A L I Z AT I O N A N D R E - C R E O L I Z AT I O N JOHN M. LIPSKI

1. Introduction1 One of the oldest surviving creole languages is spoken in the Afro-Colombian village of San Basilio de Palenque. The language, known as Palenquero by linguists and simply as lengua ‘(the) language’ by community residents, is a highly restructured Afro-Iberian contact language, strongly influenced by Kikongo and bearing unmistakable Portuguese elements as well as a lexicon substantially derived from Spanish. Community members adhere to a plausible but highly speculative account of the creation of the village, but accurate historical data are not easily obtainable. According to many popular accounts, in 1599 a group of African slaves in the Spanish port of Cartagena de Indias (now part of Colombia) revolted and fled to the partially forested interior some 60 km to the south. Their leader, Domingo Bioho “King Benko”, was apparently born in northwest Africa, in or near the contemporary nation of Guinea-Bissau, but judging by the characteristics of the Palenquero language, many of the founders were speakers of Central African Bantu languages, particularly Kikongo, and possibly also of the emerging Afro-Portuguese creole of São Tomé. In 1603, the Spanish government sued for peace, and in the following years Bioho continued to engage in anticolonial resistance, until he was captured and hanged in 1621. In the intervening years Bioho and his followers founded at least one Palenque or fortified village; the residents of San Basilio de Palenque believe that their community was the one founded by Bioho and his name has been inextricably linked with the village’s centuries old history of cultural resistance. According to historians, however, the origins of the contemporary community of San Basilio de Palenque cannot be so clearly delineated. There is some evidence that Bioho’s Palenque was not the one that has survived today but rather another site some distance away (Navarrete 2008: 22-23), one of many maroon communities that dotted the 1

An earlier version of this chapter was presented at the Hispanic Linguistics Symposium in San Juan, Puerto Rico on October 22, 2009. For comments on the oral and written versions I gratefully acknowledge Armin Schwegler, Luis Ortiz López, and several anonymous and very helpful reviewers. Any remaining rough edges are, alas, mine alone.

MURIEL/OROZCO

22

31/5/12

14:24

Página 22

JOHN M. LIPSKI

region from the 16th century through the 18th. Colombian historians generally converge on the second half of the 17th century as the probable period in which the Palenque de San Basilio came into existence (Morton 2005: 33), i.e. considerably after Bioho’s death. The first unequivocal mention of the Palenque de San Basilio (with its original name, San Miguel Arcángel) comes in a document relating a peace agreement made in 1713 (Escalante 1954: 229; Navarrete 2008: 155166; Schwegler 2011a), which is consistent with a founding date somewhere towards the end of the 17th century. In 1713 the Spanish bishop Antonio María Cassiani visited the Palenque de San Miguel Arcángel to grant this village (re-named San Basilio Magno) official recognition, and was presumably able to communicate with its residents in some variety of Spanish. A document dated 1772 stated that the Palenqueros speak with one another “un particular idioma en que á sus solas instruyen á los muchachos sinembargo de que cortan con mucha expedición el castellano de que generalmente usan” [a particular language that by themselves they teach to their children, as well as Spanish which they speak fluently] (Gutiérrez Azopardo 1980: 34)2. Ever since that time the community has been bilingual, with the Palenquero language deliberately maintained and taught to succeeding generations in a powerful affirmation of ethnic identity, as members of the “first free people of America” as the village has come to be known (Arrázola 1970). Today San Basilio de Palenque (the version of the name preferred by contemporary community leaders) is a village of more than 3500 inhabitants, most of whom are descendents of the original maroon slave population. Colombians first came to regard this village with pride rather than scorn when the Palenquero boxer “Kid Pambelé” (Antonio Cervantes) won a world championship in 1974, only seven years after the first (dirt) road was extended into the previously isolated community (Salcedo Ramos/Rodríguez 2005). The village suddenly became the scene of visits from journalists and politicians, and the fame (together with the friendship between the boxer and the son of Colombia’s president) resulted in the first electrical lines being extended into the village. In 2005 Palenque was declared Masterpiece of the Oral and Intangible Heritage of Humanity by UNESCO. The Colombian Ministry of Culture has declared San Basilio de Palenque part of the “spiritual patrimony” of Colombia, although the town is still without generalized running water or a sewer system. Several documentary films on Palenque have been produced and circulated in Colombia, and annual music and drum festivals draw even more media attention to the commu-

2 Although Morton (2005: 36) suggests that what the writer may have heard was SpanishPalenquero code-switching.

MURIEL/OROZCO

31/5/12

14:24

Página 23

THE “NEW” PALENQUERO: REVITALIZATION AND RE-CREOLIZATION

23

nity. The popular music form known as Champeta, inspired by Palenquero musicians, is very popular in coastal Colombia, which adds to the attractiveness of Palenquero culture. San Basilio de Palenque has become the site of considerable research interest, media coverage, and tourism; the village boasts a center for ethnotourism, a cultural center, and a community center, all adorned with slogans in Palenquero. Visitors from Colombia and abroad, individually and in groups, are welcomed upon arrival and are freely addressed in Palenquero as well as Spanish. The high school has Internet-equipped computers and several Palenqueros maintain web sites.

2. A profile of Palenquero language usage, past and present The existence of the Palenquero language, and its nature as a true creole rather than a non-standard variety of Spanish, was virtually unknown outside of the immediate environs of the village until the second half of the 20th century, and even those non-Palenqueros who had heard the language implicitly assumed that this was the “broken Spanish” stereotypically associated with black populations in the Americas (Lipski 1985b, 1999b, 2005). Accurate transcriptions of Palenquero phrases appeared in print as early as the anthropological study by Escalante (1954), a lengthy article in an obscure journal (later expanded in Escalante 1979), although Palenquero was not explicitly identified as a creole language. Even earlier snippets of Palenquero speech appear in the equally obscure Ochoa Franco (1945). When the field workers of the ALEC (Atlas Lingüístico-Etnográfico de Colombia) visited Palenque in 1959 as part of a nationwide dialect survey, they encountered and described only regional varieties of popular Spanish (Montes Giraldo 1962). This is apparently due to the fact that Palenqueros preferred to not reveal their local language to outsiders, several of whom were experienced dialectologists and would have immediately recognized a creole language if they had heard it. The field workers were aware of the existence of lengua but refer to this speech form – which is not described or analyzed – simply as “su dialecto local” [their local dialect] which is “notablemente diversa de la variedad general” [notably different from the general variety] (Montes Giraldo 1962: 447). Moreover, Montes Giraldo (1962: 447) affirmed that the Palenquero “dialect” “se va olvidando y [...] al menos entre los varones, apenas quedan quienes lo utilicen como único medio de comunicación” [is being forgotten and at least among the men, almost no one uses it as the only means of communication]. Montes Giraldo presents a few examples of Palenquero grammar, including the plural marker ma, postposed possessives (e.g. tatá mi ‘my father’), and the 3rd person singular pronoun ele, but still concludes (p.

MURIEL/OROZCO

24

31/5/12

14:24

Página 24

JOHN M. LIPSKI

450) that Palenquero speech “es un habla esencialmente española en la que se combinan algunos rasgos arcaicos [...] con la agudización y avanzadísimo desarrollo de numerosas tendencias vulgares, [...]” [essentially Spanish speech, combining archaic traits and acutely advanced popular tendencies], accelerated by the relative isolation of the village. Less than a decade later (in 1968) the creolist Derek Bickerton obtained field data in Palenque, and in collaboration with the anthropologist Escalante published the first article explicitly acknowledging the existence of the Palenquero Creole (Bickerton/Escalante 1970). Thereafter numerous Colombian and foreign linguists visited San Basilio de Palenque, which resulted in many monographs and articles. More importantly, it resulted in Palenqueros’ taking a fresh look at a traditional language that many younger residents felt was an embarrassing throwback, and an endangered language shunned by many Palenqueros has been revitalized and adopted as a proud symbol of ethnicity. Although the fruits of the language revitalization program can be observed today, the pathway to progress was not without obstacles. Many young learners were too embarrassed to publicly speak Palenquero, some residents felt that use of lengua by young men to young women was inappropriate, and older community members sometimes expressed disapproval or outright rejection of the use of the Palenquero language by young people. While the ethno-education program has a few detractors even today, all of the dozens of young Palenqueros interviewed by the present writer accept the importance of learning the rudiments of the ancestral language, even if only a subset of this group actually practices the language outside of school. The current vitality of the Palenquero ethno-education program and the enthusiasm with which many young people attempt to speak Palenquero is particularly striking in view of numerous predictions that the Palenquero language was on the verge of disappearance. Bickerton/Escalante (1970: 266), describing their impressions of a brief research trip conducted in 1968, echo Montes Giraldo’s (1962) observation of the decline of the Palenquero language, and affirm that “if there is no change in the material conditions of its speakers, or their attitude toward their environment, Palenquero may, despite all pressures on it, continue to survive.” Megenney (1986), based on fieldwork done in 1973-4, merely refers to Palenquero as a “post-creole language,” but does not describe it as an endangered language. The late Colombian anthropologist Nina Friedemann, who began her research in Palenque in 1976, indicated that “En la escuela de Palenque, los maestros no sabían que la manera como los niños hablaban en sus juegos y retozos no era un ‘mal castellano’ sino su idioma materno [...]” [In the Palenque school, the teachers did not know that the way that the children spoke during their games and other activities was not “bad Spanish” but rather their native language] (Friedemann/Patiño Rosselli 1983: 17). Friedemann’s research

MURIEL/OROZCO

31/5/12

14:24

Página 25

THE “NEW” PALENQUERO: REVITALIZATION AND RE-CREOLIZATION

25

collaborator in Palenque, the Colombian linguist Carlos Patiño noted that although the community was still bilingual, younger people tended to speak to one another increasingly only in Spanish (Friedemann/Patiño Rosselli 1983: 187-188). He also observed that Palenquero speakers were the object of ridicule and discrimination in the village schools and in neighboring communities, and that many Palenqueros felt that it was disadvantageous to maintain their creole language (pp. 189-190). Patiño Rosselli concludes (p. 191) that “Salvo la intervención de factores improbables aunque no imposibles –una reacción de la propia comunidad, una política oficial adecuada–, la finalización del ciclo histórico del lenguaje palenquero no parece estar muy lejos” [Barring the intervention of improbable but not impossible factors, such as a reaction from within the community, or an adequate official policy, the end of the historical cycle of the Palenquero language is not far off]. Schwegler (1996: v. 1, 42), describing the situation as of 1993, indicates that many young residents of Palenque did not even understand lengua, much less speak the language. An informal survey conducted among school children from 12 to 18 years of age revealed that only about half had some competence in Palenquero, and Schwegler speaks of the rapid decline of the language in San Basilio de Palenque. Moñino (2002: 228, fn. 2), based on extensive field work in Palenque conducted between 1994 and 1998, asserted that although classes in Palenquero were obligatory in elementary and secondary school, “los niños y adolescentes ya sólo tienen de ella un conocimiento pasivo y se limitan al uso de algunas oraciones que les sirven de emblema identitario más que de medio de comunicación” [children and adolescents now only have a passive knowledge of Palenquero and limit their use to some sentences that serve as identity markers rather than as a medium of communication]. A scant fifteen years after Schwegler’s and Moñino’s observations, however, the sociolinguistic profile of San Basilio de Palenque has changed considerably, as regards ethnic pride and community efforts at revitalizing the Palenquero language. The “improbable” events suggested by Patiño have in fact taken place in the village. Beginning slowly in 1992, and already producing some observable results by the time Morton (2005) collected data in 1998, the Colombian government’s etnoeducación, ‘ethnic education’ program, stimulated renewed interest in learning and speaking Palenquero. Morton (2005: 103f.) indicates that as of 1998 most conversations overheard in grammatically complete Palenquero (and not, e.g. just emblematic tag phrases) took place among residents over the age of 35. Pfleiderer (1998), collecting data at the same time, determined that only about 10% of school children were using the Palenquero language at home. A decade later, it is not uncommon to hear school children spontaneously addressing each other in Palenquero (although not always sustaining long conversa-

MURIEL/OROZCO

26

31/5/12

14:24

Página 26

JOHN M. LIPSKI

tions), and some young couples who themselves are Spanish-dominant make noteworthy efforts to speak Palenquero to their small children. The local elementary school and high school offer daily classes in Palenquero language and culture, beginning at the pre-school level, and students learn to read and write as well as speak the traditional community language. A small first reader (currently out of print) was developed, and a Palenquero dictionary, now in a second edition, was also produced. Since Palenquero language classes are obligatory, the number of young community members who have acquired at least a working knowledge of the Palenquero language has grown considerably, albeit not necessarily through the fully fluent acquisition of all nuances of Palenquero grammar. After all, some of the “children” referred to by Moñino now have children of their own, and at least some emblematic use of Palenquero in the home has become a de rigueur manifestation of ethnic pride for many community residents3. Since the revitalization of the Palenquero language is part of a communitywide effort, the language-planning introduced by teachers and other leaders appears to be producing tangible results, as witnessed by the fact that all school children now receive several years’ instruction in the Palenquero language. Although there is considerable optimism among the language teachers and community cultural activists, the implications for long-term language maintenance have yet to be determined. What can be objectively studied is Palenquero usage by young people whose knowledge of the language comes primarily from school classes and by traditional speakers who acquired the Palenquero language within the family. Data to be presented in the following sections reveal differences in usage patterns between these two groups of speakers, perhaps due to a partial gap in transmission represented by the generation in which the Palenquero language began a sharp decline. The rapidity with which attitudes towards the Palenquero language underwent a near total reversal (in little more than a decade) has resulted in a sociolinguistic profile in which Palenquero is spoken fluently and frequently by the village’s oldest residents, and is also used – with varying degrees of fluency but with considerable enthusiasm – by many children and young adults. In between is a partially “lost generation” of individuals who possess full passive competence in Palenquero and usually nearly complete active fluency, but who for much of their lives preferred to speak only Spanish. As a result, many children are receiving the majority of their linguistic input in Palenquero from teachers and peers, to a lesser degree from elderly residents 3

Frequently overheard phrases used by young parents to small children include miní aquí ‘come here,’ ¿bo a cuchá? ‘did you hear?’ and ¿onde bo tando? ‘where are you going?’ These phrases are known to nearly all Palenquero residents, and their use by young adults does not necessarily presuppose a more profound knowledge of the Palenquero language.

MURIEL/OROZCO

31/5/12

14:24

Página 27

THE “NEW” PALENQUERO: REVITALIZATION AND RE-CREOLIZATION

27

(who may not always speak extensively in Palenquero to children, but whose speech can be easily overheard throughout the village), and only sporadically from their parents. Such an environment, which represents a departure from the more usual vectors of cross-generational language transmission, offers the potential for the rapid emergence of linguistic innovation: the youngest speakers in effect re-create the grammar of Palenquero based on input that, while more robust than what is usually postulated for creole language formation, is relatively depleted in comparison with normal native language transmission. While not resulting in improved infrastructure for the town (there is still no indoor plumbing, stable water supply or paved streets, and dirt-floored thatchedroof huts still represent nearly half of the dwellings), the reaffirmation of Palenquero identity has produced dramatic results in the unabashed use of local linguistic patterns. Fieldwork conducted by the present researcher in 2008-2010 suggests that a major attitude reversal as regards the Palenquero language and Palenquero identity is resulting in a resurgence of local vernacular features in Spanish (e.g. the previously stigmatized intonational patterns or tono, cf. Hualde and Schwegler 2008), as well as in the increased instances in which at least some Palenquero language is employed. The following sections will present some of the characteristics of the “new” Palenquero as acquired and used by the community’s youngest residents. Included in the discussion are attempts to restore a more “pure” form of the Palenquero language, as well as innovations in young people’s Palenquero usage, innovations affecting the Palenquero article system, possessive system, copular verbs, and some prepositions. The emergence of partial gender concord is discussed in Lipski (2011), while the possible emergence of phonological tones in young speakers’ Palenquero is dealt with in Lipski (2010).

3. Back to the future: the move to “purify” Palenquero Not only are students and other young Palenquero residents enthusiastic about studying and using Palenquero, but under the tutelage of a handful of activist teachers, they strive to restore a “pure” traditional form of the language, stripping Palenquero of Spanish language accretions that had become incorporated into the speech of older generations. Already at the end of the 20th century Morton (2005: 58) observed that the traditional second-person plural pronoun enú (of Central African origin) was replacing the Spanish-derived utere (< Sp. ustedes), and the African-derived lungá ‘to die’ was displacing morí (< Sp. morir). Other “restored” forms are posá ‘house’ replacing casa, mai ‘mother’ replacing mamá, chitiá ‘to speak’ replacing jablá and conbesá, canatulé ‘hunger’ replacing jambre, ngubá ‘peanut’ replacing maní, burú replacing plata or dinero ‘money,’

MURIEL/OROZCO

28

31/5/12

14:24

Página 28

JOHN M. LIPSKI

bumbilo replacing basura ‘garbage,’ combilesa replacing amigo ‘friend,’ chepa replacing ropa/trapo ‘clothing,’ piangulí replacing ceddo or puecco ‘pig,’ vegá replacing año ‘year,’ and makaniá replacing trabajá ‘to work.’ This restoration of archaic and moribund forms felt to be less Spanish-like and more “authentically” Palenquero is not due to chance, but rather to the conscious efforts of a handful of activists, including the three or four main Palenquero language teachers. The general principle applied in the selection of Palenquero words to be introduced in school is to pick forms that differ as much from their Spanish counterparts as possible, even if the chosen item is infrequent or no longer in active use. Since these are the words learned and reinforced in school, children can be overheard using these same items, and not their more Spanish-like counterparts, when attempting to speak Palenquero spontaneously. Morton (2005: 58) discovered that some community residents were discussing the feasibility of officially coining new Palenquero replacements for patrimonial Spanish words, but such efforts have not moved forward in a concerted fashion. A few activists use the obviously concocted Etao Jundo for Estados Unidos ‘United States,’ but this option has not caught on among most young speakers. The recent appearance of a Palenquero dictionary (Cásseres Estrada 2005), a Palenquero grammar book (Simarra Reyes/Triviño Doval 2008) and a Palenquero commented glossary (Simarra Obeso et al. 2008), all emphasizing Palenquero words that differ substantially from their modern Spanish counterparts, reinforce the campaign to expunge as many Spanish words as possible from young speakers’ Palenquero. Older traditional Palenquero speakers generally regard these efforts with amusement, and there is no indication that these reintroduced Palenquero or African words are making any comeback among the community’s oldest speakers. A further indication of the impact of the reintroduced Palenquero lexical items among young speakers appears in the results of an experiment conducted in March, 2011. As part of a study of the perception of Spanish-Palenquero codeswitching, twenty-five Palenquero speakers (fifteen traditional speakers and ten adolescent speakers) were presented with a total of seventy recorded utterances gleaned from previous interviews conducted in the community. Some of the utterances were indisputably in Palenquero, a few were entirely in Spanish, and the majority contained at least some mixture of Palenquero grammar and Spanish grammar, and could thus be regarded as instances of intra-sentential codeswitching involving more than simple lexical insertions. Respondents were asked to listen to each of the sentences and decide whether the sentence was in Spanish, in Palenquero, or combined both languages. Young Palenquero speakers nearly always judged as mixed sentences containing Spanish-derived lexical items such as casa, jambre, and trabajo even when the remainder of the sentence was indisputably in Palenquero. Older traditional speakers judged such sen-

MURIEL/OROZCO

31/5/12

14:24

Página 29

THE “NEW” PALENQUERO: REVITALIZATION AND RE-CREOLIZATION

29

tences to be entirely in Palenquero. At the same time, young speakers’ hypersensitivity to targeted lexical items frequently resulted in a complete failure to notice patently Spanish grammatical constructions embedded in the sentences, including conjugated verbs, preverbal object clitics, definite articles, and feminine gender concord. {SENTENCES IN PALENQUERO; YOUNG SPEAKERS REJECT CASA INSTEAD OF POSÁ} ese casa poleba vendé nu poque ese casa era ri ma cuatro moná lo que i teneba cuné ‘that house couldn’t be sold because that house belonged to the four children I had with her’ antonce casa suto teneba Barranquilla casa quelá jue pa majaná ‘so the house we had in Barranquilla was left for the kids’ aquí teneba casa ri material nu ‘there were no cement block houses here’ {YOUNG SPEAKERS REJECT CASA INSTEAD OF POSÁ BUT FAIL TO NOTICE THE ENTIRE NOUN PHRASE LA PRIMERA CASA IN SPANISH} la primera casa de material lo que hacé aque fue casa ma tatá mi ‘the first cement block house belonged to my parents’ {YOUNG SPEAKERS REJECT CASA INSTEAD OF POSÁ BUT FAIL TO NOTICE THE ENTIRE CLAUSE TE CUENTO IN SPANISH} i asina te cuento que i acoddá ri ma nombre de to ma lo que asé miní casa mi nu ‘and so I’m telling you that I don’t remember the names of all the people who came to my house’ {YOUNG SPEAKERS REJECT JAMBRE INSTEAD OF CANATULÉ IN PALENQUERO SENTENCE} i taba cu mango p’i cumé poque i a tené jambre ‘I was holding some mangoes to eat because I was hungry’ {YOUNG SPEAKERS REJECT JAMBRE INSTEAD OF CANATULÉ IN PALENQUERO SENTENCE, BUT FAIL TO NOTICE SPANISH CONJUGATED VERB TENGA} e que tenga cumina tenga que ndá a que tenga jambre ‘those who have food have to give it to those who are hungry’ {YOUNG SPEAKERS REJECT PLATA INSTEAD OF BURÚ BUT FAIL TO NOTICE THE SPANISH VERB+CLITIC TE DAMO} si bo trae lágrima tio tigre aquí a ete vigrio lo que ta aquí suto te damo plata ‘Uncle Tiger, if you bring some tears in this glass we’ll give you money’ {YOUNG SPEAKERS REJECT TRABAJO INSTEAD OF MACANEO BUT FAIL TO NOTICE SPANISH NEGATION + VERB NO TIENE} pero si ané no tiene trabajo suto pasa trabajo aquí ‘But if they don’t have a job we have a hard time here’

MURIEL/OROZCO

31/5/12

14:24

Página 30

30

JOHN M. LIPSKI

{YOUNG SPEAKERS REJECT TRABAJO INSTEAD OF MACANEO BUT FAIL TO NOTICE SPANISH CLITIC + CONJUGATED VERB ME ABURRO AND SPANISH OTRA VEZ} de ahi me aburro trabajo ahi i ase miní camino pa Palengue otra vez ‘So there when I get bored with my job I come back to Palenque again’ {SENTENCE IN PALENQUERO; YOUNG SPEAKERS REJECT MORÍ INSTEAD OF LUNGÁ} i a teneba tre pero a murí ndo, a quelá mi ese ‘I had three [children], two died, I have that one left’ {SENTENCE IN PALENQUERO; YOUNG SPEAKERS REJECT TRAPO INSTEAD OF CHEPA} pa’o bucá trapo bo l’echá pelé cu hemano si ‘To get some clothes you fight with your brother’

4. Innovative morphology: the extension of ma In Palenquero, unlike in Spanish, nouns and adjectives are invariant for number. Plural marking in noun phrases, when it occurs at all, is effected by the preposed plural marker ma: ma jende ‘the people,’ ma ngombe ‘the cows.’ The marker ma presumably derives from one of the most frequent Bantu pluralizing prefixes, attested for Kikongo and other Central African languages known to have been present in the linguistic mix in which Palenquero was originally formed. Palenquero ma is more frequently used with definite reference, but can also be used with generic plurals, as in the potentially ma pelo asé ndrumí mucho ‘(the) dogs sleep a lot.’ The basic pattern of usage is shown in Table 1, taken from Schwegler (2007a: 62): TABLE 1 The basic Palenquero article system Definite or generic and mass nouns

Indefinite nouns

Singular

Ø

un

Plural

ma

un ma

Moñino (2007) offers the claim that ma and other African items were deliberately introduced by maroons “para despistar a los no miembros de su comunidad y no ser entendidos de ellos” [to sidetrack non community members so as to not

MURIEL/OROZCO

31/5/12

14:24

Página 31

THE “NEW” PALENQUERO: REVITALIZATION AND RE-CREOLIZATION

31

be understood by them]. Moñino asserts that combinations of ma and other determiners such as the indefinite ún and the demonstrative ése (e.g. ún ma óha bédde ‘some green leaves,’ ése ma tabáko ‘those cigarettes’) confirm that ma was originally inserted in order to create an “exotic” speech that could not be easily understood by non-initiates4. While agreeing with Moñino that Palenquero ma largely serves to disambiguate plural NPs and that superfluous ma may also occur in configurations in which plural is already indicated through other means, Schwegler (2007a: 69-70; 2007b: 216-217) argues that plural ma provides such a transparent and easily learned mechanism that it would not be effective as a means of disguising speech. Regardless of whether ma was once introduced into Palenquero in a deliberate attempt to make the language sound more exotic or to impede comprehension by outsiders, contemporary young learners of Palenquero have hit upon ma as a quintessential Palenquero element; they employ it not only for plural reference more frequently than fluent native speakers, but – in striking contrast to all previous generations of speakers – also as an unambiguously SINGULAR article. This is shown in Figure 1, extracted from a student homework assignment, which contains both the Palenquero text and the Spanish translation, in which ma is unmistakably shown to have singular reference5. In order to provide a more systematic view of young speakers’ usage in Palenquero, interviews were conducted with 25 young Palenquero speakers (15 male, 10 female), all of whom had learned Palenquero in school; some also had learned the rudiments of the language at home by listening to older relatives, but none spoke Palenquero routinely before attending school6. The ages of the intervie4 At the same time it can be noted that analytical combinations of determiners originally represented by a single inflected determiner are frequent in other creoles and semi-creoles, and do not necessarily suggest deliberate obfuscation by earlier generations. In traditional Afro-Bolivian Spanish, for example, the invariant demonstrative eje and invariant possessives such as mi ‘mine,’ nostu ‘our’ combine analytically with the plural definite article lu: eje lu mujé ( [ð
] se da solamente en comunidades de habla afrohispánicas, 2) el cambio /ɾ/ > [ð
] sólo se da en préstamos del portugués al kikongo y lenguas vecinas y 3) en las comunidades en las que se presenta el fenómeno hay tradiciones cuya denominación alude a la población de la cuenca del Congo. Finalmente, Schwegler (en preparación), ante la abrumadora influencia que parece haber ejercido el kikongo en el criollo palenquero, ha encontrado plausible la idea de un ancestro único que propone Lipski. Aunque las hipótesis no coinciden desde un punto de vista histórico ni lingüístico, el razonamiento básico parece ser el mismo. Primero, se argumenta que, en tanto la variación entre [ð
] y [ɾ] es característica del habla de poblaciones negras, el fenómeno es africano. Segundo, se apela a fuentes secundarias como testimonios literarios, datos demográficos y denominaciones culturales para poder imputar un ancestro africano propiamente dicho. Finalmente, la caracterización fonética de los sonidos estudiados se ve a la luz de las características de esta(s) lengua(s). Así, mientras Granda asocia la proximidad articulatoria entre la aproximante dental –fricativa dental en la terminología de la época4– y la vibrante simple al predominio del punto de articulación alveolar y posalveolar en la realización de /d/ en las lenguas del Golfo de Guinea, Lipski se centra el paso /ɾ/ > [ð
] (omitiendo el proceso inverso /d/ > [ɾ]) para encontrar una correlación directa entre sus ejemplos y algunos préstamos léxicos del portugués al kikongo. 4

Después de Martínez-Celdrán (2004), es claro que la /d/ intervocálica no es una fricativa, puesto que durante la pronunciación de esta consonante no se produce un flujo de aire turbulento.

MURIEL/OROZCO

31/5/12

14:24

Página 47

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

47

En una exploración de rasgos africanos en los criollos atlánticos, Parkvall (2000: 33) se ha referido al tema en otros términos: The Ibero-Romance distinction between a flap and a trill (or rather a flap and a fricative in modern standard Portuguese) is not maintained in any basilectal Creole […] Once having been merged with the flap, some variation between this [the trill] and the alveolar /d/ is only expected for reasons of articulatory proximity; indeed this is precisely what has happened in many varieties of Spanish and Portuguese spoken by descendants of Africans in the Atlantic area […] More interesting are cases of interchanges between /ɾ, r, d/ on the one hand and /l/ on the other. To be sure, many languages around the world lack a distinction between laterals and rhotics, but the distinction is generally maintained in Pidgins and Creoles if and only if it is found in the substrates […].

Esta aseveración tiene importantes implicaciones para efectos de nuestra discusión. Primero, relaciona diacrónicamente los fonemas /ɾ, d/ con otros sonidos coronales i. e., /r, l/. Segundo, se admite su origen africano en tanto la(s) lengua(s) de sustrato no presenten oposiciones entre estos elementos. La solución parece viable, pues el palenquero ha aprovechado este mecanismo en la formación de su estructura sonora. Según Patiño (1983: 93-98) en Palenquero /d/ tiende a realizarse como una vibrante simple en posición intervocálica y en posición inicial de palabra después de una vocal, por ej., no rebe parí má ‘no debe parir más’ [no ɾe∫e paɾi ma]5; simultáneamente, [ɾ] puede sustituir a una vibrante múltiple, por¯ej., barika ‘barriga’ [baɾika] y, finalmente, los tres sonidos involucrados /ɾ, r, d/ se resuelven a favor de una lateral [l], por ej., la kala ‘la cara’ [lakala], balendo ‘barriendo’ [balendo], ¿a kelá bien? ‘¿quedó bien?’ [a kela ∫jen]6. Pero aun en casos como el de este criollo, no considero plausible postu¯ un ancestro único (i.e., el kikongo) por dos razones. Primero, porque implicalar ría que los hablantes de esta lengua –aunque constituyeron una importante mayoría entre los esclavos traídos a América– predominaron en cada una de las zonas de en las que se encuentra el fonetismo, lo cual no resulta viable de acuerdo con la información histórica. Y segundo, como ha afirmado Schwegler (en preparación) a propósito del sustrato africano del palenquero, ante los vacíos históricos que limitan la investigación sustratista, deberíamos mantener una actitud prudente y estar dispuestos a aceptar la influencia de varias lenguas africanas. 5 La transcripción ortográfica de los ejemplos ha sido adaptada al Alfabeto Fonético Internacional. 6 Es importante aclarar que en palenquero esta convergencia es específica de algunos lexemas. Como anota Schwegler (en preparación) la palabra loyo ‘arroyo’ siempre es pronunciada [loo] pero nunca *[roo] o *[ɾoo], lo mismo ocurre con lexemas como kulasó ‘corazón’ o kala ‘cara’.

MURIEL/OROZCO

31/5/12

14:24

48

Página 48

JOSÉ ALEJANDRO CORREA

En este sentido, resulta de suma importancia comprobar que la variación articulatoria entre las consonantes /ɾ, r, d/ y /l/ es un rasgo que se encuentra bien difundido en el área lingüística africana. Por ejemplo, Ladefoged (1964: 29) anota que en algunas lenguas de la zona occidental de África no se presenta oposición fonológica entre /l/ y alguna vibrante como /ɾ, r, ɹ/, además, afirma que cuando tal oposición existe, /r/ se realiza como una vibrante simple [ɾ] o una vibrante aproximante [ɹ] (véase también el estudio de Lindau 1980). En la zona del África Central, la lengua bantú beembe, restringe la aparición de /d/ a la posición inicial de palabra, y /l/ se realiza como una vibrante simple frente a /i/ y entre vocales posteriores e. g., ‘piedra’ /tálì/ > [tháɾì], ‘valle’ /dú:lù/ > [dº ú:ɾù] (Jacquot 1962). De acuerdo con las pesquisas de Hyman (2003), originalmente la consonante *d del Proto – Bantú cambió a [l] o [r] en el contexto v_v (preservándose como [d] antes de [i]) en algunas lenguas como duala, tiene, bonbangi, kongo, luena, kwezo, entre otras. Concluimos que la variación fonética fue inducida simultáneamente por la estructura sonora de las muchas lenguas habladas por los africanos que llegaron a Colombia y a otras zonas de América durante los siglos XVII y XVIII. Nos enfrentamos ante un caso en el que un rasgo lingüístico característico de un área lingüística7 se transmitió al español hablado por los afrodescendientes de América (y a algunas lenguas criollas), y no un caso en el que una única lengua ha actuado como sustrato. Esta aseveración está apoyada por el hecho de que los esclavos africanos llevados al territorio estudiado provenían tanto de África occidental como de las costas atlántica y central del continente (Granda 1971, 1988; Pavy 1967 y Maya 1998). Bien, ¿en el español del Pacífico de Colombia ocurre una variación fonética entre las consonantes coronales semejante a la que ha sido reportada en el área lingüística africana? En §4.1 analizaremos auditiva e instrumentalmente una serie de muestras de habla espontánea tomadas en varias localidades de la región con el objeto de responder a esta pregunta.

2.2. LA OCLUSIÓN GLOTAL La oclusión glotal usualmente se define como una consonante producida por medio de una aducción o cierre completo de los pliegues vocales; sin embargo, 7

Thomason (2000: 311) define área lingüística en los siguientes términos: “a linguistic area is a geographical region containing a group of three or more languages that share some structural features as a result of contact rather than as a result of accident or inheritance from a common ancestor”.

MURIEL/OROZCO

31/5/12

14:24

Página 49

49

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

estudios recientes sobre el comportamiento laríngeo durante la producción de consonantes glotales, preglotalizadas y faríngeas (Esling/Fraser/Harris 2005: 386) han mostrado que la realización más común de una oclusiva glotal involucra una aducción de los aritenoides, una aducción completa de los pliegues vocales, una aducción parcial de los pliegues ventriculares y un estrechamiento del vestíbulo laríngeo. Este comportamiento canónico puede variar en el curso de la cadena hablada, de acuerdo con el contexto fonético, el hablante y la función comunicativa que cumpla la articulación laríngea. De esa manera, varias configuraciones de la glotis pueden dar lugar al estímulo que se percibe como una oclusión glotal. En un estudio realizado por Redi y Shattuck-Hufnagel (2001: 414) se identificaron, como correlatos acústicos de una oclusión glotal, descenso de la frecuencia fundamental (f0) o de la intensidad en contextos v_v, laringealización (creaky voice), sucesión irregular de pulsos glotales (aperiodicidad) y la alternancia de periodos glotales con la misma forma, duración o intensidad (diplofonía). En las lenguas del mundo, una oclusión glotal puede cumplir con varias funciones. Comúnmente ocurre como una unidad fonológica o como la realización fonética de una oclusiva. Por ejemplo, en algunos dialectos del inglés, /t/ se realiza como [ʔ] en palabras como beaten, kitten, fatten [biŸ…ʔn,kIʔn,fæʔn] (Ladefoged 2001: 48). Adicionalmente, la oclusión glotal tiene una fuerte tendencia a ocurrir en función de la estructura prosódica; en inglés puede demarcar el inicio de una frase entonativa (Dilley/Shattuck-Hufnagel 1995; Pierrehumbert/Talkin 1991), el límite entre dos frases entonativas o el final de un enunciado (Redi/ Shattuck-Hufnagel 2001). En alemán es más favorable que la glotalización ocurra antes de una vocal inicial que está entonacionalmente acentuada (Kohler 1999). La vocación demarcativa de la glotalización es tal, que en finlandés sirve incluso para que un hablante indique a su interlocutor la intención de mantener su turno en la conversación (Lennes/Aho/Toivola/Wahlberg 2006). La glotalización es común en algunos dialectos del español de América. Chela Flores (1986) menciona que en español de Maracaibo (Venezuela) la /d/ final de palabra pierde su componente oral (desbucalización) y se realiza como una glotal en palabras como verdad [beɾðaʔ] y usted [uhteʔ]. Según el autor, este fenómeno hace parte de un proceso de debilitamiento de las consonantes posnucleares que acercan la sílaba española a la estructura CV8. A este tenor, Valentín-Márquez (2006) presenta datos sociolingüísticos del español hablado en Puerto Rico, y muestra que la oclusión glotal ha comenzado a emerger en el |

|

|

8 Zamora/Guitart (1988: 109-119) presentan ejemplos similares y secundan esta línea interpretativa.

MURIEL/OROZCO

50

31/5/12

14:24

Página 50

JOSÉ ALEJANDRO CORREA

habla de los jóvenes (especialmente las mujeres) en contextos prevocálicos como realización de /s/ final, por ej., tres años [tɾeʔao]. Pero a diferencia de Chela Flores (1986) y Zamora/Guitart (1988), quienes aseguran con tino que el cambio está fonológicamente motivado, Valentín-Márquez considera que este cambio fonético es producto de un influencia de la lengua inglesa introducida “por la vía de las preferencias musicales de los jóvenes” (2006: 327). En la variedad hablada en la Costa Pacífica de colombiana, la oclusión glotal funciona como un alófono de la oclusiva velar sorda /k/, por ej., ‘le da como duro a uno’ [lea ʔomouɾo auno], ‘la casita amarilla’ [la ʔasita amaɾia]. Dentro de la región, el uso de esta articulación es privativa de los habitantes de zonas en las que se practica la minería del oro desde la colonia. Granda (1977: 94-127) propone que este rasgo proviene de las lenguas africanas que trajeron los esclavos a la región durante los siglos XVII-XVIII. Su argumentación se basa en una consideración detallada de los condicionamientos sociohistóricos que han actuado en la formación cultural de la región y, de otro lado, en la imposibilidad de explicar el origen del fenómeno a causa del desarrollo de tendencias internas del sistema fonológico del español, de influencias de adstrato o a causa de un desarrollo independiente. El autor no pasa por alto los requerimientos que, de acuerdo a los principios básicos de la lingüística histórica, deben tenerse en cuenta para afirmar que un rasgo lingüístico surgió de una influencia de sustrato determinada. No obstante, un vistazo a algunos de sus argumentos permite poner en duda esa hipótesis. El primer punto problemático en la argumentación del autor se desprende de la posibilidad de que el fenómeno sea producto de un desarrollo independiente, esto es, de la puesta en marcha de principios regulares del cambio fonético. Con respecto a este punto, Granda afirma: Del mismo modo, aunque más matizadamente, me parece rechazable (al menos en el momento actual y en el presente estado de conocimientos) la suposición de la génesis del proceso fonético [k] > oclusiva glotal en Guapi e Iscuandé por una evolución local, independiente, no condicionada por factores externos (entre los cuales se incluirían también, las tendencias generales hispánicas) y basada en posibles, oscuras y no identificadas propensiones articulatorias regionales (1997: 118; las cursivas son mías).

Lo problemático de la afirmación anterior es que no tiene en cuenta cambios lingüísticos paralelos en lenguas no relacionadas, es decir, no se explora a fondo la posibilidad de que el paso /k/> [ʔ] sea un cambio recurrente en el desarrollo histórico de diferentes lenguas, motivado, lingüísticamente, por la evidente proximidad perceptual que existe entre una oclusiva velar sorda y una oclusiva glotal, lo cual, convendrá el lector, no necesariamente implica apelar a “oscuras y no identificadas propensiones articulatorias regionales”. Por ejemplo, Blust

MURIEL/OROZCO

31/5/12

14:24

Página 51

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

51

(2004: 366) reporta que el cambio */t/ > /k/ se dio en las lenguas polinesias hawaiano, samoano, tahitiano y luanguiano simultáneamente al cambio */k/ > /ʔ/, y Donohue (2002: 198) nota que el fonema */k/ del proto-skou cambió a /ʔ/ en las lenguas wutung, dumo y dusur. Pero la posibilidad de que el fenómeno que discutimos sea producto de una tendencia regular del cambio fónico no quedaría abierta a exploración si las lenguas traídas por los esclavos africanos manifestaran en la actualidad este rasgo con suficiente claridad o, cuando menos, si fuera recurrente en algunas de las hablas criollas atlánticas. Pero ninguna de estas dos alternativas es sustentada por los datos obtenidos. Como vimos, de acuerdo a los datos históricos, los esclavos que arribaron a la región que estudiamos provenían del Golfo de Guinea y el África centro-occidental (Pavy 1967; Granda 1988), y los apelativos asignados a los esclavos durante la colonia sugieren una alta presencia de hablantes de las familias Kwa, Gur, Mande, Kru, entre otras. De acuerdo con Ladefoged (1964), la oclusiva glotal /ʔ/ no parece ser un fonema recurrente dentro de las lenguas del África occidental, y su aparición se restringe –en las lenguas estudiadas en su obra– al fula (Atlántico-Occidental), bambari, ngwo, bafut, kom (Benue-Congo) y al hausa, bura y margi (familia Afro-asiática). Por el contrario, una característica de las lenguas guineanas es la presencia de oclusivas labiovelares /kp b/, cuyos rastros sí son tangibles en el criollo saramacano de Surinam (Good 2005). Quizá la falta de recurrencia de la oclusiva glotal en las lenguas africanas sea una de las razones por las que no figura dentro de las aportaciones africanas en la formación de los criollos atlánticos (Parkvall 2000: 145-148). El segundo punto problemático de la hipótesis africanista de la oclusión glotal es, entonces, que no hay una fuerte correlación entre las estructuras fonéticas de las lenguas que, de acuerdo con los datos históricos, ejercieron un mayor influjo en el área y el rasgo en discusión. A pesar de ello, Granda plantea una posibilidad con respecto al origen del fenómeno que resulta bastante lógica desde un punto de vista fonético: Dicho proceso pudo haberse desarrollado […] por medio de la articulación, en un primer estadio temporal, de la [k] castellana como una consonante eyectiva de doble oclusión (velar y glotal), similar a las que aparecen abundantemente en las lenguas guineanas, de la cual pasaría, en un segundo estadio, posterior, a la pérdida de la oclusión velar, perdurando sólo la glotal. Este segundo tipo de proceso evolutivo encuentra significativos paralelismos en diferentes lenguas africanas en las que, partiendo de una consonante eyectiva, se ha pasado, por pérdida de la oclusión anterior, a una articulación glotal simple. Así ha ocurrido en bari, serer, margi y otras (1977: 126-127; las cursivas son mías).

MURIEL/OROZCO

31/5/12

52

14:24

Página 52

JOSÉ ALEJANDRO CORREA

Esta posibilidad se ve fuertemente limitada por la falta de recurrencia de las consonantes eyectivas en las lenguas guineanas. En el estudio fonético de las lenguas del África Occidental de Ladefoged (1964: 5), una de las obras de referencia de Granda, se afirma claramente: “In the sixty-one languages investigated, ejectives occurred only in Hausa, which has c’ k’ kw s’ (with some dialectal variation: Katsina Hausa has ts’ instead of s’)”. De esa manera, resulta difícil sostener –en el estado actual de nuestros conocimientos– que la oclusión glotal del Pacífico de Colombia se deriva de alguna lengua o grupo de lenguas africanas. El principal argumento en contra de dicho origen lo constituye la falta de correlación entre los datos históricos y lingüísticos; más bien este rasgo parece coincidir con la tendencia general de las oclusivas orales a realizarse como una oclusiva glotal. Desde esta perspectiva resulta útil preguntarnos si además de su relación segmental con una oclusiva velar sorda, ¿la oclusión glotal está en función de la estructura prosódica? y si es así ¿en qué contextos ocurre y cuáles son sus correlatos acústicos? En pocas palabras, ¿el comportamiento de la oclusión glotal coincide con el observado en otras lenguas como el inglés y el alemán? En § 4.2 presentaremos el análisis auditivo y acústico de este sonido y formularemos, sobre bases más seguras, nuestra hipótesis sobre este fonetismo dialectal.

3. Método 3.1. LOS PUNTOS DE ENCUESTA El trabajo de campo fue realizado en el departamento del Chocó en el mes de julio de 2006 en los municipios de Tutunendo, Cértegui, Condoto y Nóvita, en el último de los cuales pudimos contactar a un hombre adulto natural del municipio de Santa Rita de Iró (véase Mapa 1). Desde la colonia, estas localidades han sido centros mineros de gran importancia, de tal suerte que la explotación artesanal de este recurso continúa siendo la base de la economía local. No obstante, como escribió don Luis Flórez hace sesenta años, aun hoy “las condiciones de vida son miserables: mala es la salud, la comida, malas las habitaciones, malos los transportes, casi nulo el desarrollo económico…” (1950: 110). Pero vale la pena contarle al lector urbano que ante este panorama aciago, los habitantes de la región responden siempre con una inexorable sonrisa y una alegría inquebrantable9. 9

Como demostró Granda (1977), una de las consecuencias del aislamiento geográfico y del poco impacto de la educación y el desarrollo económico es la conservación de la tradición oral. En Nóvita, por ejemplo grabé este canto: Aunque mi amo me mate a la mina no voy/yo no quiero morir en un socavón/el blanco vive en su choza/de madera y con balcón/el pobre en

MURIEL/OROZCO

31/5/12

14:24

Página 53

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

53

La actual capital del Chocó, Quibdó, fue el centro administrativo de la provincia de Citará (siglos XVII y XVIII). Esta provincia –delimitada por el río Atrato– incluía a los actuales municipios de Tutunendo y Cértegui. Por su parte, el municipio de Condoto pertenecía a la provincia de Nóvita, cuya extensión estuvo determinada por el río San Juan (sobre este punto véase Cantor, 2000 y Velásquez, 1960). De estas provincias dependieron otras poblaciones mineras –reales de mina– y los resguardos indígenas que se encargaban de la producción agrícola para el mantenimiento de las cuadrillas de esclavos (las fotos 1 y 2 muestran la vista panorámica de Nóvita para 1973 y 2006, nótese el poco cambio que ha sufrido el poblado) A diferencia de lo que ocurrió en la Costa Atlántica de Colombia, en la Costa Pacífica no se habla ninguna lengua criolla. Este hecho resulta enigmático puesto que, como anota McWhorter (2000: 7-10), las condiciones prototípicas para la formación de una lengua criolla estaban dadas desde el inicio: poco contacto de los esclavos con población blanca, aislamiento geográfico, alto número de esclavos africanos, etc. No obstante, podemos decir que el habla del Chocó y del resto de la región es, sin duda alguna, un dialecto del español. Para explicar esta situación se ha dicho que el español chocano es producto de un proceso de descriollización (Granda 1978: 209). También se ha dicho que en el Chocó nunca se habló una lengua criolla debido a que las cuadrillas mineras estaban conformadas por esclavos mulatos y criollos (nacidos en América y hablantes nativos del español), lo cual habría generado una fragmentación lingüística que favoreció la imposición del español como código comunicativo (Cantor 2000: 45-57). Dado que esta cuestión se sale de los alcances de este estudio, sólo es posible decir que la solución del problema requiere un trabajo exhaustivo de descripción lingüística apoyado con información histórica de primera mano. Ahora bien, por la importancia histórica de la extracción del oro de aluvión, los mineros tradicionales representan un sector poblacional de especial importancia para conocer la evolución de la lengua española (incluidos los fenómenos estudiados) en esta porción de la América negra. Por esta razón se buscó que los hablantes encuestados, además de haber nacido y crecido en los puntos seleccionados, practicaran la extracción del oro como modo de subsistencia.

su choza de paja/con un solo napoleón/y aunque maten cadena a la mina no voy/yo no quiero morir en un socavón.

MURIEL/OROZCO

54

31/5/12

14:24

Página 54

JOSÉ ALEJANDRO CORREA

FOTO 1 Vista panorámica de Nóvita para 1973 (archivo Instituto Caro y Cuervo)

FOTO 2 Vista panorámica de Nóvita para 2006 (archivo personal)

3.2. PROCEDIMIENTO DE ANÁLISIS Dos consideraciones nos impulsaron a elegir estos puntos de encuesta. Primero, en la zona se encuentran los cursos altos de los ríos San Juan y Atrato, en cuyos tributarios se ha practicado la minería del oro desde la colonia. Segundo, en estos puntos se realizaron parte de las encuestas que dieron lugar a los trabajos pioneros de Flórez (1950, 1951), Montes Giraldo (1974) y Granda (1977), de tal suerte que se hiciera posible detectar la vitalidad de los fenómenos estudiados y poner a prueba nuestras hipótesis ante el habla de los lugares que antes fueron fuente de información. Para responder a las preguntas formuladas al final de las secciones § 2.1 y 2.2 fueron analizadas un total de trece muestras de habla espontánea. Los participantes se dividen en dos grupos generacionales: uno de hablantes adultos y otro de hablantes jóvenes, y sus edades oscilan entre los 40-75 y 12-25 años respectivamente. Dentro del grupo de participantes adultos fueron grabados siete hombres y dos mujeres, y dentro del grupo de participantes jóvenes, tres mujeres y un hombre (sus datos se utilizaron para el análisis de la oclusión glotal únicamente). Con el objeto de tener acceso a un registro de habla espontánea, en la recolección de muestras uno de los dos miembros del equipo se encargó de iniciar la conversación y crear un ambiente comunicativo cálido, en tanto, el segundo miembro se encargó de adecuar los equipos, presto a dar inicio a la grabación en

MURIEL/OROZCO

31/5/12

14:24

Página 55

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

MAPA 1 Puntos de encuesta en el departamento del Chocó

55

MURIEL/OROZCO

31/5/12

56

14:24

Página 56

JOSÉ ALEJANDRO CORREA

el momento apropiado. Dadas las habilidades comunicativas del encuestador, y gracias a la expresividad de los habitantes de la región, nuestro procedimiento tuvo éxito en la mayoría de los casos. Las grabaciones fueron capturadas con una consola digital y un micrófono omnidireccional, y se cuidaron al máximo las condiciones externas para asegurar la calidad de las muestras. Las señales se digitalizaron a 22.100 Hz y fueron segmentadas y analizadas en el programa Praat (Boersma/Weenink 2007). El análisis fue auditivo e instrumental: se analizaron los fonemas /r, ɾ, d/ frente a las variantes [r, ɾ, ð
, l] en posición inicial de palabra después de una vocal y en posición intervocálica, para lo cual se cotejó el espectrograma de cada aparición. En el caso de la oclusión glotal, se buscó su estímulo auditivo en todos los contextos posibles, y se consultó el oscilograma prestando atención a las categorías acústicas reportadas por Redi/Shattuck-Hufnagel (2001).

4. Resultados 4.1. LAS CONSONANTES CORONALES La Figura 1 muestra los porcentajes (abscisas) y el número de variantes fonéticas de los fonemas /d, r, ɾ/ en posición intervocálica y en posición inicial de palabra después de una vocal (los números enteros que aparecen sobre cada barra). Las estimaciones fueron hechas a partir de las muestras tomadas a los hablantes adultos (nueve grabaciones) puesto que, de acuerdo con análisis previos, la norma de habla de este grupo etario (40-75 años) conserva con mayor vigor el fenómeno estudiado. En otras palabras, se decidió mantener controlada la variable edad para poder responder los interrogantes diacrónicos planteados en §2.1. Las variantes estudiadas son [ɾ, ð
, r, l] y, además, se considera la elisión [ø] de estos sonidos como otra resolución articulatoria. En primer lugar, el diagrama de barras indica que el fonema /ɾ/ se mantiene como una vibrante simple en la mayoría de los casos (61,3%), pero su producción está gobernada por un proceso de relajamiento articulatorio que permite como variantes a la aproximante dental [ð
] (8,6%) o la elisión completa de la consonante (29,6%). La pronunciación de /ɾ/ como una consonante lateral [l] que, como vimos anteriormente, tiene cierta relevancia desde un punto de vista diacrónico, parece estar perdiendo vitalidad, pues sólo tiene una frecuencia de dos apariciones (0.4%). Ejemplos que ilustran las tendencias: pero ['peɾo], claro ['klað
o], aire [ai e] y él diría ['eð
i'lia]. En segundo lugar, es patente que el fonema /d/ está dejando de ser pronunciado como una aproximante dental, puesto que esta variante del español estándar

MURIEL/OROZCO

31/5/12

14:24

Página 57

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

57

FIGURA 1 Diagrama de barras con el promedio (los valores exactos se especifican en el texto) y número de realizaciones fonéticas de las consonantes coronales del español del Chocó

sólo alcanza un 24,3%. En su lugar, los hablantes propenden a eliminar el segmento (63,9%) y a veces se interpreta como una vibrante simple (11,30%). La variante lateral [l] también constituye una posibilidad sonora, pero solo hemos podido reportar un caso (0.4%) como realización de /d/. Ejemplos que ilustran las tendencias: cada uno ['kað
a uno], miedo ['mjeo], se quedó [se ke'ɾo] y queda bueno ['kela'∫
weno]. Finalmente, los datos correspondientes a la vibrante múltiple /r/ señalan que de las tres variables categóricas estudiadas, ésta es la que presenta menor variación. Esto significa que los hablantes mantienen su identidad fonética como vibrante múltiple (86,6%), y sólo en algunas situaciones comunicativas se desvían de la norma y acuden a una vibrante simple (8,9%), a una aproximante dental (2,8%) o a la variante lateral (2,8%). Ejemplos que ilustran las tendencias: la res [la 'res], la res [la 'ɾes], arranqué [að
an'ke] y arranqué [alan'ke]. Cuando tomamos enunciados completos y no unidades aisladas, la variación descrita trasciende el dominio de la palabra y determina la forma de todo el enunciado. La Figura 2 muestra los espectrogramas de las expresiones: y arranqué y me fui [jað
an'ki me '§wi], aire respiratorio [air epjatoð
jo] y si Dios quiere [si 'joh kje:]. Nótese que la variación articulatoria que hemos descrito provoca que el número de segmentos se reduzca y/o la estructura silábica se simplifique, de

MURIEL/OROZCO

58

31/5/12

14:24

Página 58

JOSÉ ALEJANDRO CORREA

manera que las palabras tienden a fusionarse sintagmáticamente. En el primer ejemplo, que en el español estándar puede transcribirse [ja. ran.'kei .me.'fwi] es pronunciado, por un hablante del español chocoano, como [jað
an'ki me '§wi]. Las dos versiones mantienen el mismo número de sílabas, pero en la versión chocoana la vibrante [r] pasa a una “aproximación” de la lámina de la lengua hacia la protuberancia alveolar, y la consonante posvocálica /n/ se deriva de un breve descenso del velo del paladar, casi imperceptible en el espectrograma que mostramos en la parte superior de la Figura 2. Además, el diptongo [ei ] cambia a [i], y la fricativa labiodental sorda /f/ manifiesta una sutil fricción producida al nivel de la glotis [§], que es sonora por la naturaleza de los sonidos vecinos e_w. En el segundo ejemplo, aire respiratorio [ai ɾepjatoð
jo]10, la sílaba final de aire se fusiona con la primera de respiratorio y, adicionalmente, la vibrante simple se elide y se realiza como una aproximante [ð
] en la misma expresión. El espectrograma de la parte inferior de la figura 2, correspondiente a si Dios quiere [si 'joh kje:], también ilustra cómo la elisión de /ɾ/ y /d/, sumadas al debilitamiento de /s/ al inicio de si y al final de Dios, generan una señal reducida en elementos pero suficientemente informativa. Estas reducciones masivas que involucran menor número de sílabas y/o el cambio en la cualidad fonética de los segmentos esperados, también han sido registradas en el habla espontánea de otras lenguas como el alemán, inglés y el sueco (véanse los trabajos de Kohler 1998, 1999; Johnson 2004 y Engstrand/Krull 2001). Los resultados pueden resumirse en los siguientes términos: la alternancia fonética entre [ð
~ ɾ ~ r ~ l], común en el área africana, ha sido un hábito que ha logrado mantenerse y recrearse en el español hablado por los mineros tradicionales del Pacífico de Colombia. Digo que la variación se mantiene porque aun cuando ciertas variantes fonéticas (por ej., [l]) no parecen incidir en el estatus fonético/fonológico de las variables estudiadas, si tomamos in toto la variación descrita en la Figura 1, ésta se explica como parte de un proceso de reducciones articulatorias que afectan el número de sílabas y la cualidad fonética de los segmentos11. Se ha recreado porque las reglas fonotácticas12 que regulan la

10

Nótese que a lo largo de este documento he adoptado la tradición de transcribir los diptongos ascendentes con las semivocales [j] o [w] más un núcleo, y los diptongos descendentes con un núcleo seguido de las semivocales [i ] o [u

]. Para estas y otras convenciones véase Martínez-Celdrán et al. (2003). 11 A propósito de una interpretación novedosa de datos semejantes en alemán, sugiero al lector la lectura de los revolucionarios trabajos que ha venido produciendo Klaus Kohler y colegas en el Instituto de fonética y procesamiento digital de señales del habla, Universidad de Kiel.

MURIEL/OROZCO

31/5/12

14:24

Página 59

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

FIGURA 2 Fusión de palabras en las que se involucran consonantes coronales

59

MURIEL/OROZCO

31/5/12

14:24

60

Página 60

JOSÉ ALEJANDRO CORREA

variación de las consonantes coronales en las lenguas africanas no han logrado transmitirse, pero la variación se ha integrado al sistema sonoro de la lengua española.

4.2. LA GLOTALIZACIÓN El fenómeno de la glotalización actualmente no es recurrente en el habla de las poblaciones que se incluyeron dentro del estudio. Entre los hablantes adultos sólo dos –un hombre y una mujer– usaron glotalizaciones consistentemente. Pero aun cuando la mayoría de adultos no usa esta articulación en su habla cotidiana, pudimos notar que los jóvenes (cuatro muestras) tienen cierta propensión a usarla. La razón por la cual son escasas las glotalizaciones entre los adultos resulta ser una cuestión problemática, puesto que, de acuerdo con mis observaciones, la edad y el grado de educación no parecen ser variables determinantes; la mayoría de nuestros informantes reportaron un nivel bajo o ningún grado de escolaridad, y ninguno de aquellos que sobrepasaba los setenta años presentó ejemplo alguno. El hablante masculino tenía –en el momento de la grabación– treinta y nueve años de edad, y nació en Santa Rita de Iró. La mujer, natural del municipio de Cértegui, al momento de la grabación tenía sesenta y nueve años de edad, y aun cuando está alfabetizada, su habla no se asemeja a la de ningún otro vecino de este municipio. Para despejar este interrogante considero que es necesario explorar en un futuro los cursos altos de las zonas mineras para establecer posibles patrones diatópicos, e indagar a fondo la historia de vida de los hablantes para poder extraer generalizaciones sociolingüísticas. A pesar de su poca difusión en la comunidad de habla, a nivel idiolectal la glotalización es un fenómeno con formas y funciones variadas. En el habla del hombre adulto se identificaron un total de cincuenta y dos realizaciones, de las cuales cuarenta y cinco (86,5%) se utilizaron como un alófono de /k/. Las siete realizaciones restantes (13,5%) sirvieron para demarcar el límite entre dos frases entonativas, el inicio de una frase entonativa, para denotar énfasis y en una ocasión fue producto de una desbucalización, por ej., me devolví no, no quise coger

12 Por ejemplo, vimos anteriormente que la lengua bantú beembe, restringe la aparición de /d/ a la posición inicial de palabra, y /l/ se realiza como una vibrante simple frente a /i/ y entre vocales posteriores por, ej., ‘piedra’ /tálì/ > [tháɾì], ‘valle’ /dú:lù/ > [dú:ɾù] (Jacquot º 1962). Las restricciones de esta lengua –y de muchas otras– no se han conservado en la variedad que estudiamos, pero la variación de las coronales si ha logrado afectar la posición intervocálica del español.

MURIEL/OROZCO

31/5/12

14:24

Página 61

61

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

TABLA 1 Promedio y número de realizaciones acústicas de la oclusión glotal en dos hablantes adultos Oclusión canónica

Pulso glotal breve

Aperiodicidad Laringalización

Vocal laringalizada

Descenso de la intensidad

Total

M

9 (17,3%) 5 (9,6%)

13 (25%)

7 (13,4%)

9 (17,3%)

9 (17,3%)

52 (100%)

F

20 (24,3%) 6 (7,3%)

–

19 (23,2%)

37 (45,2%)

–

82 (100%)

bocado más [meo l'bi nõ ʔ no kise ko§eɾ β ao 'mas], es muy poquito [ʔe 'mwi po'kito], me ayudó [me ʔa'wo], ˇpodemos ¯matar (a)l tipo [poe mo ma'ta 'tiʔe]. La Tabla 1 muestra el promedio y número de realizaciones acústicas de la oclusión glotal para los dos hablantes adultos. En total se reportaron seis correlatos para la articulación en cuestión (Figura 3). En la muestra analizada, el hombre adulto (M) hizo uso de las seis articulaciones posibles, la distribución de valores es proporcional para cada variante, y sólo se aprecia una preferencia por la aperiodicidad. No obstante, siete de las nueve oclusiones glotales canónicas (cuando hay un cese completo de los pulsos glotales) se utilizaron para demarcar los límites de la frase entonativa, énfasis y la desbucalización. La mujer adulta (F) produjo un total de ochenta y dos realizaciones, de las cuales setenta y cinco (91,5%) fueron alófono de /k/. Los siete casos adicionales (8,5%) sirvieron para demarcar el inicio de una frase entonativa o para denotar énfasis, por ej., así como [ʔa'si omo], tiene seis años, ['tjene 'sei s ʔ'ao]. Como muestra la Tabla 1, la hablante hizo uso de cuatro de seis realizaciones fonéticas posibles, no haciendo parte de su repertorio la aperiodicidad ni el descenso de la intensidad. Hay una clara preferencia por las laringelizaciones, ya sea como variante consonántica o como vocal laringalizada (creaky voice), por ej., por(que) ya esta(ba) malita [po ata ma'lita]. Al igual que el hablante masculino, siete de las veinte oclusiones˜ glotales canónicas se usaron para demarcar los límites de la frase entonativa o énfasis; no se presentaron desbucalizaciones en este caso. Gracias a la técnica de la observación participante, pudimos notar que esta articulación también es usada por los jóvenes. La inspección de la señal del habla de tres mujeres y un hombre (veinte minutos de conversación en cada caso) mostró que el comportamiento del sonido estudiado es bastante homogéneo en este grupo generacional. Cada participante produjo un promedio de tres oclusiones glotales a lo largo de veinte minutos de conversación, ninguno de ellos utilizó el sonido como alófono de una consonante oclusiva velar sorda, todos utilizan la

MURIEL/OROZCO

31/5/12

14:24

Página 62

62

JOSÉ ALEJANDRO CORREA

oclusión glotal para demarcar los límites –inicial o medio– de la frase entonativa, el contexto de ocurrencia fue v_v, y la realización fonética típica fue el cierre completo del articulador laríngeo. Algunos ejemplos: nosotros, el grupo de nosotros [no'sotɾo ʔel ' ɾupo de no'sotɾo], como yo, él es el que está dando la plata [komo 'o ʔel 'e ke 'ta 'ando la 'plata], o sea nosotras [ʔo sea 'nota]. Aunque la ocurrencia del fenómeno es baja por el tipo de corpus seleccionado (habla espontánea), resulta interesante comprobar que los datos concuerdan con los presentados por Valentín-Márquez (2006) para el español de Puerto Rico, en donde se reporta que la oclusión glotal está emergiendo en el mismo contexto prevocálico, por ej., tres años [ tɾeʔ'ao], las únicas [la ʔ 'unikah]. Además, debo decir que en recientes visitas a San Basilio de Palenque he notado que este fenómeno también está ocurriendo en el español hablado por sus hablantes más jóvenes. Nuestros resultados indican, pues, que la oclusión glotal no sólo es utilizada en el español del Pacífico de Colombia como un alófono de /k/, sino que sus funciones se extienden a nivel suprasegmental para demarcar los límites inicial y medio de la frase entonativa. De otro lado, vemos que tras el símbolo [ʔ], el cual usamos inadvertidamente a causa de nuestra familiaridad con el alfabeto, se esconde un fino mecanismo que genera una impresión perceptual por medio de diferentes configuraciones del tracto laríngeo13. Si sumamos nuestros resultados a los de otros autores que han estudiado los dialectos del español americano (Chela Flores 1986; Zamora/Guitart 1988), encontramos viable afirmar que la oclusión glotal ha surgido de un cambio fonético independiente cuyos mecanismos acústicos y funciones comunicativas coinciden plenamente con los reportados en otras lenguas del mundo (Pierrehumbert/Talkin 1991; Dilley/Shattuck-Hufnagel/ Ostendorf 1995; Kohler 1999; Redi/Shattuck-Hufnagel 2001; Esling/Fraser/ Harris 2005). De otro lado, el trabajo de Valentín-Márquez (2006), el cual muestra cómo la oclusión glotal está surgiendo entre los jóvenes de Puerto Rico, coincide con nuestros resultados sobre el uso de la glotalización entre los jóvenes del Pacífico de Colombia. Esto significa que resulta plausible que este rasgo fonético sea un cambio lingüístico en curso, y que afecte a diferentes dialectos de la lengua española. x

5. Resumen y conclusiones La intención del presente estudio ha sido evaluar el presunto origen africano de dos rasgos fonéticos del español hablado en el Pacífico de Colombia, i.e., la

13

Para el desarrollo de este concepto véase Esling (2005b).

MURIEL/OROZCO

31/5/12

14:24

Página 63

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

63

FIGURA 3 Correlatos acústicos de la glotalización: a) cese prolongado de los pulsos glotales, b) un pulso glotal breve, c) aperiodicidad, d) laringalización y vocal laringalizada –porción de onda a la derecha del recuadro– y e) reducción de la intensidad. Cada recuadro indica la parte de la señal que corresponde a la oclusión glotal

MURIEL/OROZCO

64

31/5/12

14:24

Página 64

JOSÉ ALEJANDRO CORREA

alternancia entre los fonos [ð
] ~[ɾ], y la realización de la oclusiva velar sorda /k/ como una oclusiva glotal. Con respecto al primer fenómeno, vimos que Granda (1988) identificó en los archivos históricos que los africanos llevados al territorio que estudiamos fueron principalmente hablantes de lenguas de las familias Kwa (akan y ewe, especialmente) y Bantú. Para él, puesto que en la mayoría de las lenguas pertenecientes a estas familias la articulación de /d/ es alveolar y posalveolar (no dental como en español), resulta natural que los hablantes hayan adoptado una vibrante simple y no una aproximante dental i. e., [ð
] > [ɾ]. A este tenor, Lipski (2007) considera que en las áreas en las que se presenta este fonetismo (incluido el Pacífico de Colombia) las tradiciones aluden esencialmente a la cuenca del Congo. Luego muestra algunos préstamos del portugués al kikongo en los que se presenta el cambio [ɾ] > [ð
] (inverso al descrito por Granda), y concluye que el fenómeno proviene del kikongo y lenguas vecinas (véase también Schwegler, en preparación). Un análisis más detallado de la bibliografía (Jacquot, 1962; Ladefoged, 1964; Lindau, 1980; Parkvall 2000; Hyman 2003) nos ha permitido establecer que en las lenguas africanas (las habladas en el Golfo de Guinea y en la zona centrooccidental de África) no es común la distinción fonológica entre los fonemas /d, r, ɾ/ y /l/. Nos preguntamos entonces ¿en el español del Pacífico de Colombia ocurre una variación fonética entre las consonantes coronales semejante a la que ha sido reportada en el área lingüística africana? Así, dijimos que si este planteamiento resultaba válido, estaríamos ante un caso en el que un rasgo fonético de un área lingüística (i. e., la africana) se transmitió al español hablado por los afrodescendientes de América, pero no estaríamos ante un caso en el que una única lengua (por ej., el kikongo) ha actuado como sustrato. Con el planteamiento anterior en mente, decidimos examinar auditiva e instrumentalmente una serie de muestras de habla espontánea para comprobar si en el territorio estudiado existe una variación semejante a la que exhiben gran parte de las lenguas africanas. Examinamos los fonemas /d, r, ɾ / frente a las variantes [ð, ɾ , r, l] y [ø] en posición intervocálica y en posición inicial de palabra después de una vocal (sección § 4.1). Los resultados arrojados por el análisis mostraron que es plausible afirmar que la variación fonética característica de las lenguas africanas ha logrado mantenerse (en mayor o menor medida, dependiendo de la variante) en el habla de la población estudiada. Así, podemos concluir que: (1) /ɾ/ se realiza como una vibrante simple en la mayoría de los casos, pero puede debilitarse y realizarse como una aproximante dental [ð
], elidirse completamente o realizarse como una lateral [l]; (2) /d/ se elide con gran frecuencia, se pronuncia como aproximante dental [ð
], como una vibrante simple [ɾ] y en menor medida como una consonante lateral [l]; (3) /r/ es el fonema que menos variación presenta, puesto que en la mayoría de casos se realiza como una vibrante múltiple

MURIEL/OROZCO

31/5/12

14:24

Página 65

EL ESPAÑOL HABLADO EN EL PACÍFICO DE COLOMBIA

65

[r], pero sufre un relajamiento articulatorio que hace que se pronuncie como una vibrante simple [ɾ], como una aproximante dental [ð
] o como una lateral [l]. En cuanto al segundo caso estudiado (la realización de /k/ como una oclusiva glotal) se mostró (sección §2.2) que el procedimiento usado por Granda (1977) estuvo basado principalmente en el estudio de los acontecimientos históricos que determinaron la preponderancia de la población negra en el Pacífico de Colombia y en la identificación de las poblaciones africanas que formaron las cuadrillas mineras de la región. Una vez hecho esto, Granda propuso la hipótesis de que la oclusión glotal pudo haberse originado como producto de la pérdida del componente oral de una consonante eyectiva i. e., /k’/ > [ʔ]. Como vimos, esta hipótesis no resulta viable porque las eyectivas no son comunes en las lenguas Kwa ni en las lenguas bantúes (para las lenguas guineanas véase Ladefoged, 1964), y el rasgo tampoco aparece entre los rasgos de sustrato africano de los criollos atlánticos (véase, Parkvall 2000). Basados en la abundante bibliografía que se ha producido sobre el tema durante los últimos años (por ej., Chela Flores, 1986; Zamora/Guitart, 1988; Pierrehumbert/Talkin 1991; Dilley/Shattuck-Hufnagel/ Ostendorf 1995; Kohler 1999; Redi/Shattuck-Hufnagel 2001; Esling/Fraser/ Harris 2005) dijimos que existen dos razones para pensar que la oclusión glotal es producto de un cambio fonético regular. Primero, es sabido que el cambio /k/ > [ʔ] ha sido recurrente en la evolución de varias lenguas no africanas y, segundo, la oclusión tiene una función demarcativa y sirve para delimitar los límites inicial y medio de las frases entonativas (ejemplos clásicos son el inglés y el alemán). Así, planteamos dos preguntas: ¿en qué contextos ocurre la oclusión glotal y cuáles son sus correlatos acústicos? y ¿el comportamiento de la oclusión glotal coincide con el observado en otras lenguas como el inglés y el alemán? En la sección § 4.2 analizamos el habla de dos hablantes adultos (mineros de profesión) y cuatro jóvenes, quienes fueron, dentro del corpus recogido, los únicos en presentar oclusiones glotales. En el habla de los adultos la oclusión glotal ocurrió como un alófono de /k/ en la mayoría de los casos, pero también sirvió para demarcar los límites (inicial y medio) de las frases entonativas, para denotar énfasis y, marginalmente, como producto de desbucalizaciones. Los hablantes hicieron uso de por lo menos seis estrategias articulatorias para producir el estímulo auditivo de una oclusión glotal: 1) un cese prolongado de los pulsos glotales, 2) un pulso glotal de breve duración, 3) aperiodicidad o sucesión irregular de pulsos glotales, 4) laringalización, 5) asimilación de la laringalización a la vocal contigua (creaky voice) y 6) descenso de la intensidad sin cese de los pulsos glotales. En los hablantes jóvenes la oclusión demarcó el límite de las frases entonativas. En ambos grupos generacionales el cese prolongado de los pulsos glotales (oclusión canónica) tuvo una relación directa con la función demarcativa. A pesar de los pocos datos que encontramos para explicar este fenómeno conside-

MURIEL/OROZCO

66

31/5/12

14:24

Página 66

JOSÉ ALEJANDRO CORREA

ramos que, como planteamos en la primera parte del estudio, la oclusión puede analizarse como un cambio fonético independiente. Esta posición tiene la ventaja de centrar la discusión en las propiedades fonéticas y comunicativas del fenómeno, y no nos enfrenta ante divagaciones históricas que, dada la falta de información pertinente, resultan infructíferas. Finalmente, se han dado algunas indicaciones que pueden guiar la investigación futura. Primero, el comportamiento de los dos fenómenos estudiados produce una reducción considerable de las palabras y de los enunciados. Sería interesante estudiar cómo estos procesos alteran (o no) la comprensión de los mensajes. Puede pensarse en un ejemplo extremo como porque ya esta(ba) malita [po ata malita] en el que la segunda sílaba de porque se integra a la primera /po/ por medio de la laringalización del núcleo y, además, el morfema {-ba} está ausente. Segundo, resulta importante estudiar a fondo el comportamiento de la oclusión glotal entre los jóvenes puesto que el fenómeno se da en varios dialectos caribeños. En este aspecto un estudio interdialectal resultaría óptimo para saber cómo está emergiendo este sonido en el español de América.

MURIEL/OROZCO

31/5/12

14:24

Página 67

L O C AT I O N A L A D V E R B S I N C O L O M B I A N S P A N I S H C O N V E R S AT I O N CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

1. The Spanish locational adverb system The Spanish locational adverb system comprises five terms which are traditionally divided in two ways. First they are divided into three sets on the basis of some idea of distance: aquí and acá (both corresponding in some sense to English here) are treated as indicating location close to the speaker; ahí (corresponding to there) is considered either to indicate a location close to the hearer or else a location medium-distant from speaker; and allí and allá (also corresponding to there) are considered to be either distant from both speaker and hearer, or alternatively far-distant from the speaker (cf. Section 4). On another axis, these adverbs are treated as belonging in two series, those ending in -i and those ending in -a (cf. Section 6). The two series are typically understood in terms of either location (-i) versus direction (-a), or specific (-i) versus more general locations (-a), although advocates of both theories admit that the distinction is not absolute. While there is disagreement about precisely what the distinction between the two series is, it is widely agreed that the formal difference between the adverbs corresponds to a semantic or functional difference (cf. Terrado Pablo 1990: 63), that is, that “the adverbs ending in -í seem to fall into one category, and those ending in -á in another” (Sacks 1987: 292). In this paper, we analyze the use of these five terms in the conversational Spanish of Cali, Colombia, and demonstrate that their use differs markedly from that traditionally described in the literature. We find that ahí is not fully entrenched in the system in that, when used spatially, it is by no means restricted to a medial sense and furthermore, it exhibits a wide range of non-spatial uses. We also find that the -i and -a adverbs do not pattern as two distinct series, as neither series exhibits homogeneous behavior, and that this is only partially attributable to the behavior of ahí. Furthermore, we note that the adverbs occur repeatedly in certain fixed constructions such as más acá / más allá ‘closer / further away’, para acá / para allá (or pa’cá / pa’llá) ‘in this / that direction’, por ahí ‘around there’, and ahí vemos ‘we’ll see’, or frequently occurring collocations, such as estar aquí ‘be here’ and allá es … ‘(that place) there is ...’. We propose that the meaning differences that have been ascribed to the -i and -a series

MURIEL/OROZCO

31/5/12

14:24

68

Página 68

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

are in many cases attributable to these constructions, rather than the adverbs themselves. We briefly compare our results with those reported for other varieties of Spanish, and argue that this comparison suggests that these findings are not unique to this variety of Spanish, nor to the conversational genre studied, though the patterns we observe in our data have not yet been systematically studied for other varieties.

2. Corpus and distribution of the adverbs The data for this study are drawn from approximately eight and a half hours of conversational Colombian Spanish (~ 90,000 words), collected in the city of Cali, Colombia (cf. Travis 2005). Two native speakers recorded spontaneous interactions with their husbands, families and friends over periods of several months during 1997 and 2004. A total of 27 speakers are represented, including 19 women and 8 men, all between the ages of 20 and 60, from the middle class. The data have been transcribed following the method outlined in Du Bois/Schuetze-Coburn/Cumming/Paolino (1993). The frequencies with which the adverbs occur in the corpus are given in Table 1. There are several important points to observe about this distribution. Firstly, note that for the proximal terms, aquí (from the -i series) is the overwhelmingly preferred adverb (accounting for 79% of the proximal adverbs), while for the distal terms, it is allá (from the -a series) which is the preferred form (accounting for 86% of the distal adverbs). Thus, merely in terms of distribution we note a discrepancy across the two series. Also of interest is the fact that the so-called medial term, ahí, is the most frequent adverb overall, accounting for close to 40% of the total number of tokens of all adverbs (394/1,048), which, we will see below, is due to the high frequency with which it is used in a non-spatial sense (cf. Table 8). This frequency data alone, then, gives reason to question the systematicity of the two series, an issue we return to below. TABLE 1 Frequencies of occurrence in the data (N = 1,048) Proximal

N

%

Medial

-i series

aquí

291

79%

ahí

-a series

acá

77

21%

—

368

100%

N

%

394 100%

394 100%

Distal

N

%

allí

41

14%

allá

245

86%

286 100%

MURIEL/OROZCO

31/5/12

14:24

Página 69

69

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

3. Distribution in other varieties of Spanish In comparing this distribution with other varieties of Spanish for which such data are available, we note that skewed behavior of the proximal and distal terms is by no means unique to our corpus. Table 2 presents a summary of the distribution of -i versus -a pairs of adverbs in our conversational data from Cali with that reported by Miyoshi (1999) for other varieties of spoken Spanish (the source of each data set is given below the table). TABLE 2 Distribution of aquí/acá and allí/allá in six varieties of spoken Spanish

1 2 3 4 5

aquí

acá

allí

allá

Cali, Colombia

79%

21%

14%

86%

Bogotá, Colombia1

85%

15%

31%

69%

Caracas, Venezuela2

85%

15%

18%

82%

Mexico City, Mexico3

91%

9%

61%

39%

Buenos Aires, Argentina4

20%

80%

48%

52%

Madrid, Spain5

99%

1%

91%

9%

Habla culta (Miyamoto 1993). Sociolinguistic interviews (Sedano 1996). El habla de la ciudad de México (Miyoshi 1999). Habla culta (Sedano 1995a); Habla culta (Sedano 1995a).

This comparison reveals that in all dialects reported here, except Buenos Aires (cf. Kany 1975: 269, Sacks 1987: 312), aquí is the overwhelmingly preferred proximal term, and in Madrid, this is nearly categorical. For the distal terms, we see more variability: allá is preferred in Cali, Bogotá and Caracas, while allí is preferred in Mexico City and Madrid, with Buenos Aires showing an even distribution. What is most interesting from our point of view is that few dialects show consistent favoring of one or the other series. Madrid is the only variety to do this strongly (showing near categorical use of the -i series, as has been noted in the literature, cf. Carbonero Cano 1979, Nilsson 1983, Richardson 1996, Schmidely 1975), and while Mexico City does show proportionally higher use of the -i series for both the distal and proximal terms, the difference is much more marked for the proximal terms (91% aquí) than for the distal terms (61% allí).

MURIEL/OROZCO

31/5/12

70

14:24

Página 70

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

Table 3 compares the distribution of ahí and allí in the data from Cali with Sedano’s (1999a) data from Caracas, Venezuela (the only study which provides such token counts), and shows that in both varieties, ahí is used substantially more frequently than allí (respectively 10 times and 7 times more frequently). In contrast, work on written Peninsular Spanish (Nilsson 1983, Richardson 1996, Schmidely 1975) shows that the two adverbs either occur equally frequently, or allí occurs with a higher rate than ahí. This striking difference in relative frequency between these spoken Latin American varieties and written Peninsular Spanish may be due to a genre or dialect effect, something which we leave for future analysis. TABLE 3 Distribution of ahí/allí in two varieties of spoken Spanish

1

ahí

allí

Cali, Colombia

91%

9%

Caracas, Venezuela1

88%

12%

Sociolinguistic interviews (Sedano 1999a)

It has been proposed that ahí is taking over from allí in Latin American Spanish (Kany 1975, Sacks 1987), or that the contrast between the two is being lost (Moreno F. 1985: 160, Sacks 1987: 308). In her study comparing the use of these two adverbs in Venezuelan Spanish, Sedano (1999a: 52) notes that some of her speakers reported allí to be simply a more precise pronunciation than ahí. The difference in pronunciation in that variety, as in the one studied here, is that allí occurs with a palatal approximant: ahí [ai] / allí [aji]. Sedano’s finding that speakers of middle and upper socio-economic status favor the use of allí more than speakers of lower socio-economic status suggests that allí may be a prestige form (1999a: 57), and that the choice of one form or the other may depend on social rather than functional or distributional principles. The low number of tokens of allí in our corpus means that the patterning of this adverb that we observe must be taken as preliminary, and in need of further examination in a larger corpus where the set of factors that jointly affect speakers’ choices between the two forms can be identified (cf. Sedano 1999a)1. 1

As pointed out by one of the editors of the volume, experimental studies evaluating the ability of speakers / transcribers to distinguish between [a'i] and [a'ji] in different contexts could also be insightful here.

MURIEL/OROZCO

31/5/12

14:24

Página 71

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

71

4. Distance axis As noted earlier, the Spanish locational adverbial system is considered to mark a three-way spatial distinction, though there is much discussion about precisely how to interpret the degrees of distance. They are typically analyzed as consisting of terms indicating location close to the speaker (aquí, acá), close to the hearer (ahí), and distant from both (allí, allá); or two proximate, one medial and two distal locational terms (e.g. Brzozowska-Zburzynska 2005, Butt/Benjamin 2004, Hottenroth 1982, Moreno F. 1985, RAE 1973, Richardson 1996, Schmidely 1975). Although it is readily apparent from our data that aquí and acá refer to a location closer to the speaker than do allí and allá, the place of ahí in this spatial system is far from clear. We observe ahí used for ‘proximal’ locations, as in (1) where it refers to the speaker’s finger, as well as ‘distal’ locations, as in (2), where it refers to a location a day’s drive from where the conversation is being held2. (1) Re marks on his finger, where he pricked himself with a chicken bone S: ahí tambien me chucé con el pollo. [chuzo 304]3 Santi: ‘ahí also I jabbed myself with the chicken.’ (2) Re the ecological damage caused by the guerrilla in comparison to a major oil company R: un oleoducto, .. no es !nada, en comparación con lo que bota= -.. sólo Ecopetrol, .. ahí en la bahía de Cartagena, todos los días. [contamination 380] Rocío: ‘((blowing up)) an oil pipeline, is nothing, in comparison with what Ecopetrol alone puts ahí into the bay of Cartagena, every day.’

Further evidence for the fact that ahí is not restricted to medial distance is seen in the apparent interchangeability between aquí and ahí, as well as between ahí and allí, in some contexts in our data; the latter in particular has been noted

2

One reviewer notes that the use of a non-distal adverbial to refer to a distant location can serve to bring that location closer, as when recounting memories, for example. While this does occur in the data, the use of ahí to refer to remote locations is not limited to this kind of ‘psychological’ closeness. In (2), for example, the speaker is recounting something which someone else has told her, with which she has no personal connection. 3 This gives the name of the transcription from which the example is drawn and the line number of the occurrence of the adverb. Transcription conventions are listed in the appendix.

MURIEL/OROZCO

31/5/12

72

14:24

Página 72

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

in the literature (e.g. Moreno F. 1985: 160, Sacks 1987: 308, Sedano 1994b: 226, Sedano 1999a: 52). This can be seen in the following examples, where ahí is used by the same speaker to refer to the same location as aquí in (3) and as allí in (4). (3) A:

Tenía pensado que eso era aquí=, .. donde te lavaron el carro. Como ahí fue que te lo quitaron cierto? [almuerzo 394] Ángela: ‘I had thought that it was aquí, where they washed the car. Since ahí was where they removed it ((the carpet)) for you, right?’

(4) A: M: A: Ángela: María: Ángela:

.. Y en dónde lo echo, allí mismo? Hm? Ahí mismo? ‘And where shall I put it? Right allí?’ ‘Hm?’ ‘Right ahí?’

[calima2 138]

Thus while aquí/acá and allí/allá can be understood to be distinguished by referring to a location which is relatively close or distant from the speaker, in its spatial use ahí should be understood as a general term, with no restrictions in terms of distance from speaker or addressee (cf. Curnow/Travis 2008). This generalized use of ahí has been noted in the literature (e.g. Molho 1992, Nilsson 1983, Sedano 2000: 926), and thus is likely to be a pan-Hispanic phenomenon.

5. Deictic vs. anaphoric uses These locative adverbs have a deictic function, where they point out a location that is physically present in the place where the speech event takes place. This use can be seen in (1), referring to the speaker’s finger, and in (4), referring to the interlocutor’s dinner plate. Deictic forms can often also be used anaphorically, making reference to prior mentions in the discourse (Levinson 1983, Lyons 1977), as can be seen in the use of ahí in (3) to refer back to the location just mentioned by the speaker (aquí donde te lavaron el carro ‘here where they washed your car’). It has been proposed that the -i adverbs show greater anaphoric use than the -a adverbs, a retention of the anaphoric uses of hic in Vulgar Latin from which they derive (e.g. Molho 1992: 207). Although the anaphoric / deictic distinction is widely recognized in the literature, it is extremely difficult to operationalize for these locative adverbs in the conversational data under study here, as Sedano also found in her data from sociolinguistic interviews (1999a: 54). In particular, there are many tokens that would qualify as both anaphoric and deictic (referred to as ‘anadeictic’ by

MURIEL/OROZCO

31/5/12

14:24

Página 73

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

73

Sedano 1996, 2000). Example (5) below provides an illustration, where Santi’s use of ahí refers to a location which is at the same time physically present in the context (‘there on the plate in front of us’), and referred to previously (‘there where you just mentioned’). In cases such as these, it is impossible to determine whether one or the other use is ‘primary’. (5) S: A: S: A: S: Santi: Ángela: Santi: Ángela: Santi:

Quieres un poquito de esto? ... Ahorita lo pruebo. Pruébalo. ...(3.0) @@ Sino que tiene [muchas salsitas] -[Por aquí]? Ahí no hay nada mi amor, [restaurant 730] ‘Would you like a bit of this?’ ‘I’ll try it in a moment.’ ‘Try it. …(3.0) @@ But it’s got lots of sauces.’ ‘Around aquí?’ ‘Ahí there’s nothing, my love,’

Rather than attempt to determine whether each adverb is used anaphorically, here we consider merely whether the location referred to by the adverb has been mentioned previously in the discourse. While a previous mention of the same location does not necessarily mean that the adverb is referring anaphorically to that mention, the absence of a previous mention does mean that the adverb cannot be anaphoric, and this therefore gives some indication of the anaphoric nature of these adverbs. As deixis and anaphora are most readily applied to physical locations, we restrict our analysis of anaphoric and deictic uses of these adverbs to those cases where they are used with spatial meanings. Table 4 provides the distribution of the adverbs according to previous mention, and demonstrates a clear distinction between the proximal (aquí and acá) and non-proximal forms (ahí, allí and allá): the former are used proportionally less often in contexts where the location has been referred to previously (38% and 29% respectively), and the latter are used proportionally more often where it has been (62%, 61% and 73% respectively)4.

4

These figures include tokens that have a prior mention in the discourse and which are also present in the physical environment of the speech context (and thus are potentially deictic as well). A similarly contrastive distribution is obtained if we consider only those tokens that are not present in the speech environment but have been mentioned in the preceding discourse: we have only a single token of the proximals (one token of acá) which is mentioned previously and is not present, while the non-proximals are used over 50% of the time to refer to places with a previous discourse mention and no presence in the speech environment (ahí 51%, allí 58%, allá 64%).

MURIEL/OROZCO

31/5/12

74

14:24

Página 74

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

TABLE 4 Frequencies of tokens with previous mention in discourse (spatial uses only, N = 832)5 With previous mention

No previous mention

Unclear6

Total

aquí

104

38%

151

56%

17

6%

272

acá

22

29%

48

64%

5

7%

75

ahí

134

62%

69

32%

13

6%

216

allí

20

61%

12

36%

1

3%

33

allá

172

73%

53

22%

11

5%

236

Total

452

55%

333

40%

47

5%

832

These results suggest that allá is the most highly anaphoric adverb in these data: while allá is used over three times as often in contexts where the location had been mentioned in the preceding discourse than when it has not been (172 vs. 53), ahí and allí are used slightly less than 1.5 times as often (134 vs. 69 and 20 vs. 12 respectively). This appears to contrast with some other varieties, where ahí has been found to be the most highly anaphoric adverb (Sedano 1996, 2000) or all -i adverbs have been considered to be more highly anaphoric (Molho 1992). Turning now to deixis, Table 5 gives the distribution of the adverbs according to physical presence in the speech environment. Given this definition of deictic, the proximal terms are rarely used to refer to a location that is not in sight (one token of aquí and four of acá), and when they are, they are used in a non-straightforward way, as in (6), where the speaker is making a general statement, and thus the adverb does not refer to any real place. The distal terms are all more likely to be used in a non-deictic sense, but again the results for allá are most marked:

5

Of the total of 1,048 tokens that were initially extracted, 56 were excluded from analysis as the surrounding discourse is unclear, and thus the meaning cannot be conclusively determined. Of the remaining 992 tokens, 160 are used in non-spatial contexts (see section 7 for details). 6 This refers to those tokens where it is unclear whether the referent of the adverb under study has a preceding coreferential use or not, including tokens used too early in the recording to tell, and tokens for which the referent of either the target token or of a potential coreferential earlier mention is unclear.

MURIEL/OROZCO

31/5/12

14:24

Página 75

75

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

TABLE 5 Frequencies of deictic tokens (spatial uses only, N = 832) Deictic

Non-deictic

Unclear7

Total

aquí

271

100%

1

0%

—

272

acá

71

95%

4

5%

—

75

ahí

80

37%

132

61%

4

2%

216

allí

10

30%

19

58%

4

12%

33

allá

38

16%

186

79%

12

5%

236

Total

470

56%

342

41%

20

2%

832

non-deictic uses of allá are nearly five times as frequent as deictic uses (186 vs. 38), while for ahí and allí they are less than twice as frequent (132 vs. 80 and 19 vs. 10 respectively). (6) Re what used to happen with parents and children in general, when on Sundays they did completely different things C: Ellos iban por allá, y los hijos quedaban acá. [almuerzo 1683] Celia: ‘They ((parents)) went allá, and the children stayed acá.’

These results once again demonstrate a lack of systematicity in the behavior of the -i and -a series, the key distinction being one based on proximal vs. nonproximal terms: regardless of the series to which they belong, the proximals are significantly less likely to refer to a location that has been mentioned previously in the discourse than the non-proximal terms (Chi-squared = 77.096, p < 0.0001) and are significantly more likely to refer to a location that is present in the physical context of the speech situation (Chi-squared = 506.645, p < 0.0001).

6. Two series: -i vs. -a There has been much discussion of the differences between the two series, and in general two explanations have been put forward. One is that the -i series refers to 7 This refers to those tokens where it is unclear from the discourse context whether the referent is in sight or not.

MURIEL/OROZCO

76

31/5/12

14:24

Página 76

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

location, while the -a series refers to movement (Bello 1978, Nilsson 1983, Salvá 1830/1988). The second (not entirely unrelated to the first) is that the -i series refers to a precise location, while the -a series has a more approximate sense (Hanssen 1913: 261, Kany 1975: 269, Miyoshi 1999, Moliner 1984, RAE 1973, Sacks 1987, Schmidely 1975, Sedano 1994b). These different meaning relations are reported to be hangovers from the Latin system from which the adverbs arose: hic/illic (from which the Spanish -i series developed) marked ‘location’ or ‘rest’; and hac/illac (Spanish -a adverbs) indicated some idea of ‘adjacency’ (Hanssen 1913: 261, Sacks 1954: 264, Sacks 1987: 290, Sedano 1999b: 343). However, this distinction between these series is considered to have been lost in Vulgar Latin and to no longer be encoded in the modern-day Romance languages (Hanssen 1913: 262). Nevertheless, the -i and -a adverbs are usually treated as belonging to two functionally distinct series with a semantic distinction based on those that existed in Latin, so we explore it further here through consideration of co-occurring elements in the discourse context, based on the understanding that the semantics of the adverbs are borne out in the contexts in which they occur.

6.1. LOCATION VS. MOVEMENT As a first step towards identifying the semantics of the adverbs, all adverbs used with a spatial sense were analyzed as having either a ‘movement’ or a ‘location’ sense. The results are given in Table 6. As can be seen, all adverbs are used more in a locative sense, with this use accounting for 86% of the data overall. Thus, it is clearly not the case that the -a series is reserved for movement. Nevertheless, there is a significantly higher proportion of use of the locative sense with the -i series (ranging from 88% to 94%) than for the -a series (69% and 79%, Chisquared = 623.053, p < 0.0001), thus partially supporting this understanding. When we look more closely at the distribution of the adverbs in the data, however, we find that this is in large part due to certain frequent collocations. First, let us consider the use of the preposition para, described by Sedano (1994b: 228) as marking “movement towards a goal”, and thus associated with a movement sense. The 34 tokens of para in the data occur exclusively with acá and allá, and account for 21% (16/75) of the tokens of acá and 8% (18/236) of the tokens of allá (see example (7)). Furthermore, para acá and para allá account for a large proportion of the movement senses of each adverb (70%, 16/23, of acá encoding movement and 37%, 18/49, of allá encoding movement). If we exclude the use of para + acá / allá, the proportion of use of acá and allá that encode location rises to 88% (52/59) and 85% (186/218) respectively, thus

MURIEL/OROZCO

31/5/12

14:24

Página 77

77

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

TABLE 6 Distribution of location and movement senses (spatial uses only, N = 832) aquí

location movement

ahí

N

%

238

88%

34

12%

272

N

allí %

204 94% 12 216

6%

acá

allá

total

N

%

N

%

N

31

94%

52

69%

187

79% 712 86%

2

6%

23

31%

49

21% 120 14%

236

832

33

75

%

N

%

much closer to that of the adverbs from the -i series. In addition, para acá and para allá often undergo phonological reduction to be produced as [paka] and [paa] respectively. This, coupled with their high frequency of use and the lack of use of para with adverbs from the -i series, suggests that para acá and para allá function as fixed constructions or prefabs (Bybee 2006), which contribute to the association between the -a series and movement. (7) P:

.. Trae pues para la casa de ella, Y para acá para la casa de nosotros. [paras 160] Paco: ‘She brings ((flowers)) to her house and para acá to our house.’

We also find certain ‘adverb + verb’ combinations occurring repeatedly in the data. Here we analyze only those 424 adverb tokens that occurred as complements of the verb (rather than adjuncts), where complements mark the direction of a motion verb (such as ir(se) ‘go’, venir ‘come’), or the location of a locative verb (estar ‘be’, haber ‘there is’), a placement verb (dejar ‘leave’, poner ‘put’) or a posture verb (sentarse ‘sit’, pararse ‘stand’), and adjuncts play a peripheral role in the event. In the following examples we can see the distinction between complements and adjuncts: allá in (8) is an argument of the verb dejar, while acá in (9) provides more tangential information. (8) D: Dejaste el diccionario allá? Diana:‘Did you leave the dictionary allá?’

[comida 377]

(9) D: Qué celebran acá? Cumpleaños? Dora: ‘What are they celebrating acá? A birthday?’

[venta 1821]

Table 7 gives the distribution for the use of the adverbs across the different verbal types with which they occur as complements. The breakdown is given just

MURIEL/OROZCO

31/5/12

78

14:24

Página 78

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

TABLE 7 Distribution by adverb of most frequent verbs to occur with these adverbs as complements (N = 424)8 aquí N

ahí %

N

allí

acá

allá

total

%

N

%

N

%

N

%

N

%

0

–

17

4%

28 21%

40

9%

15

11%

89 21%

venir

7

6%

2

1%

0

–

8

28%

ir(se)

7

6%

3

2%

1

5%

1

3%

estar

37 32%

29 23%

5

26%

3

10%

haber

15

13%

15 12%

1

5%

1

3%

25 19%

57 13%

8

7%

11

9%

2

11%

3

10%

26 20%

50 12%

other

44

37%

67 53%

10

53%

13

45%

37

28% 171 40%

Total

118

131

424

ser

127

19

29

for the two most frequent motion verbs (venir ‘come’ and ir(se)‘go’), and the three most frequent locative verbs (the copulas estar and ser, and the existential haber) that occur with these adverbs in the data. We observe a strong tendency for acá and allá to occur with venir and ir(se) respectively: venir accounts for 28% of the uses of acá as a complement, and ir(se) 21% of the uses of allá, while these verbs account for minimal uses of the other adverbs. The frequent co-occurrence of venir and ir with acá and allá is in accordance with the notion that the -a series expresses movement. This is further supported by the fact that three out of the seven tokens of venir + aquí are in fact used in a locative (or ‘fictive motion’) sense, as in (10), and the same is true for four out of the seven tokens of ir + aquí. (10) D: Claro aquí -- aquí viene la ve=na. Doris: ‘Of course aquí -- the vein goes aquí.’

[chuzo 321]

It should be noted that seven of the eight tokens of venir + acá and 10 out of the 28 tokens of ir(se) + allá occur with the preposition para (venir para acá, ir

8 As the complement / adjunct distinction functions in the same way for both spatial and non-spatial uses, all are included here.

MURIEL/OROZCO

31/5/12

14:24

Página 79

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

79

para allá), and thus it is not the verb alone that favors the adverb, but the preposition also plays a role. We find support for the association between the -i series and location in the behavior of estar, which accounts for 32% of the uses of aquí, 23% of ahí and 26% of allí as a complement, but only approximately 10% of acá and allá. For haber and ser, illustrated in (11), however, we find that the favored adverb is allá, which goes against the patterning predicted by an association between the -i series and location. These two verbs account for approximately 20% of the uses of allá as a complement. On the other hand, acá occurs 3% of the time with haber and 10% of the time with ser, rates similar to, or less than, those observed for the -i series with these verbs. (11) A: F: A: Ángela: Fabio: Ángela:

Conocen la Carbonera? .. No. .. A- -- allá es un poquito carito, pero, !huy. ‘Do you know (the restaurant) la Carbonera?’ ‘No.’ ‘Allá is a little expensive, but, boy!’

[comida 583]

6.2. PRECISE VS. APPROXIMATE LOCATION We now move on to the second key proposal in the literature, namely that the -i adverbs represent a more precise location and the -a adverbs a more approximate location. In fact, ‘precise’ versus ‘approximate’ location is often unmeasurable in discourse data, because the scope of the adverbs is not always apparent from the discourse context (cf. Sedano 1994b: 320), and even when it is, it often cannot be determined whether the speaker had intended a precise or approximate sense. Once again we can look for co-occurring features in the discourse environment that express more precise or general meanings in order to determine whether certain adverbs are favored in more ‘precise’ or ‘approximate’ contexts. It has been proposed that más ‘more’ implies approximation or the lack of clear boundaries (Miyoshi 1999: 362). This is apparent in the following example. (12) A:

.. no es exactamente allá sino, Más acá. Más hacia el sur. Pero sí. más o menos. Ángela: ‘It’s not exactly allá, but rather, more acá. More towards the south. But yes. More or less.’

[install 594]

Más only occurs with adverbs from the -a series in our data, as has also been reported for other varieties (Nilsson 1983, RAE 1973, Sacks 1987, Schmidely

MURIEL/OROZCO

31/5/12

80

14:24

Página 80

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

1975, Sedano 1994b). However our data present just five tokens of más + adverb, and thus the numbers are too low to draw any conclusions. Another ‘approximate’ context is with the preposition por to mean ‘around’ as in (5) and (6) above and in (13) below. If -a adverbs are less precise, we would expect them to appear with the preposition por more than the -i adverbs. Although the numbers are low and thus only suggestive, we do see some evidence for a favoring of por with acá and allá as compared with the corresponding -i terms: por acá accounts for 7% (5/75) of the tokens of this adverb used in a spatial sense, but only 4% (10/272) of aquí; and it accounts for 12% (29/236) of the tokens of allá, but does not occur at all with allí. Por does, however, account for 7% (16/216) of the spatial tokens of ahí, and thus ahí does not show the same disfavoring of this preposition that the other two -i terms do. Furthermore, as we discuss below (Section 7.4), por ahí is also frequently used with a non-spatial sense (N = 26), and when all tokens of por ahí are considered, this collocation accounts for 12% (42/365) of the tokens of ahí. (13) A: .. Me pone esto por ahí. Ángela: ‘Put this por ahí for me.’

[breakfast 868]

In sum, we do observe some differences in the patterning of the adverbs from the -i and -a series that appear to correspond with the notion of the -i series marking a location or having a more precise sense (their co-occurrence with the verb estar, the low rate of use of por with aquí and allí) and the -a series marking movement or having a more approximate sense (their co-occurrence with the prepositions para and por, the verbs venir and ir(se), and the adverb más). However, we also see several exceptions to this (the high rate of use of all adverbs from both the -i and -a series with a location sense, the co-occurrence of allá with haber and ser and the high frequency of por ahí). This leads us to propose that, although these co-occurrence patterns may originally have been attributable to a meaning distinction between the two series, in contemporary Spanish they no longer reflect a difference of meaning per se, but rather represent conventionalized collocations. That is, the reason that speakers prefer an expression such as ven pa’cá ‘come here’ or estoy aquí ‘I am here’ is not solely because of the semantics of the adverbs, but because this has become the conventionalized way of expressing such notions.

7. Functions In order to complete our overview of the locative adverbs, we now move on to consider the functions they play. As was seen in Table 1, ahí is the most frequent

MURIEL/OROZCO

31/5/12

14:24

Página 81

81

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

TABLE 8 Frequencies and functions of forms (N = 992) aquí

%

ahí

%

allí

%

acá

%

allá

%

272

99

216

59

33

83

75

100

236

100

situational

3

1

88

24

6

15

—

—

—

—

temporal

1

—

30

8

1

3

—

—

—

—

emotive

—

—

10

3

—

—

—

—

—

—

approximative

—

—

21

6

—

—

—

—

—

—

276

100

365

100

40

100

75

100

236

100

spatial

adverb overall, but when we consider the different functions with which they are used (provided in Table 8), we see that aquí, ahí and allá occur at a similar rate with a spatial sense (272, 216 and 236 respectively). The spatial uses, however, account for only 59% of the total number of tokens of ahí, while they account for 83% of allí and for aquí, acá and allá are (near) categorical. In a comparison of ahí and allí in sociolinguistic interviews from Caracas, Sedano (1999a) similarly finds that ahí is used proportionally more with a non-spatial (what she terms ‘pragmatic’) sense than allí, although for both adverbs she observes a much lower rate of non-spatial uses than in our data (13% (213/1,697) for ahí and 2% (5/225) for allí). Whether this is representative of a difference between the two regional varieties, the genres studied or different applications of the terms remains an open question. The majority of examples given so far in the paper represent the spatial uses, which are also those uses that have received the most attention in the literature. Here, we will focus on the less commonly discussed non-spatial uses, namely, what we have termed ‘situational’, ‘temporal’, ‘emotive’ and ‘approximative’. It is interesting to note that the situational uses, which are the closest in meaning to the spatial uses (as we outline below), are the most frequent of the extensions away from the spatial sense for ahí, and also account for almost all of the nonspatial uses for aquí and allí. Furthermore, while the situational uses occur in a broad range of contexts, the other extensions we see are restricted to a limited set of relatively defined contexts. Thus rather than the adverbs themselves extending into these other domains, it is specific constructions that facilitate these extensions (cf. Bybee 2003, Traugott 2003, on grammaticization in constructions). As we have seen above, the repeated use of specific collocations plays a role in the overall patterning of the adverbs.

MURIEL/OROZCO

31/5/12

82

14:24

Página 82

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

7.1. SITUATIONAL What we have termed ‘situational’ are those uses that have moved away from a locative sense in that they include anaphoric reference to a previously mentioned situation, which may or may not correspond to a specific location. Often in these cases, the adverb retains a possible spatial interpretation, but refers more generally to a whole situation rather than simply the location where it occurs. Thus in example (14), Ángela is talking with a friend about conversations with her husband about sensitive issues that she had recorded. Her interlocutor suggests that Ángela’s husband had not realized he was being recorded, to which Ángela responds that he did but that they continued talking with the recorder on regardless. It could be proposed that ahí makes reference to the location where the conversations took place, understood in a broad sense, because the recordings mentioned were made at multiple locations. At the same time, ahí refers to the apparently humorous situation of being recorded (note the laughter following). (14) Re recording conversations (including arguments) with her husband A: él si se daba cuenta. Pero seguíamos hablando ahí. @@[@@@] R: [@@@] [arepas 186] Ángela: ‘he did realize ((that he was being recorded)). But we kept talking ahí. @@[@@@]’ Rosa: ‘[@@@]’

While the few situational uses we see of aquí and allí all similarly allow a spatial interpretation, for ahí there are examples referring more clearly just to a situation, where a spatial interpretation is not possible. One example can be seen in (15), where ahí refers to the state of being a drug addict. (15) M: se dejó llevar del vicio, y hasta ahí llegó. [campaign 849] Milena: ‘he let himself be taken along by the addiction, and ahí is as far as he got.’

A further extension can be seen in the use of ahí to refer to the way in which something is done, often paraphrasable with así. In example (16), Santi is describing the faces a comedian makes when imitating the Colombian president; ahí does not mean that the comedian raises his mouth to any particular place, but that he raises it in a certain way. Note the parallelism in Santi’s two lines, with ahí in the second line being used in a very similar fashion to así in the first (and in Ángela’s further comment).

MURIEL/OROZCO

31/5/12

14:24

Página 83

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

(16) S: A: Santi: Ángela:

hace esa cara así=, .. sube ese pico ahí. Y e- -- .. hace así, la lengua, ‘he makes a face like this (así), lifts his lip ahí.’ ‘And he goes like this (así), sticking out his tongue,’

83

[pizza 1190]

7.2. TEMPORAL The semantic extension from space to time is a common one cross-linguistically, which has been widely discussed in the grammaticalization literature (e.g. Bybee/Perkins/Pagliuca 1994, Hopper/Traugott 1993), and which has also been noted for the locative adverbs in Spanish (Carbonero Cano 1979, Kany 1975: 270, Nilsson 1983: 262, Schmidely 1975: 247, Sedano 1999a). Based on our data, the temporal use appears to be an extension from the situational (rather than spatial) use, as a situational interpretation is available alongside the temporal interpretation in many cases. We also observe that all temporal uses occur in highly lexically specified constructions (namely ahí mismo, ahí sí, ahí vemos, de ahí (a X), and por ahí). We discuss the first four of these below, and consider por ahí in Section 7.4, because, due to its high frequency and varied uses, it warrants independent discussion. The construction ahí mismo occurs in the data with a spatial sense, as an emphatic meaning something like ‘right there,’ and with a temporal sense, illustrated in example (17). Here, Santi states that if you miss a loan payment, you are sued right away. Note that this is closely tied to the situational sense, as it also implies something like ‘in the circumstances that you miss a payment, …’. (17) S: se cuelga en una letra, .. Y carta ahí mismo. Santi: ‘You miss one payment, and you get sued ahí mismo.’

[familia 549]

Ahí sí occurs with an emphatic spatial, situational and temporal sense, often interrelated. Example (18) illustrates the temporal (related to a situational) sense. Here the participants are discussing all the work involved in a political campaign, and Omar states that if the candidate wins, then (at that time, and in that situation), the work really starts. (18) O: R: O: Omar: Rocío: Omar:

Pero que si gana, .. ya la cosa se pone -.. [Ah hah]. .. [!peor]. que ahí sí, ahí sí, es donde comienzan a trabajar. [Tumaco 411] ‘But if you win ((the election)), things get --’ ‘Ah-hah.’ ‘worse. Ahí sí, ahí sí is when they ((politicians)) start working.’

MURIEL/OROZCO

31/5/12

84

14:24

Página 84

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

Ahí vemos (also ahí verás) is a construction that has moved far from any spatial interpretation (cf. Gutiérrez-Rexach 2001: 168). It is used with a temporal sense to refer back to a prior mentioned time, but again, a time that corresponds to a certain situation. In (19), Santi is saying that it is only at the (hypothetical) time that his wife gets pregnant that they will decide what to do about the care of the child. (19) S: ahí vemos cómo hacemos. Santi: ‘ahí vemos what we’ll do’

[almuerzo 1829]

And finally, the construction de ahí (a X) is used to treat a point in time (or a situation) as a starting point for another event or situation. In example (20), ahí refers both to the time when a coconut tree starts producing coconuts, and the situation in which a tree has begun to produce coconuts. (20) O:

Un coco comienza a producir sus cocos a los ocho años. Y de ahí, eso es constante. [contamination 829] Omar: ‘A coconut palm begins to produce coconuts at eight years. And de ahí, it is constant.’

We find both aquí and allí used in related constructions (with de or desde), suggesting that while the other temporal constructions are specific to ahí, here we have a more generalized schema, namely ‘de / desde + locative adverb (a + time).’ This construction has been noted in the literature, indicating that it is also found in other varieties of Spanish (e.g. Kany 1975: 271, Schmidely 1975: 247). (21) On the phone, talking about when he will have a jingle ready O: Pues, por la noche? de aquí a la noche tenemos algo. [campaign 298] Omar: ‘Well, by tonight? de aquí to tonight we’ll have something.’

7.3. EMOTIVE It is not difficult to trace a possible path of evolution of the extensions away from the spatial sense that we have seen so far: we have noted the link between spatial and situational uses, and that between situational and temporal. Perhaps because of the obvious semantic links, we find similar extensions not just for ahí but for other adverbs as well. However, what we are classifying as the emotive use is a much more marked extension, and one which we only find for ahí. The adverb is found in a construction where it immediately follows a noun or adjective, or occasionally

MURIEL/OROZCO

31/5/12

14:24

Página 85

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

85

immediately precedes an adjective. Here, ahí is used to express the speaker’s emotional reaction to the topic being discussed. This has been described in various ways in the past: Kany (1975: 270) terms this use an “expletive or emphasizing particle”; Sedano (1999a: 55) describes it as being used to mark “una desvalorización de la entidad o de la acción sobre la que indica ahí” ‘a devaluing of the entity or action to which ahí refers’. It should be noted that there are several tokens in the data that appear to express emotional overtones alongside one of the other meanings we have discussed here (as in (14) and (16), for example). We coded as emotive only those tokens that do not allow for a spatial, situational or temporal interpretation, and leave the discussion of ambiguous cases and consideration of the path of development of this function for future research. The emotion expressed in this use is always strong, and most commonly negative. This can be seen as in (22), where Santi is discussing an actor whom he does not like at all. However, in distinction to Sedano’s (1999a) understanding of the meaning of this construction, in our data it is also used with a positive emotional or even affectionate reading, as in (23). This suggests, then, that the positive or negative overtone comes from the discourse context, and ahí is playing a role more akin to that of a discourse marker that has undergone bleaching, not just of the spatial meaning, but also of any specific emotional reaction that it may have originally had. (22) S:

... un Cubanito ahí. .. chimbo. ... Un actorcito Cubano. [breakfast 169] Santi: ‘A little Cuban ahí. A jerk. A little Cuban actor.’

(23) N:

él es toda buena gente, ahí. Te trata como, por igual, a todo el mundo, [estudios 1151] Nury: ‘He’s all nice ahí. He treats everyone the same way,’

7.4. APPROXIMATIVE As discussed above, one of the functions of the preposition por ‘through, around’ in Spanish, in combination with a location, is to indicate something approximate, ‘in the area of.’ Use of a preposition meaning ‘around’ in this way is relatively frequent cross-linguistically (Haspelmath 1997: 48), and, as was discussed in Section 6.2, por occurs with several of the locative adverbs in Spanish. However, it is only with por ahí that we find an extension beyond spatial contexts in our data. One common use of non-spatial por ahí is preceding or following a temporal expression to mark it as non-precise.

MURIEL/OROZCO

31/5/12

86

14:24

Página 86

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

(24) A: Yo vengo aquí a las dos, por ahí. Ángela: ‘I will come here at two o’clock, por ahí.’

[Colombo 952]

Por ahí is also used in a similar way more broadly in conjunction with any number, not explicitly a time, as in (25). (25) P: Vos parecés por ahí de veinticuatro. Patricia: ‘You look por ahí twenty-four.’

[comida 128]

Finally, there are two tokens in our data where por ahí has extended beyond the association with space, time and numbers. These two tokens are produced together in a repair by the same speaker, and are illustrated in example (26), where David is commenting on the idea that strong currents have reportedly brought penguins to Colombia from the polar regions. He suggests, the punchline of a joke, that rather than penguins, maybe what has been brought by the currents is ‘nuns or something like that:’ (26) D:

O serán por ahí, monjas -- O serán por ahí monjas de= -[contamination 704] David: ‘Or might they be por ahí nuns -- Or might they be por ahí nuns from --’

It is hard to fully interpret precisely what is meant by por ahí here. It may be modifying the noun monjas ‘nuns,’ just as it modifies a numeral in (24) and (25). However there are suggestions that, in some varieties of Caribbean Spanish, por ahí has grammaticized as a clausal marker, undergoing phonological reduction to be produced with a diphthong [poraj], and adopting the meaning of ‘perhaps’ (Rafael Orozco, p.c.). It may be that the use we observe in (26) is evidence of a step in that direction in this variety.

8. Conclusion In this paper we have looked at the five locational adverbs in the spoken Spanish of Cali, Colombia, both in terms of their behavior within this variety, and in comparison with other dialects and genres where data is available. Our comparisons raise many questions about how widespread the patterning we observe may be. While we do find several commonalities across dialects (e.g. skewed distribution, constructions such as más acá / allá or para acá / allá, non-spatial uses of ahí), and have also noted some differences from previous findings (anaphoric use of allá, the very low rate of use of allí), much of what we have noted regard-

MURIEL/OROZCO

31/5/12

14:24

Página 87

LOCATIONAL ADVERBS IN COLOMBIAN SPANISH CONVERSATION

87

ing the patterning of the adverbs in our data has received little attention in the literature. This highlights the importance of taking into account the possible effects of regional variation and differences in genre, particularly with spoken varieties of a language such as Spanish, with a wide geographical range. While formally the five locational adverbs may represent two different series (-i and-a), the use of these adverbs in these data does not suggest any coherence in this categorization. This is seen in the distribution patterns of the adverbs as well as in their functions. Looking purely in terms of overall frequency, the -i adverb aquí is overwhelmingly preferred as the proximal term, while the -a adverb allá is the preferred distal term. The same forms are favored by speakers in Bogotá and Caracas, but not in Mexico City, Buenos Aires or Madrid, suggesting that this may be a feature of Colombian and Venezuelan Spanish. In terms of the functions of the -i and -a series, we again have not found a neat distinction in these data, and we propose that the distinctions that have been noted in the literature are to a large degree attributable to the constructions in which the forms repeatedly occur. Future research could quantify this in larger corpora, which would allow for more detailed frequency analyses, taking into account transitional probabilities, as well as the set of factors that jointly account for speakers’ use of the forms. We propose that the association between the -a series and ‘movement’ or ‘approximate location’ comes about in part from the co-occurrence of acá and allá with the preposition para, the adverb más and the verbs venir and ir(se), while the association of the -i series and ‘location’ or ‘precise location’ comes about from the lack of use of these adverbs in these environments, alongside their use with estar. These constructions may be retentions of older usage patterns from Vulgar Latin where such meaning distinctions existed, retentions which we can attribute to the high frequency of use of such constructions (cf. Bybee/Thompson 1997, on the conserving effect of high token frequency). Further weakening the possibility of any clear functional distinction between the -i and -a series is the behavior of ahí, which is the ‘odd man out’ among these adverbs in our data. It is not restricted to a medial sense when used spatially, and it has a very wide range of non-spatial uses, much more so than the other adverbs. Once again, we find that collocations play a major role in this patterning, as the majority of the non-spatial uses in the data are not simply non-spatial interpretations of the adverb alone, but rather of the adverb in certain repeatedly occurring collocations (e.g. ahí sí, ahí vemos / verás, de aquí / ahí / allí (a X)). This study thus contributes to an accumulating body of research which reveals a role for frequent lexical items and collocations or chunks in the patterning of grammar. It demonstrates that the behavior of these adverbs is in part determined by the constructions in which they repeatedly occur, which represent conventionalized ways of saying things.

MURIEL/OROZCO

31/5/12

88

14:24

Página 88

CATHERINE E. TRAVIS/TIMOTHY JOWAN CURNOW

APPENDIX Transcription conventions (Du Bois/Schuetze-Coburn/Cumming/Paolino 1993) .

final intonation contour

...

medium pause (> 0.7 secs)

,

continuing intonation contour

..

short pause (about 0.5 secs)

?

appeal intonation contour

…(N.0)

timed pause (N seconds)

--

truncated intonation contour

[]

speech overlap

-

truncated word

@

one syllable of laughter

=

lengthened syllable

words spoken while laughing

!

word produced emphatically

(( ))

researcher’s comment

Capital initial

restart

MURIEL/OROZCO

31/5/12

14:24

Página 89

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI, COLOMBIA: O N E VA R I A B L E O R T W O ? EARL K. BROWN/ESTHER L. BROWN

1. Introduction The variable aspiration and deletion of syllable- and word-final /s/, characteristic of multiple varieties of Spanish, is a well-studied phonological phenomenon. This reduction is illustrated in examples (1) and (2)1. (1) Para bu[h]car en el diccionario. ‘In order to search in the dictionary.’ [diccionario] (2) Yo tenía muchos amigo[Ø] de Gringolandia. ‘I had many friends from Gringoland.’ [install]

This type of /s/ variation has been examined in several varieties of Colombian Spanish (Lafford 1986; Earl K. Brown 2009a; File-Muriel 2009). Within Colombian national boundaries, two macrozones have been identified, based largely upon different realizations of syllable- and word-final /s/: lowland (coastal) regions with rampant final /s/ reduction, and the more conservative /s/-retaining varieties of the interior highlands (Montes Giraldo 1982; Flórez 1973; 1978). Colombian Spanish has also been characterized, however, as being a variety that evidences reduction not only in syllable-final position, but in syllable-initial position as well (cf. Hualde 1991; Cuervo 1939; Flórez 1973). This type of reduction can be seen in examples (3) and (4), also taken from our corpus. (3) Me está doliendo la cabe[h]a horrible. ‘My head hurts horribly.’ [calima] (4) Le gusta la [Ø]ebolla. ‘S/he likes onions.’ [empanada]

In fact, Lipski (1994: 209) notes that some varieties of Colombian Spanish are “unique in the Spanish-speaking world in reducing /s/ more frequently in syl-

1

The examples are taken from the corpus used in this study, with the title of the corpus file indicated in brackets. We thank Catherine Travis for her generosity in making this corpus available to us for this paper.

MURIEL/OROZCO

31/5/12

90

14:24

Página 90

EARL K. BROWN/ESTHER L. BROWN

lable-initial than in syllable-final position.” This phonological reduction has not gone unnoticed, but there has been a “scarcity of available studies, especially those of a comprehensive nature, on which to base … affirmations” regarding the nature and extension of this variable reductive process (Obaid 1973: 63). Although large-scale quantitative analyses have been conducted on other dialects (e.g. López Chávez 1977; Esther L. Brown/Torres Cacoullos 2002; Esther L. Brown 2005), very little is known about the linguistic factors that constrain variation of initial /s/ in Colombia. Indeed, we would argue that it continues to be the case that initial /s/ reduction as an object of study has not received the attention it deserves (cf. Obaid 1973: 63) and very little is known about this linguistic variable in general. Through an analysis of 11,517 tokens of /s/ in the spoken Spanish of 38 speakers from Cali, Colombia, our study elucidates the linguistic factors that constrain the realizations of /s/ in all syllable and word positions in this variety. Furthermore, by examining four separate syllable and word positions, these data are uniquely able to address the question of whether initial /s/ reduction is a straight-forward extension of processes occurring in syllable-final position, as has been proposed (e.g. Méndez Dosuna 1987; 1996; Penny 2000), or whether weakening processes such as /s/ reduction can begin in syllable-initial position (e.g. Ferguson 1990; Mason 1994; Esther L. Brown/Torres Cacoullos 2002; Esther L. Brown 2005).

2. Background Despite being dialectally noted by many, and considered by some to be “the most interesting phonological phenomenon taking place in the Spanish language” (Obaid 1973: 63), initial /s/ reduction has generally been treated by researchers as sporadic (cf. López Scott 1983), and perhaps anomalous (cf. Hualde 1991: 56). In fact, it has been the case that “the dropping or aspiration of the /s/ phoneme in Spanish is generally regarded as occurring only in syllable-final and word-final positions….” (Resnick 1975: 13, emphasis ours). Certainly a survey of /s/ reduction studies would lead one to such a conclusion. Bibliographic indexes are replete with valuable studies of /s/ variation in syllable-final position. In stark contrast to this abundance of research is the paucity of available studies on /s/ realizations in syllable-initial position. Notwithstanding the lack of empirical analyses, there is documented mention of Colombian Spanish as a variety with frequent reduction of syllable- and wordinitial /s/ (e.g. Resnick 1975: 13; Hualde 1991: 55-56; Lipski 1994: 209). As Flórez (1973: 83) notes:

MURIEL/OROZCO

31/5/12

14:24

Página 91

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

91

La s intervocálica, con articulación aspirada, débilmente faríngea, se oye hoy en diversas palabras del habla espontánea, rápida y descuidada de muchos colombianos en diversas regiones del país, incluyendo la ciudad de Bogotá y el discurso de locutores de radio y televisión. Algunos ejemplos del uso actual: nojotros, diejisiete, nejesita, …sí jeñor, la jeñora, no jabemos, etc. Este fenómeno … creo que se está desarrollando y extendiendo en Colombia. (emphasis ours)

The extent to which initial /s/ reduction has developed or spread in Colombian Spanish is unknown owing to the lack of research on the subject. A primary goal of this investigation, therefore, is to quantify and describe the degree of phonological reduction of initial /s/ using quantitative variationist methodology. We will also identify the linguistic variables that significantly constrain, and hence promote and retard, /s/ reduction in this syllable position. A secondary goal of this analysis is to determine the source of this reduction. Many traditional accounts of aspiration of Spanish /s/ assert that /s/ weakening must begin in preconsonantal contexts (Terrell 1977a; 1979; Barrutia/Schwegler 1994: 216). Instances in which reduction occurs phrase-finally or prevocalically stem from analogical extensions of preconsonantal reductive processes. To account for these cases of prevocalic aspiration and deletion, the rule that generates syllable-final /s/ reduction is said to generalize to include more conditioning environments (Terrell 1979) or is re-ordered to apply before resyllabification, and hence produces reduced allophones in a context otherwise unspecified (Harris 1983). Penny (2000: 149-150) describes a hierarchy of /s/ reduction in America that reflects a pattern of extension that emanates from phrase-final and syllable-final, preconsonantal contexts (word-medial, word-final) to later include prevocalic, word-final /s/ and lastly an extension to word-initial position. The hierarchy is reflected in the diffusion pattern proposed by Lipski (1999: 198-199), for example. According to this model, /s/ begins its weakening process in preconsonantal position (Stage 1) in either word-medial or word-final position. Generalizing from this position, the reduction may extend its phonetic context to include absolute final (Stage 2). Stage 3, in which word-final /s/ extends its phonetic context to include prevocalic position, is found in the “phonologically most advanced dialects” (Lipski 1999: 198). Once Stage 3 has been generalized in a dialect, reduction may extend to word-initial (post-vocalic) position (Stage 4). Lastly, “the ultimate generalization would include word-internal /VsV/ combinations, and in some marginal varieties of Spanish, this extension is in its incipient phase” (Lipski 1999: 199). This cline has been tested in other initial /s/ reducing varieties (New Mexico; Chihuahua, Mexico) that are historically closely related. The possible causal

MURIEL/OROZCO

31/5/12

14:24

92

Página 92

EARL K. BROWN/ESTHER L. BROWN

TABLE 1 Proposed diachronic weakening pathway for /s/ [adapted from Lipski (1999)] Stage

Phonological environment

Examples

1

s > h / __C (word-medial, word-final)

los mismos ‘the same ones’

2

s > h / __# (phrase-final)

cantas ‘2sg sing’

3

s > h / __V(word-final, prevocalic)

los amigos ‘friends’

4

s > h / V#__ (word-initial, postvocalic)

la semana ‘week’

5

s > h / V__V (word-medial, intervocalic)

veces ‘times’

relationship between final /s/ reduction and initial /s/ reduction, which was not sustained in previous analyses (Esther L. Brown/Torres Cacoullos 2002; Esther L. Brown 2005), has not been empirically tested outside of those varieties. In this current analysis, therefore, using variationist methodology and a usage-based approach, we will: 1. Quantify and describe the variable realizations and lexical extension of syllable-initial /s/ (word-initial, word-medial) in Cali, Colombia. 2. Quantify and describe the variable realizations and lexical extension of syllable-final /s/ (word-medial, word-final) in Cali, Colombia. 3. Identify the linguistic factor groups that significantly constrain realizations of /s/ in syllable-initial and syllable-final positions. 4. Identify the magnitude and direction of effect of the linguistic factors that significantly constrain realizations of /s/ in syllable-initial and syllablefinal positions. 5. Illustrate that syllable-initial /s/ reduction is a separate variable from syllable-final /s/ reduction in Cali, Colombia and, hence, need not be considered exclusively an extension of final reduction.

3. Data and methods The data in this study come from spontaneous conversations between native speakers of Cali, Colombia (cf. Travis 2005). With relatively high levels of initial /s/ reduction, this dialect lends itself well to a quantitative analysis of this phenomenon, as large numbers of tokens are necessary to detect patterns within otherwise apparently unsystematic variation.

MURIEL/OROZCO

31/5/12

14:24

Página 93

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

93

The recordings were made by a university student who recorded friends and family members whenever spontaneous conversations arose. Consequently, these conversations approximate the vernacular perhaps more than any other register, and are therefore essentially free of hypercorrection and irregular patterns of variation (cf. Tagliamonte 2006: 8). The corpus consists of the speech of 38 speakers, 14 men and 24 women, between the ages of 21 and 55 years. The majority of the speakers are considered middle-class Colombians, as many of them hold or are in the process of obtaining a university degree and many work in a related profession, such as university professor or high school teacher. Some exceptions include house cleaners and salespeople, who pertain to the working class. The larger number of female and well-educated speakers is undoubtedly the result of the fact that the research assistant is a female university student. All speakers live in Cali, and were either born there or moved there as children. The length of the conversations ranges from eight to forty minutes. The recordings were made in places where the research assistant happened to be: at her home or that of a family member, in a car traveling, in her or a family member’s kitchen eating a meal, in the university cafeteria, in a restaurant, in the university library, and other places. The conversation topics vary greatly and cover those topics normally discussed between friends and family members: family news, gossip about colleagues, jobs, current events, plans for the weekend or the upcoming holidays, car trouble, computer trouble, food, etc. The total number of words in the corpus is 105,248 (cf. Travis 2005: 13-16 for more information about the corpus used in this study, which is also used in Travis and Curnow, this volume). In order to analyze /s/ reduction in this corpus, a sequential set of tokens of /s/, starting from at least five minutes into each conversation, was coded impressionistically as either maintained or reduced. The choice to code the dependent variable by ear rather than instrumentally was based on the fact that “discrete variants, as the label implies, tend to be easily distinguished and their coding is typically a relatively straightforward process” (Milroy/Gordon 2003: 144). In short, if we heard sibilance, we coded the token as maintained, otherwise, it was coded as reduced. Tokens were passed over if there was so much background noise that the token could not be distinguished with confidence. A large number of tokens was analyzed in this study: N = 11,517 in 2,241 different lexical types. The outlier behavior of entonces made necessary its exclusion because the entire last syllable of this word was usually deleted, as in /entón/, leaving neither a syllable-initial nor a word-final /s/ to analyze (N = 415). Concerning cases of outliers, Tagliamonte (2006: 91) asserts: “When the variable under investigation occurs in a context which is anomalous with respect to the variation of forms within it, these are typically removed from the analysis.”

MURIEL/OROZCO

94

31/5/12

14:24

Página 94

EARL K. BROWN/ESTHER L. BROWN

Research conducted on the reduction of /s/ in both syllable-initial and syllable-final positions in multiple varieties of Spanish has shown that factors such as age, gender, speech style, level of education, and socioeconomic status correlate with varying degrees of reduction (Fontanella de Weinberg 1973a; López Chávez 1977; Cedergren 1978; López Scott 1983; Lafford 1986; Terrell 1986; Cid-Hazard 2003; López Morales 1989; Lipski 1986b; 1990; Dolores Brown 1993; Esther L. Brown 2000). We have no reason to presume that the Spanish of Cali, Colombia would lack such social and stylistic variation with regard to realizations of /s/. Social and stylistic considerations were not the primary motivators in the selection of consultants for this corpus, however, and we therefore do not consider extralinguistic factors in this analysis. The seven independent variables analyzed in this study are: syllable-position (initial, final), word-position (initial, medial, final), the preceding and following phonological contexts, the prosodic stress of the syllable in which /s/ occurs (stressed, unstressed), the length of the word in segments, and the lexical frequency of the word in which /s/ occurs. In order to calculate lexical frequency, the number of occurrences of the individual words was calculated using the same corpus from which the words were extracted. This number was entered into the logistic regression models and thus, word frequency was measured as a gradient variable rather than a categorical one, as would be the case if the tokens were placed into groups based on their general frequency (e.g. high, mid, and low). One of the advantages of analyzing frequency as a gradient variable is that the researcher does not have to make the (oftentimes) arbitrary decision of where to place the dividing lines between the frequency groups (cf. File-Muriel 2010). Rather than the actual (or raw) number of occurrences, the natural logarithm of the number was entered into the logistic regressions. The reason for this decision is that the highest-frequency word types invariably have frequencies substantially higher than the rest of the tokens. This distribution gives the impression that the increase in frequency among the lowest-frequency tokens is negligible. However, there may be significant differences between tokens with raw frequencies of, for example, one and fifty, but this information is overshadowed by the huge differences when compared to the highest-frequency tokens, which have raw frequencies of 1,500. By taking the log of the raw frequencies the huge disparity between the highest-frequency tokens and the rest is assuaged. Gries (2009: 255) points out that word frequencies are usually skewed and that when there are outlier frequencies, “it makes sense ... to log the frequencies.” The following section provides a summary of the results of the quantitative and statistical analyses, as well as a discussion of the theoretical implications of these results.

MURIEL/OROZCO

31/5/12

14:24

Página 95

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

95

4. Results 4.1. SYLLABLE-INITIAL /S/ REDUCTION IN CALI, COLOMBIA (WORD-INITIAL AND WORD-MEDIAL) Overall, reduction in syllable-initial position is 12%. Reduction affects 14% of the lexical types in our data: of the 1,143 lexical types with a word-initial or syllable-initial, word-medial /s/, 158 different words show /s/ reduction. Among the most highly reducing lexical types are nosotros ‘we’ with 92% reduction (N = 38), necesito ‘1sg need’ with 68% reduction (N = 22), señora ‘woman’ with 63% reduction (N = 19), and se ‘3sg clitic pronoun’ at 37% (N = 437). The results of a series of stepwise binary logistic regressions performed with the statistics software R (R Development Core Team) reveal results which most often seem to concur with the findings of the few previous studies on initial /s/ reduction in Spanish. Word- and syllable-initial /s/ is significantly conditioned by the following variables in these data: preceding segment, stress, word frequency, following segment, word position, and word length. See Table 2. The size of the conditioning effect is represented by the order in which the variables are listed from top to bottom. The coefficients give the magnitude of effect of variants in comparison to the application value within each variable; the further away from zero, the stronger the effect of the variant on /s/ reduction. The p-value reports the statistical significance of the difference between the conditioning effect of each variant and the application value. The coefficients that are not statistically significant are struck-through. The coefficient and the p-value of the application values are listed as “NA” as a self-comparison is not applicable. The two gradient variables, word frequency and word length, do not have a “% reduced” listed as there are no groupings (e.g. high versus low) within these variables. The R2 value measures the proportion of variation that is accounted for in the statistical model, with values ranging between zero and one in which more variation is accounted for as the R2 value increases towards one. As seen in the table, the variable with the strongest effect is the previous segment. In comparison with a preceding pause, preceding non-high vowels (/a, e, o/) significantly condition word- and syllable-initial /s/ reduction. In this context, syllable-initial /s/ reduces at a rate of 18%. On the other hand, there is no significant difference between the effect of a preceding pause and a preceding high vowel, nor between a pause and a consonant, and reduction is disfavored in this context (with reduction rates of 4% and 5% in these contexts). The fact that a preceding segment conditions word- and syllable-initial /s/ the most in these data concurs with previous studies of word- and syllable-initial /s/ in New Mexican

MURIEL/OROZCO

31/5/12

14:24

96

Página 96

EARL K. BROWN/ESTHER L. BROWN

TABLE 2 Distribution of word- and syllable-initial /s/ Preceding segment

% reduced

% data

Coefficient

p-value

Pause

4

20

NA

NA

Consonant

4

14

0.078

0.750

High V

5

9

0.228

0.395

18

57

1.628

0.000

Tonic

6

48

NA

NA

Atonic

17

52

0.812

0.000

100

0.246

0.000

4

31

NA

NA

15

69

0.877

0.000

Medial

11.5

54

NA

NA

Initial

12.3

46

0.719

0.000

NA

100

0.109

0.000

N=

6,069

R2 =

0.178

df =

8

p=

0.000

Non-high V Prosodic stress

Word frequency NA Following segment High V Non-high V Word position

Word length

MURIEL/OROZCO

31/5/12

14:24

Página 97

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

97

Spanish (Esther L. Brown 2005) and in the Spanish of rural Chihuahua, Mexico (Esther L. Brown/Torres Cacoullos 2002; 2003). Prosodic stress also significantly conditions word- and syllable-initial /s/, with more reduction in unstressed syllables. When syllable-initial /s/ is realized in an unstressed syllable, the reduction rate is 17%, whereas initial /s/ in a stressed syllable reduces at a lower rate of 6%. This concurs with studies that demonstrate the influence of stress on the reduction versus maintenance of sounds in general. In fact, longer and larger articulatory gestures can be considered part of the definition of stressed syllables (cf. Hualde 2005: 239; Beckman 1999). Thus, it is expected that /s/ in unstressed syllables will have shorter and smaller articulatory gestures and therefore will be perceived more often as reduced. The third most influential variable in the reduction of word- and syllable-initial /s/ is the frequency of the word in which /s/ occurs. This sound is reduced more often as word frequency increases, as indicated by the fact that the coefficient returned by the regression model is greater than zero (0.246). As previously mentioned, this variable was entered into the regression model as a gradient variable, rather than as a categorical one. As such, the model returned a single coefficient and p-value for the variable itself rather than individual variants, as there are no individual variants in gradient variables. This result is in line with the many studies that have noted significant correlations between increased frequency and higher reduction (Bybee 2001; 2002; Pluymaekers/Ernestus/Baayen 2005; Raymond/Dautricourt/Hume 2006; Gahl 2008; Esther L. Brown/Torres Cacoullos 2002; Esther L. Brown/Torres Cacoullos 2003; Earl K. Brown 2009b; File-Muriel 2009). The following segment is the fourth significant variable. Similar to the influence of the preceding segment, word- and syllable-initial /s/ is reduced more often when followed by a non-high vowel (15%) than by a high vowel (4%). The /s/ in Colombia typically has a predorsal articulation that can be alveolar (Montes Giraldo 1982: 37) or dental-alveolar (Flórez 1978: 211). Production of /s/ requires increased tongue height to create the friction necessary to articulate a sibilant. By definition, the high vowels /i/ and /u/ are also produced with an increased tongue height in the oral cavity and do not put pressure on the tongue to move away from the alveolar ridge. Articulation of a non-high vowel, however, creates a more open space in the oral cavity due to the lower tongue height and hence increases the propensity towards a lax articulation of the sibilant, either as aspiration or deletion. Word position is the variable with the fifth strongest conditioning effect: /s/ is reduced more often in word-initial position than in word-medial position. This is the case even though the word that provides the largest number of tokens, and

MURIEL/OROZCO

98

31/5/12

14:24

Página 98

EARL K. BROWN/ESTHER L. BROWN

which has the highest frequency, sí, shows only 1% reduction while the rest of the words with word-initial /s/ have 17% reduction. The lexical item sí is most often used in utterance-initial position (post-pause), which is a phonological context that does not favor reduction. Further, the word-initial /s/ is followed by a high-vowel, is stressed, and is often articulated in isolation. All of these factors may contribute to the exceedingly low reduction-rate for sí, despite its exceptionally high word frequency; it is the highest in our data, with a frequency of 1,774, which compares to frequencies of 975, 782, 397, and 317 (or 55%, 44%, 22%, and 18% as frequent as sí) for the next four most-frequent words: se, eso, así, si. Finally, word length significantly conditions word- and syllable-initial /s/, with increased reduction mirroring increased word length. This result seems to support Terrell’s (1979: 609) assertion about the influence of word length on word-final plural /s/ reduction. He argues that as word-final /s/ reduction increases, “the danger of misunderstanding increases and some effort (unconscious) is made to preserve the phonetic substance of monosyllabic words.” Our results show that reduction of syllable-initial /s/ in Cali, Colombia is significantly conditioned by lexical factors (frequency, word length, word position, stress) and extra-lexical or online contextual factors (phonological context). Reduction rates of initial /s/, at 12% overall, are not as high as other varieties of Spanish with initial /s/ reduction, such as New Mexico with 24% (Esther L. Brown 2005) or Chihuahua with 28% reduction (Esther L. Brown/Torres Cacoullos 2003), but these results do suggest that, despite the geographic separation, syllable-initial /s/ reduction is constrained in nearly identical ways in these varieties of Spanish.

4.2. SYLLABLE- AND WORD-FINAL /S/ REDUCTION IN CALI, COLOMBIA The overall rate of reduction in syllable- and word-final position is 36%. Of the 1,098 lexical types with a word-final or word-medial, syllable-final /s/, 467 different words show /s/ reduction in our data. These 467 words represent 43% of the lexical types. Among the most highly reducing lexical types are desde ‘from’, reduced 84% of the time (N = 19), mismo ‘same’ also reduced 84% of the time (N = 25), vas ‘2sg go’ with a reduction rate of 82% reduction (N = 22), nosotros ‘we’ with 81% reduction (N = 36), todos ‘all’ with 75% reduction (N = 48), and nos ‘1pl clitic pronoun’ with 70% reduction (N = 63). Most of the variables selected as having a significant influence in the reduction of word-initial and syllable-initial, word-medial /s/ also make a significant contribution to the reduction of syllable- and word-final /s/. However, it is impor-

MURIEL/OROZCO

31/5/12

14:24

Página 99

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

99

tant to note the differences in hierarchies of significant variables between the two regressions. This difference in hierarchies suggests that /s/ is conditioned differently in the two syllable positions. The following variables significantly condition /s/ reduction in syllableand word-final position: word position, following segment, prosodic stress, word frequency, and word length. See Table 3. Again, the magnitude of effect of each variable is represented by its position in the table, with the variables with the strongest effect at the top and the variables with the weakest effect on the bottom. The variable that most conditions syllable- and word-final /s/ reduction is the position that /s/ occupies within the word. /s/ is reduced more often in word-final position (43%) than in word-medial position (13%). This result is in line with many other studies of syllable- and word-final /s/ reduction (e.g. Earl K. Brown 2009b) that show increased rates of /s/ reduction in word-final position. The following segment is the second most influential factor group in conditioning realizations of /s/ in syllable-final position. In comparison with a following pause, all other segments (coronal and non-coronal consonants, and high and non-high vowels) favor reduction. The size of the conditioning effect is gauged by comparing the coefficients: the further away from zero, the stronger the effect. As seen in the table, of all the variants, a following non-high vowel favors /s/ reduction the most among the variants. This result differs from most other studies of syllable- and word-final /s/ reduction, which show that a following consonant favors /s/ reduction more than other segments. The factor with the third strongest influence on /s/ reduction is syllabic stress. As expected, syllable- and word-final /s/ is reduced more often in atonic syllables than in tonic ones. /s/ in unstressed syllables reduces at a rate of 36%, while /s/ in stressed syllables has an overall reduction rate of 35%. While small, this difference was selected as significant in the logistic regression. Again, this result concurs with the findings of many other studies of reductive processes in phonology. The fourth variable to significantly condition syllable- and word-final /s/ reduction in these data is word frequency. As anticipated, /s/ is reduced more often in more frequent words, evidenced by the positive coefficient value. This result falls in line with previous research on the reductive nature of frequency (e.g. Hooper/Bybee 1976; Bybee/Scheibman 1999; Esther L. Brown 2004; FileMuriel 2009; Earl K. Brown 2009a). The fifth and final variable to condition syllable- and word-final /s/ reduction is word length: reduction increases as the word length increases. This finding matches the effect of word length on word- and syllable-initial /s/ reduction seen above in Table 2, and with other studies of syllable- and word-final /s/ reduction (Terrell 1979: 603).

MURIEL/OROZCO

31/5/12

14:24

100

Página 100

EARL K. BROWN/ESTHER L. BROWN

TABLE 3 Distribution of syllable- and word-final /s/ Word position

% reduced

% data

Coefficient

p-value

Medial

13

26

NA

NA

Final

43

74

1.939

0.000

Pause

17

14

NA

NA

Consonant

36

68

1.610

0.000

High V

37

5

1.073

0.000

Non-high V

53

13

1.717

0.000

Tonic

35

39

NA

NA

Atonic

36

61

0.237

0.001

Word frequency

NA

100

0.066

0.000

Word length

NA

100

0.062

0.004

Following segment

Stress

Word length N=

5,448

R2 =

0.195

df =

9

p=

0.000

These findings suggest that syllable-final /s/ reduction in Cali, Colombia is constrained in ways that match previous research on the phenomenon. The linguistic factor groups that favor reduction in this study have also been found to favor reduction in other varieties. One notable exception is the finding regarding following phonological context. In Cali, Colombia, word-final /s/ reduction is favored even when the following word starts with a vowel. This contrasts sharply

MURIEL/OROZCO

31/5/12

14:24

Página 101

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

101

with studies that note higher rates of retention in prevocalic contexts. We will return to a discussion of this in the next section. The results presented thus far reveal the rates of /s/ reduction in all four syllable and word positions2 in Cali, Colombia, and give us a complete picture of the extent of aspiration and deletion in this variety. A comparison of Tables 2 and 3 shows certain similarities between initial /s/ and final /s/ reduction. Perhaps not surprisingly, certain factor groups are shown to have similar effects on /s/ reduction initially as well as finally. For example, prosodic stress has the same direction of effect in both analyses: unstressed syllables favor reduction over stressed syllables. However, the comparison of results summarized in Tables 2 and 3 also reveals certain differences. For example, previous phonological environment is the most important conditioning factor group for initial /s/, but was not selected at all for final /s/. What can we infer from these results regarding the notion that initial /s/ reduction is an extension of final /s/ reduction in this variety of Spanish? The following section outlines our attempts to address this question.

4.3. FINAL > INITIAL CONTINUUM If /s/ reduction were progressing down a cline as summarized in Table 1, we might expect rates of reduction to be highest in preconsonantal positions (stage 1), and to decrease successively in subsequent stages. Table 4 shows rates of reduction in Cali, Colombia for each of the phonological positions proposed in the cline. It can be noted that the rates of reduction do not appear to decrease with each purported stage in the cline, and hence do not appear to suggest a pathway of change. This cline also failed to be supported in New Mexico and Chihuahua, Mexico, two other dialects with initial /s/ reduction (Esther L. Brown/Torres Cacoullos 2002; Esther L. Brown 2005). The results in Table 5 summarizing data from a variety of Spanish dialects that are not as closely related historically as the ones previously studied, provide further support of the notion that initial /s/ reduction need not proceed from final /s/ reduction. It has been proposed (Esther L. Brown/Torres Cacoullos 2003: 35) that “syllable initial /s/ reduction will occur in Spanish varieties where: (1) word-final /s/ reduction is greater before a pause and a vowel than before a consonant, and (2) overall final /s/ reduction rates are relatively low”. The Cali data partially support the first (1) part of this assertion: reduction is high in word-final prevocalic

2 That is, (1) word-initial, (2) word-medial, syllable-initial, (3) word-medial, syllablefinal, and (4) word-final.

MURIEL/OROZCO

31/5/12

102

14:24

Página 102

EARL K. BROWN/ESTHER L. BROWN

TABLE 4 Rates of /s/ reduction in Cali Colombian Spanish Stage

Phonological environment

Examples

1

s > h / __C (word-medial)

13% (1436)

2

s > h / __ #C (word-final, preconsonantal)

51% (2354)

2

s > h / __# (phrase-final)

17% (824)

3

s > h / __#V(word-final, prevocalic)

49% (1047)

4

s > h / V#__ (word-initial, post-vocalic)

23% (266/1142)

s > h / V__V (word-medial)

13% (370/2788)

(5)

position (49%). Nevertheless, in Caleño Spanish, rates of reduction are not high in phrase-final position (17%). Prevocalic reduction rates do not exceed preconsonantal reduction rates (51%) as would be expected; however, differences between these reduction rates are not significant. Indeed, a cross-dialectal comparison of reduction in seven varieties of Spanish shows that dialects with noted initial /s/ reduction (Cali, Colombia, New Mexico, Chihuahua) all show no significant differences between word-final preconsonantal /s/ reduction and word-final pre-vocalic /s/ reduction. These data are summarized in Table 5. This lack of significant difference contrasts with other /s/ reducing varieties such as Argentina, Cuba, Puerto Rico, and Venezuela, where, in word-final position, there are significant differences between pre-vocalic and pre-consonantal positions. Preconsonantal word-final /s/ reduces at a significantly higher rate than prevocalic word-final /s/ in these varieties. This suggests that there may be something special about the intervocalic phonological context in word-final position, and further that this lack of significant difference could potentially be used as a diagnostic to predict the existence of initial /s/ reduction in varieties of Spanish. Given the comparatively higher rates of prevocalic word-final /s/ reduction in initial /s/ reducing varieties, if there were a final > initial continuum, it seems likely the word-final, pre-vocalic position would be a key phonological context for the possible extension of reduction to word-initial position; a gateway of change, so to speak. One of the methodological weaknesses of the two regressions reported in Tables 2 and 3 is that the variables are not orthogonal: not every variant analyzed in one regression is analyzed in the other. For example, word-initial /s/ can only

MURIEL/OROZCO

31/5/12

14:24

Página 103

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

103

TABLE 5 Crossdialectal comparison of /s/ reduction in 4 word and phonological contexts in Spanish varieties with and without noted initial /s/ reduction3 Context

Cali, Colombia

___C

New Mexico

Mérida, Venezuela

San Juan, Puerto Rico

13% (1436) 25% (1310) 22% (569)

88% (4150) 97% (1714) 90% (1675)

90% (1737)

___##C

51% (2354) 61% (1875) 42% (656)

89% (5475) 98% (3265) 97% (2769)

97% (2706)

___##V

49% (1047) 62% (962)

47% (221)

12% (2649) 82% (1500) 87% (1189)

91% (1176)

___//

17% (824)

54% (230)

22% (2407) 39% (1776) 58% (777)

82% (903)

44% (811)

Chihuahua, Argentina Mexico

Cuba

New Mexico data taken from Esther L. Brown (2005). Chihuahua data taken from Esther L. Brown/Torres Cacoullos (2002; 2003). Argentina and Cuba data taken from Terrell (1977b; 1978a; 1979), as summarized in Bybee (2001: 140). Venezuela and Puerto Rico data taken from Earl K. Brown (2009b). Differences in reduction rates word-finally __##C vs. __##V: Cali, Colombia p = 0.269, X2 = 1.22 not significant, New Mexico p = 0.4802, X2 = 0.498 not significant, Chihuahua, Mexico p = 0.1823, X2 = 1.78 not significant, Argentina p = 0.001, X2 = 4588.33, Cuba p = 0.001, X2 = 403.02, Mérida, Venezuela p = 0.001, X2 = 162.28, San Juan, Puerto Rico p = 0.001, X2 = 76.08

be followed by a vowel (e.g. señora) whereas word-medial, syllable-final /s/ is only followed by a consonant (hasta), and word-final /s/ can be followed by both vowels and consonants, as well as by pauses (los niños, los otros #). A similar situation holds when analyzing the preceding segment. Word-final /s/ is only preceded by vowels (los niños), while word-initial /s/ can be preceded by vowels, consonants, and pauses (este señor, un señor, # señor). Given this distributional asymmetry, and considering the importance of word-final, prevocalic context revealed in the cross-dialectal comparison summarized in Table 5, we conducted a stepwise binary logistic regression using only the intervocalic data (N = 4,992), which includes VsV in word-initial, word-medial, and word-final contexts. The results of the statistical analysis are summarized in Table 6. If we were to find evidence of an extension process proceeding from final > initial, these tokens of /s/ would, in principle, be ones in which this pathway of change would be most evident. 3

These are varieties that have not been cited repeatedly in the literature as having notable initial /s/ reduction. It is possible that instances of variable /s/ aspiration occurs in initial position, particularly with specific lexical items, but that rates of reduction are not high enough to be noteworthy.

MURIEL/OROZCO

31/5/12

14:24

104

Página 104

EARL K. BROWN/ESTHER L. BROWN

TABLE 6 The conditioning effect of the main-effect variables4 on intervocalic /s/ reduction Variable/variant

Coefficient

p-value

–4.606

0.000

Word position = Initial

NA

NA

Word position = Medial

-1.098

0.194

Word position = Final

2.130

0.015

Stress = Tonic

NA

NA

Stress = Atonic

1.961

0.000

NA

NA

0.921

0.009

NA

NA

Following segment = Non-hi V

2.343

0.000

Frequency

-0.002

0.981

Word length

–0.330

0.004

(Intercept)

Previous segment = Hi V Previous segment = Non-hi V Following segment = Hi V

As is made quite evident from this analysis of intervocalic tokens, the factor with the greatest conditioning effect, word position, makes apparent a significant difference (p = 0.015) between /s/ reduction in word-initial position (V#sV) and /s/ reduction in word-final position (Vs#V). Even when considered in identical phonological contexts, there is a significant difference between these two instances of reduction. Crucially, it is important to also note the lack of a significant difference (p = 0.194) between word-initial /s/ reduction and syllable-initial /s/ reduction in word-medial position (VsV). Thus, it appears that initial /s/ reduction (word-initial and word-medial) is one process that contrasts significantly with intervocalic /s/ reduction in word-final position. What might cause word-final /s/ to behave differently? One likely reason is the difference in its distribution across syllable-positions. While word-initial /s/

4 The pairwise interactions between the main effects are not listed here even though they were entered into this model.

MURIEL/OROZCO

31/5/12

14:24

Página 105

SYLLABLE-FINAL AND SYLLABLE-INITIAL /S/ REDUCTION IN CALI

105

is always in syllable-initial position, word-final /s/ occurs in syllable-initial position (because of resyllabification) only 24% of the time in these data (e.g. ‘vas a’ [bá-sa] vs. ‘vas con’ [bás-kon]). Thus, even when word-final /s/ is resyllabified into syllable-initial position when followed by a vowel, it remains lexically final and the cumulative experience in syllable-final position appears to influence its behavior. Overall, the results of our quantitative analyses summarized in sections 4.14.3 suggest that in this variety of Spanish, initial /s/ reduction is not an extension of final /s/ reduction. The rates of reduction in different word and phonological contexts are not suggestive of a cline of development proceeding from final > initial (Table 1). Furthermore, the separate analyses of initial vs. final /s/ reduction (Tables 2 & 3) show a different ranking of constraints, and consequently, magnitude of effect, for the linguistic factors. Lastly, when considering /s/ only in the intervocalic context shared across three word positions, we report no significant difference between word-initial and word-medial /s/ reduction, yet a significant difference between reduction in word-initial and word-final positions (Table 6). Together, these results suggest that the processes occurring in syllable-final position and those occurring in syllable-initial position are independent. These results are in line with the notion put forth by Bybee (2001: 86) that “it is probably fair to say that the ‘same’ consonant almost never appears in both syllableinitial and syllable-final position without some modification.” It seems, therefore, that when talking about aspiration and deletion in Cali, Colombia, we are talking about two variables, (s-) and (-s), as opposed to one.

5. Conclusions The previous discussion outlines the results of a large-scale quantitative analysis of /s/ reduction in all syllable and word positions (word-initial; syllable-initial, word-medial; syllable-final, word-medial; word-final) in Cali, Colombia. The results of the logistic regressions reveal the linguistic factor groups that significantly constrain both initial and final /s/, and the direction of effect for individual factors is determined (Tables 2 and 3). These results give us a better picture of the nature and extent of /s/ reduction in Cali, Colombia and address the paucity of empirical analyses of syllable-initial /s/ reduction. Future research is needed in other initial /s/ reducing varieties of Spanish (i.e. Honduras, Nicaragua, El Salvador) to address the dearth of empirical analyses on initial /s/ reduction. Individual analyses of initial /s/ tokens and final /s/ tokens revealed different orderings of factor groups (by magnitude of effect) and were suggestive of two

MURIEL/OROZCO

106

31/5/12

14:24

Página 106

EARL K. BROWN/ESTHER L. BROWN

separate variables (/s-/ and /-s/) as opposed to one (/s/). Also, quantitative analyses of reduction rates of /s/ tokens in different syllable positions and phonological contexts (Table 4) do not point to a unified process of /s/ reduction proceeding from preconsonantal contexts to ultimately include word-medial intervocalic contexts. Yet a comparison of rates of /s/ reduction in seven different dialects those with initial /s/ reduction and those without noted initial /s/ reduction - does reveal an interesting difference between the two types of /s/ reducing varieties (Table 5). Those varieties with no noted word- and syllable-initial /s/ reduction (Cuba, Argentina, Venezuela, Puerto Rico) all present significant differences in rates of reduction in prevocalic contexts and preconsonantal contexts in wordfinal position. Conversely, initial /s/ reducing varieties (New Mexico, Chihuahua, Cali) share the commonality of having no significant difference in reduction rates word-finally in prevocalic and preconsonantal contexts. To determine if this commonality implicated intervocalic /s/ as a crucial component acting as a portal through which final /s/ could proceed to initial, we conducted an analysis of all intervocalic tokens of /s/ (Table 6). Results of this analysis clearly suggest that even in the shared intervocalic context, /s-/ and /-s/ are significantly different. The final > initial continuum reviewed in this work, therefore, does not seem to hold for the dialect under discussion here. Our data are taken from the Spanish of Cali, which, although classified as a highland variety (e.g. Montes Giraldo 1982), may, in fact, be better considered a transitional variety with influences from the nearby coastal regions. The ‘transitional’ nature of this variety may have implications for the analysis that were not identified in this work, and the applicability of the final > initial cline should be tested in future research on different varieties. Additionally, social variables were not included in this analysis. Phonological variation of /s/ has been shown to correlate significantly with social variables (e.g. Lafford 1986) that help elucidate mechanisms of change, and as such, should be considered in subsequent works on Cali /s/ reduction.

MURIEL/OROZCO

31/5/12

14:24

Página 107

O N T H E S Y N TA X O F T H E F O C A L I Z I N G S E R ( ‘ T O B E ’ ) S T R U C T U R E I N T H E S PA N I S H OF BUCARAMANGA CATALINA MÉNDEZ VALLEJO

1. Introduction In Spanish, a declarative sentence (1) can be transformed into a pseudo-cleft construction, such as (2): (1) Adelaida tiene una casa en Real de Minas (2) Lo que Adelaida tiene es una casa en Real de Minas

Pseudo-cleft constructions are often used in Spanish to focalize certain constituents within a sentence. In (2), for example, the direct object (una casa en Real de Minas) is the focused constituent and is c-commanded by the copula verb es (> ser). In certain Spanish varieties (i.e. Venezuelan, Ecuadorian, Panamanian, Colombian, and Dominican), a sentence involving the same focus interpretation can also be created without the relative clause (lo que), as shown in (3): (3) Adelaida tiene es una casa en Real de Minas

The structure in (3), here referred to as the Focalizing Ser structure (henceforth FS), has not been widely studied. It has been reported to occur in Brazilian Portuguese (Oliveira/Braga 1996, Mikolajczak 2003, Kato 2009), Venezuelan Spanish (Sedano 1988, 1990, 1994a, 1995b, 2003a, 2003b), Colombian Spanish (Albor 1986, Curnow/Travis 2004), Dominican Spanish (Toribio 1992, 2002), and Caribbean Spanish (Bosque 1999, Camacho 2006). In terms of the syntactic study of the FS, Bosque, Toribio, Camacho, and Kato have each proposed their own theoretical analysis of this dialectally-marked form. Despite the progress made in these proposals, there are still certain aspects that have not been explained or that are not entirely supported by empirical data. This chapter aims to provide an innovative account of the syntactic properties of FS, specifically in terms of the kind of constituents that can (or cannot) be FS-

MURIEL/OROZCO

108

31/5/12

14:24

Página 108

CATALINA MÉNDEZ VALLEJO

focused and the morphology of FS. Based on the data presented here1, I propose that the FS is a sentence-internal focus structure, and I discuss some of the formal features that I have extensively described in previous work (Méndez Vallejo 2009). This chapter is organized as follows: section 2 briefly reviews previous studies of the FS; section 3 provides an alternative analysis of FS as an independent syntactic structure; section 4 describes the type of constituents that can (and cannot) be FS-focused; section 5 analyzes the morphology of FS and shows that the agreement patterns found between FS and certain sentential constituents are not random; section 6 discusses the FS as a dialectal manifestation of sentence-internal focus; and, section 7 states some conclusions.

2. Previous approaches to the FS puzzle 2.1. THE FS AND THE PSEUDO-CLEFT

The similarity found between examples (2) and (3) shown above has prompted some scholars to claim that FS is a reduced version of the pseudo-cleft construction. In fact, according to Albor (1986), Sedano (1988, 1990, 1994, 1995, 2003a, 2003b), and Toribio (1992, 2002), the only structural difference between the pseudo-cleft and the FS is that the relative clause present in the former (lo que ‘what’) is not pronounced in the latter. Albor (1986: 184), for example, describes FS as a syntactic derivation of the pseudo-cleft. He attributes the FS to Colombian and Ecuadorian dialects, and claims that its use is geographically, socially, and stylistically more extended than usually thought: it is found in various dialectal regions, in educated and non-educated populations, in colloquial speech, and in literary and news texts2. Based on her statistical analysis of FS in Caracas Spanish, Sedano (2003a, 2003b) suggests that although the FS and the pseudo-cleft are practically equivalent, FS is structurally simpler (it lacks a relative clause), which facilitates the production of focus (it allows last-minute focalization).

1

Most of the data used here come from utterances evaluated by native Colombian speakers. 45 university students from Bucaramanga conducted acceptability judgments on 125 sentences (84 containing the FS). Each sentence was part of a mini-dialogue and participants were asked to rate them using a scale from 1 (least acceptable) to 5 (most acceptable). 2 I personally have found the FS in written comments on the internet (e.g. personal blogs, newspapers, etc.), and I have heard it being used numerous times on Colombian TV (e.g. news broadcasts, soap operas, talk shows, etc.).

MURIEL/OROZCO

31/5/12

14:24

Página 109

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

109

Finally, Toribio (1992, 2002) proposes a unified syntactic analysis of the FS and the pseudo-cleft, and claims that FS “is the null operator counterpart of the traditional pseudo-cleft” (Toribio 2002: 134). Thus, according to her analysis, examples (2) and (3) above would be structured as shown in (4) and (5), respectively: (4) Lo [lo (pro)

que Adelaida tiene [CP OP que [IP Adelaida tiene t…

es una casa en Real de Minas

(5) Lo [lo (pro)

que Adelaida tiene es una casa en Real de Minas [CP OP que [IP Adelaida tiene t…

Although analyzing the FS as an incomplete form of the pseudo-cleft seems to be a simple and appealing way to explain the syntactic configuration of the FS, later research shows that FS is indeed an independent structure. According to Curnow/Travis (2004: 3), for example, it is clear that the pseudo-cleft is sensitive to clitic climbing (6)-(7), whereas FS is not (8)-(9): (6)

ok

Lo que quiero es irme

(7) * Lo que mei quiero es iri (8)

ok

Quiero es irme

(9)

ok

Mei quiero es iri

In addition, both Curnow/Travis (2004) and Bosque (1999) claim that the pseudo-cleft cannot focus negative polarity items (10), whereas the FS can (11): (10) * El que no vino fue nadie (11)

ok

No vino fue nadie

Bosque also shows that focused constituents can be extracted out of a pseudo-cleft structure (12), but not out of an FS structure (13): (12)

ok

¿Quiéni fue el que salió ________?

(13) * ¿Quiéni salió fue ________?

Finally, I have previously pointed out (Méndez Vallejo 2009: 103) that progressives cannot be focalized by the pseudo-cleft (14), but they can by FS (15):

MURIEL/OROZCO

31/5/12

110

14:24

Página 110

CATALINA MÉNDEZ VALLEJO

(14) * Lo que ha estado es estudiando lingüística (15)

ok

Ha estado es estudiando lingüística

Looking at the data in (6)-(15), it becomes evident that the FS and the pseudo-cleft do not react similarly to certain syntactic phenomena (e.g. occurring in cases of clitic climbing and focusing negative polarity items, extracted constituents, and progressives). Thus, the fact that the FS and the pseudo-cleft are not interchangeable in all syntactic contexts strongly suggests that they do not share the same structural configuration. In fact, examples (6)-(9) and (12)-(13) indicate that the verb ser (‘to be’) does not occupy the same syntactic position in pseudo-clefts and in the FS structure: in pseudo-clefts ser is placed within CP, whereas in the FS construction it seems to be placed within IP. More specifically, cases of wh-movement (affecting two clauses) are only felicitous with the pseudo-cleft, and cases of clitic climbing (affecting one clause) are only felicitous with the FS. Hence, while FS involves only one clause, the pseudo-cleft involves two distinct clauses. Given this, the analysis that I present here takes as a starting point the assumption that the FS is not a pseudo-cleft but a completely independent syntactic structure. Particularly, as I will propose in section 3, FS should be viewed as a TP-internal functional projection, always preceding focused (discourse-new) constituents.

2.2. SYNTACTIC ANALYSES OF THE FS Starting from a view of the FS as an independent structure (not as an incomplete form of the pseudo-cleft), Bosque (1999), Camacho (2006), and Kato (2009) have proposed their own syntactic accounts of this form. Bosque, for example, argues that FS is a Focus Phrase generated inside VP, where the non-copulative verb ser functions as the head of the focus projection. Hence, a sentence such as (16) is analyzed by Bosque as shown in (17): (16) Juan comía era papas (17) [IP Juani [VP ti [V comía [FP [[F° era ] papas ]]]]] (Bosque 1999: 4)

Camacho also suggests that FS is a VP-internal structure. For him, it is an equative structure (i.e. a Copula Phrase) that originates as an adjunct of VP. In this kind of construction, the subject is null and the predicate is the focused VP. Thus, following this proposal, a sentence such as (18) is derived as illustrated in (19):

MURIEL/OROZCO

31/5/12

14:24

Página 111

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

111

(18) Los pájaros se comieron fue las migas (19) [IP Los pájaros [I’ se comieron [VP [VP [V] ei] [IP COP x [I’COP fue [las migasi ]]]]]] (Camacho 2006: 19)

However, although analyzing the FS as a VP-internal structure correctly predicts many cases of FS-focalization, it fails to account for FS-focused constituents generated above vP. For example, as shown in (15) above, FS may focus progressives, which are thought to originate in Auxiliary Phrases (Progressive Phrases) above vP (Klein 1968, Zagona 2002). Furthermore, under Camacho’s view the FS must only focus single, domainfinal constituents. Given this, sentences such as (20)-(21) are ruled out by Camacho because in the former case there are more than two constituents being focused ([pan][a su abuela]), and in the latter the focused constituent ([pan]) is not placed at the end of the relevant domain (the TP clause): (20) * Marta le compró fue pan a su abuela (21) * Marta le compró fue pan a su abuela (Camacho 2006: 20)

However, I have found that the FS can focus more than one constituent, or a constituent that is not domain-final (Méndez Vallejo 2009: 185). Thus, the data collected indicate that Colombian speakers (e.g. students from Bucaramanga) accept sentences such as (20)-(21)3. Finally, based on her previous work in Brazilian Portuguese (Kato 2008), Kato (2009) proposes an innovative analysis of FS applicable to Caribbean Spanish. According to her, a series of remnant movements leads to the occurrence of a Copula Phrase (placed in CP), where the FS is ultimately generated. Thus, if we applied Kato’s analysis of Portuguese to the Spanish data, a sentence such as (22) would be derived through a series of steps, as in (23):

3 The acceptability judgment tests that I administered to university students from Bucaramanga included sentences such as (20)-(21) above. The results obtained show that these speakers find these kinds of sentences acceptable. It is also worth mentioning that I have collected sentences similar to (20) and (21) from TV shows (e.g. soap operas, news programs, etc.) and from naturally-occurring conversations.

MURIEL/OROZCO

31/5/12

14:24

Página 112

112

CATALINA MÉNDEZ VALLEJO

(22) Eva traía era uvas (23) a. [FP F [IP Eva [I’ traía [vP tV uvas ]]]] b. [FP [vP tV uvas] [F F [IP Eva [I’ traía [vP tVP ]]]]] ↑

c. [IP era [VP tcop [FP [vP tV uvas ] [F F [IP Eva [I’ traía [vP tVP ]]]]]]] d. [TopP [IP Eva [I’ traía [vP tVP ]]] [TopP’ Top [IP era [VP tcop [FP [vP tV uvas] [F F [IP tIP]]]]]] ↑

Although Kato’s analysis correctly predicts most cases of FS-focalization, it fails to account for cases in which the FS is followed by already-mentioned elements. As I have pointed out above, Colombian speakers accept sentences such as (21) above, where an old-information segment follows the FS-focused constituent. Hence, following Kato’s remnant analysis, it is not possible to obtain the FS sentence in (24) after applying the derivation in (25): (24) A: Escuché que Verónica le dio un carro a Eric B: No, le dio fue una moto A ERIC (25) a. [FP F [IP [ I’ le dio [TopP A ERIC [vP tV una moto tDP]]]]] b. [FP [vP tV una moto tDP]] [IP [ I’ le dio [TopP A ERIC [vP tVP ]]]]] ↑

c. [IP fue [FP [vP tV una moto tDP]] [IP [I’ le dio [TopP A ERIC [vP tVP]]]]] d. [TopP [IP[I’ le dio [TopP A ERIC [vP tVP ]]]] [IP fue [FP [vP tV una moto tDP] [F F [IP tIP ]]]]]] ↑

So far, I have discussed three syntactic analyses of FS. As I have shown, they constitute interesting approaches to solving the FS puzzle and they have greatly contributed to our understanding of this structure. However, all of these analyses run into some serious theoretical and empirical difficulties. In the next section, I will describe my syntactic proposal for the FS, and I will show that this form should be viewed as a functional projection located inside TP.

3. An alternative view of the FS Before I describe my proposal for the syntactic configuration of FS, it is important to mention some of its informational characteristics. First, as shown in (26)(27) below, the FS only renders felicitous outcomes when it precedes focalized (discourse-new) constituents:

MURIEL/OROZCO

31/5/12

14:24

Página 113

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

113

(26) A: ¿Dónde llovió esta mañana? B: ok Llovió fue en Piedecuesta (27) A: ¿Dónde llovió esta mañana? B: # Llovió fue esta mañana

Second, similarly to Curnow/Travis (2004), I have observed that FS may introduce both contrastive and non-contrastive focus in Colombian Spanish. In (28), for example, the question uttered by A does not provide a closed set of alternatives from which B needs to choose (non-contrastive focus), whereas in (29) B is prompted to choose one element from the closed set of alternatives in A’s question (contrastive focus): (28) A: ¿Qué compraste? B: ok Compré fue vino (29) A: ¿Compraste vino o cerveza? B: ok Compré fue vino

Third, the verb ser in the FS construction functions as a link between new information introduced in discourse and what has already been mentioned or is presupposed. As I explained in previous work (Méndez Vallejo 2009), FS does not act as a copulative or an auxiliary verb, and it seems to have acquired a new pragmatic function as a focus link. In fact, examples (30)-(31) below show that FS does not seem to hold any specific verbal functions (although it still maintains verb morphology), and it only appears to emphasize or intensify the focused element: (30) Estoy es triste (31) Todo ha sido es un problema

Based on this observation, I suggest that FS is generated inside a Focus Phrase (FocP), in which the head (Foc) is null and the specifier is eventually occupied by FS: (1)

FocP SER

FocP Foc

XP

MURIEL/OROZCO

31/5/12

114

14:24

Página 114

CATALINA MÉNDEZ VALLEJO

In terms of the syntactic position of this FocP, I claim that it must occur within the internal periphery of TP. Given that the FS can only focus post-verbal constituents (see (33)-(34) below), I argue that FocP must be placed below T4: (33)

ok

[TP Juan Carlos [T nos prestói [FOCP fue [vP ___ una mecedora ]]]]

(34) * [TP Juan Carlos [FOCP fue [T nos prestói [vP ___ una mecedora ]]]] (35) * [FOCP Fue [TP Juan Carlos [T nos prestói [vP ___ una mecedora ]]]]

Furthermore, since FS may focus constituents that are generated above vP (such as perfectives (36) and progressives (37)), I argue that this FocP must also be placed above vP: (36)

ok

[TP Lucy [T había [FOCP era [PERFP estado [PROGRP saliendoi [vP __ [PP con Pepe]]]]]]]

(37)

ok

[TP Lucy [T había [PERFP estado [FOCP era [PROGRP saliendoi [vP __ [PP con Pepe]]]]]]]

Finally, the data that I have collected show that the FS may be preceded by constituents that have already been mentioned in previous discourse (topicalized elements): (38) A: Fernando dizque está podrido en plata B: Pues tendrá PLATA será la familia

In brief, the syntactic structure that I propose for FS can be illustrated as in (39) below5:

4 I adopt here the by-now standard analysis according to which verbs in Spanish overtly move from v to T (Edmonds 1978; Contreras 1991; Zagona 2002). 5 It is important to mention at this point that I have employed the mechanism of Agree (Chomsky 2000) to argue that focus features are matched and valued between FocP and the focused element. In Méndez Vallejo (2009), I show that only constituents that have [+focus] features and fulfill Chomsky’s Conditions on Matching can be FS-focused. This allows me to explain why in sentences such as (20)-(21) FS can focus both constituents ([pan] [a su abuela]), or only one of them ([pan] [a su abuela]). However, such discussion is beyond the scope of this study.

MURIEL/OROZCO

31/5/12

14:24

Página 115

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

(39)

115

TP T’ T

TopP FocP SER

PerfP ProgrP vP v’ v

V V

XP

Hence, analyzing FS as a TP-internal structure makes the correct predictions regarding FS focalization (only post-verbal elements can be FS-focused, and elements above vP may be FS-focused). It also correlates with previous claims according to which the low-IP area (i.e. the left periphery of vP/VP) reflects the low-CP area (i.e. the left periphery of IP) in that it is rich with topic and focus projections (Horvath 1986, Yanagida 1995, Belletti 2004, Hsu 2008). In the next section, I will concentrate on the kind of constituents that may be FS-focused, and in section 5, I will explore the morphological properties of FS.

4. FS-focused constituents In Bucaramanga Spanish, FS is used to focalize a wide variety of phrases. For example, it can be found preceding prepositional phrases (PPs), as shown in (40)-(41): (40) [TP [T Vamos [FOCP es [vP [PP para el concierto ]]] (41) [TP [T Estaba nevando [FOCP era [vP [PP en Moscú ]]]]]

The FS can also be used to focus single adjectives (42), or entire adjectival phrases (43):

MURIEL/OROZCO

31/5/12

14:24

116

Página 116

CATALINA MÉNDEZ VALLEJO

(42) [TP La falda

[T me quedó [FOCP fue [vP grande ]]]]

(43) [TP Los visitantes [T están

[FOCP es [vP [ADJP súper interesados ]]]]]

Furthermore, FS may precede single adverbs (44), or entire adverbial phrases (45): (44) [TP [T Llegaron [FOCP fue [vP (45) [TP [T Llegó

ayer ]]]]

[FOCP fue [vP [ADVP como ayer ]]]]]

The adverbs in (44)-(45) have been referred to in the literature as ‘VPadverbs’ (cf. Zagona 2002) or ‘lower adverbs’ (cf. Cinque 1999). As shown, this kind of adverb is felicitously FS-focused. However, as illustrated in (46)-(47), ‘IP-adverbs’ (cf. Zagona 2002) or ‘higher adverbs’ (cf. Cinque 1999) must not be FS-focused: (46) * [TP No [T veo (47) * [TP

[FOCP es francamente [vP nada ]]]]

[T Estás [FOCP es naturalmente [vP loca ]]]]

In addition to VP-adverbs, the FS may precede determiner phrases (DPs) serving as direct objects (48), indirect objects (49), and even subjects (50)6: (48) [TP La tía Carmen

[T

nos mandó [FOCP fue [vP [DP panuchas ]]]]]

(49) [TP [T Les mostramos [TOPP las fotos

[FOCP fue [vP [DP a ellos ]]]]]]

(50) [TP En clase

[FOCP fue [vP [DP Álvaro ]]]]]

[T

habló

Importantly, FS may not precede DPs that have moved to positions higher than T’, such as CP or Spec, TP. For example, in sentences (51)-(53), the object DPs have moved from vP to a Topic Phrase above TP, and the subject DP has moved from vP to Spec, TP7:

6 Toribio (1992, 2002) claims that post-verbal subjects must not be focused by FS in Dominican Spanish. However, based on the acceptability judgment tests that I conducted and data that I have gathered from Colombian TV shows, I have found that FS-focused post-verbal subjects are completely acceptable in Colombian Spanish. 7 Much debate still remains regarding the final placement of pre-verbal subjects in Spanish. It has been proposed, for example, that pre-verbal subjects check nominative case within

MURIEL/OROZCO

31/5/12

14:24

Página 117

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

(51) * [TOPP Panuchasi [TP

la tía Carmen [T

117

nos mandó [FOCP fue [vP ___ ]]]]]

(52) * [TOPP A ellosi

[TP [T les mostramos [TOPP las fotos

[FOCP fue [vP ___ ]]]]]]

(53) * [TOPP Álvaroi

[TP

[FOCP fue [vP ___ ]]]]]

en clase

[T

habló

Besides DPs, the FS may precede various types of embedded CPs (54)-(55) and IPs (56): (54) [TP [T Daba tristeza (55) [TP

Ella

[FOCP era [CP porque se fuera solo ]]]] [T

(56) [TP [T Vi

trataba

[FOCP era [CP de imponerle todo a Jairo ]]]]

[TOPP a Tatiana [FOCP fue [IP llegar ]]]]]

As I mentioned in section 3, I have also observed that the FS can focalize perfectives (57), progressives (58), or a combination of a perfective and a progressive (59): (57) [TP

La porción [T

la han [FOCP

es

[PERFP disminuidoi [vP ___ ]]]]]

(58) [TP [T Voy

[FOCP es

[PROGRP saliendoi [vP ___ ]]]]]

(59) [TP [T Había

[FOCP era

[PERFP

estado

[PROGRP estudiando [vP ___ mucho ]]]]]]

Interestingly, the FS cannot focus auxiliary verbs that have moved to T: (60) * [FOCP Era [TP [T había [PERFP estado [PROGR estudiando [vP ___ mucho ]]]]]]

Furthermore, I find that FS may precede in-situ interrogative words uttered in echo questions (61), but, as initially pointed out by Bosque (1999: 18), the FS must not precede interrogative words that have been extracted to CP: (61) A: El profesor dijo que había examen y trabajo final B: ok¿ [TP [T Dijo [FOCP fue [vP qué ]]]]? (62) B: *¿ [CP Quéi [TP [T dijo [FOCP fue [vP ____ el profesor ]]]]]?

VP and later move to Spec, IP (TP) to check EPP features (e.g. Contreras 1991). Other authors argue that pre-verbal subjects are placed in a Topic Phrase and move to Spec, IP (e.g. MejíasBikandi 1992) or above IP (e.g. Ordóñez 1997). Given that the placement of pre-verbal subjects does not affect my theoretical proposal for the FS, I will simply assume here that pre-verbal subjects move to Spec, TP.

MURIEL/OROZCO

31/5/12

118

14:24

Página 118

CATALINA MÉNDEZ VALLEJO

Finally, as shown in previous work (Méndez Vallejo 2009: 205), I have found that FS may focus certain types of small clauses. Starting from observations made by Sedano (1990: 93)8, I have discovered that the FS may focus small clauses containing stage-level predicates (63), but not containing individuallevel predicates (64): (63)

ok

[TP [T Trajo [TOPP el perroi [FOCP fue [vP [SC ____ encadenado ]]]]]]

(64) * [TP [T Trajo [TOPP el perroi [FOCP fue [vP [SC ____ dálmata ]]]]]]

So far, I have shown that FS may focus a wide variety of phrases (e.g. PPs, AdjPs, AdvPs, DPs, PerfPs, ProgrPs, CPs, IPs, and Small Clauses). Looking in particular at FS-focused DPs, it is clear that a phrase may be FS-focused, irrespective of its syntactic or semantic role within the sentence (e.g. subject, direct object, indirect object). However, as illustrated in examples (46)-(47), (51)-(53), (60), and (62), FS never precedes a constituent that has moved to a pre-verbal position, or a verb placed in T (i.e. a matrix or an auxiliary verb). Based on these empirical observations, the generalization that I establish here is that the FS may focus any type of constituent as long as it is post-verbal. This finding becomes particularly crucial for my syntactic analysis of the FS. As I explained in section 3, having the FS preceding only post-verbal elements clearly suggests that it constitutes a TP-internal functional projection.

5. FS The morphological characteristics of FS have remained largely under investigated. Although Bosque (1999) and Curnow/Travis (2004) have mentioned a few

8

For Sedano (1990: 93), the sentence in (i) suggests that the FS may occur inside DPs (NPs in her terminology): (i) Me gusta la música es moderna However, looking at examples (63)-(64), I would argue that Sedano’s example (i) does not constitute a case of DP disruption, but of a stage-level predicate being FS-focused. That is, when (i) entails an individual-level reading, the sentence becomes ungrammatical (ii), but when it entails a stage-level reading, it remains grammatical (iii): (ii)i * Me gusta la música es moderna (no rock) (iii) ok Me gusta la música es moderna (no bailable)

MURIEL/OROZCO

31/5/12

14:24

Página 119

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

119

peculiarities regarding FS morphology, this topic has received little attention. In this section, I would like to emphasize the complexity of FS morphology and to describe two agreement processes that I have previously unveiled (Méndez Vallejo 2009): one, observed between FS and the verb in T, and another, between FS and the focused constituent9. In terms of the first agreement process, FS must agree in tense with the verb in T. As shown in (65)-(66), for example, the sentence becomes felicitous only when both the matrix verb and FS agree in tense: (65)

ok

Esa mujer está es loca

(66) * Esa mujer estaba es loca

Furthermore, FS and the verb in T must also agree in aspect: (67)

ok

Esa mujer estaba era haciéndose la loca

(68) * Esa mujer estaba fue haciéndose la loca

However, FS does not agree with the verb in T in either person or number: (69)

ok

Les traje fue unas empanadas de carne

(70) * Les traje fui unas empanadas de carne (71)

ok

Les trajimos fue unas empanadas de carne

(72) * Les trajimos fuimos unas empanadas de carne

As for the second agreement process, FS shows a more complicated pattern as it appears to hold different agreement relations with post-verbal subjects and objects (DOs and IOs). First, when the FS focuses post-verbal subjects, FS agrees with them in both person and number:

9 Given the scope of this chapter, I will only focus on describing the agreement patterns between FS and the verb in T, and between FS and the focused constituent. However, I should mention that I use the mechanism of Agree (Chomsky 2000) to explain how: a) tense/aspect features are matched and valued between FocP and the verb in T, and b) person/number features are matched and valued between FocP and the focused element. For a more thorough account of FS morphology, please see Méndez Vallejo (2009).

MURIEL/OROZCO

31/5/12

14:24

120 (73)

Página 120

CATALINA MÉNDEZ VALLEJO

ok

Saqué la basura fui yo

(74) * Saqué la basura fue yo (75)

ok

Sacamos la basura fuimos nosotros

(76) * Sacamos la basura fue nosotros10

Second, when FS precedes direct objects, it optionally agrees with them in number. For example, when the DO is plural FS may remain in singular (77), or become plural too (78): (77)

ok

Tengo es dos gatos

(78)

ok

Tengo son dos gatos

Importantly, FS does not establish any person agreement with the DO. As shown in (79)-(80), the sentence is felicitous only when FS is conjugated in the third person singular: (79)

ok

Te viste fue a ti misma

(80) * Te viste fuiste a ti misma

The agreement variation shown between FS and DOs (in (77)-(78)) resides entirely on contrast. That is, if FS is used in a non-contrastive context (81) number agreement is not allowed. However, if the FS is used in a contrastive context (82), number agreement is required11: (81) A: ¿Tienes mascotas? B: Sí, tengo (ok es / *son) dos perros y un gato (82) A: ¿No tenías dos gatos y un perro? B: No, tengo (* es / ok son) dos perros y un gato

10

In Méndez Vallejo (2009), I show that cases such as (76) may be acceptable only with a non-contrastive reading. 11 A more detailed discussion concerning the relation between contrastive focus and FS morphology can be found in Méndez Vallejo (2009).

MURIEL/OROZCO

31/5/12

14:24

Página 121

121

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

Interestingly, FS never establishes person or number agreement with IOs (83), or with DOs showing Differential Object Marking (84)12: (83) Les habíamos comprado un videojuego (ok era / *eran) a mis primos (84) Besó (ok fue / *fueron) a mis primos

In fact, this lack of agreement between FS and IOs (and DOs with Differential Object Marking) holds in both non-contrastive (85) and contrastive (86) contexts: (85) A: ¿A quién le diste tus joyas? B: Se las di (ok fue / * fueron) a los revendedores (86) A: ¿No les diste tus joyas a los fabricantes? B: No, se las di (ok fue / * fueron) a los revendedores

Up to this point, the data presented in this section show that FS establishes two separate agreement relations: one the one hand, it agrees with the verb in T in both tense and aspect (87a); on the other, it agrees in person and number with post-verbal subjects, and in number with DOs lacking Differential Object Marking (87b): (87) a. T [Tense: α] [Aspect: β]

b. FS [Tense: α] [Aspect: β]

FS [Number: α] [Person: β]

Focused subject [Number: α] [Person: β]

FS [Number: α] [Person: β]

Focused DO (no DOM) [Number: α] [Person: β]

It is important to clarify that FS does not establish any kind of agreement with IOs or with DOs showing Differential Object Marking. Also, the number agreement variability found between FS and DOs can be explained in terms of 12

Differential Object Marking consists of using the preposition a before animate objects. This phenomenon is also referred to in the literature as personal a (Zagona 2002, Ordóñez 1997).

MURIEL/OROZCO

122

31/5/12

14:24

Página 122

CATALINA MÉNDEZ VALLEJO

contrastive focus. That is, if the FS is produced in a non-contrastive context, it does not agree with the DO in number; but if the FS is uttered in a contrastive context, it must agree with the DO in number. Also, it is important to note that the examples provided up to this point can account for the arguments that I develop in this chapter. I have presented some highlights of the structure of the FS, but I have not elaborated on certain theoretical details. Please refer to my earlier work (Méndez Vallejo 2009) to find a more complete account of FS morphology.

6. FS as a dialectal manifestation of TP-internal focus In section 3, I provided empirical and theoretical evidence to support my claim that the FS is a TP-internal focus structure. In this section, I extend this idea to suggest that the syntactic structure illustrated in (39) above may be applicable to all Spanish dialects, not just to FS-dialects. More specifically, I hypothesize that all dialects of Spanish have TP-internal projections destined to host elements which are informationally specified (i.e. elements that convey informational features such as focus, topic, (non)contrastiveness, evidentiality, etc.). Thus, depending on the dialect in question, this TP-internal area may be occupied by one (or more) lexical item(s) to convey one (or more) informational feature(s). I will clarify here that the purpose of this section is to propose a new avenue of investigation, rather than to report findings that are empirically-well established (contrary to what I have done in sections 2-4). In previous work (Méndez Vallejo 2009: 314), I have explored the idea that speakers from non-FS-dialects may choose certain dialectally-marked expressions in order to express particular semantic and/or pragmatic meanings (e.g. topic, focus, evidentiality, contrast, etc.). These expressions may often be placed within TP, preceding segments of new information. For example, in some Spanish dialects (e.g. Peruvian Spanish) it seems that the conjunction pero may precede the focused constituent in contrastive contexts13: (88) A: Pensé que tenías dos perros B: ok No, tengo pero dos gatos

13 Sentences such as (88) are acceptable among some Peruvian and Colombian speakers that I have consulted. However, a few Mexican speakers have reported to me that they do not find this sentence acceptable. A more comprehensive test should be designed to pinpoint exactly which speakers (i.e. speakers of which particular dialects) accept these sentences, and in which particular discourse contexts.

MURIEL/OROZCO

31/5/12

14:24

Página 123

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

123

However, pero does not seem to render grammatical outcomes when used in non-contrastive contexts (89): (89) A: ¿Tienes mascotas? B: *Sí, tengo pero dos gatos

Contrary to pero, the expression dizque in Colombian Spanish may be used TP-internally to express evidentiality in non-contrastive contexts (90). However, when dizque is used in contrastive contexts (91), the sentence is infelicitous14: (90) A: ¿Qué se compró Julián? B: ok Se compró dizque un BMW (91) A: ¿Julián no compró un Mercedes? B: *No, se compró dizque un BMW

As shown in (88)-(91), other expressions (aside from the FS) may be placed TP-internally before discourse-new elements, in order to convey focus, contrast, or evidentiality. Interestingly, both pero and dizque may occur alongside the FS: (92) A: Pensé que tenías dos perros B: ok No, tengo es pero dos gatos (93) A: ¿Julián no compró un Mercedes? B: ok No, se compró fue dizque un BMW

Based on my own intuitions as a Colombian speaker, sentences (92)-(93) do not convey the same meanings as sentences (88) and (90) above: in (92), the contrast becomes more emphatic than in (88), and in (93) both evidentiality and contrast are expressed. Thus, pero and dizque seem to have different discourse functions and are not semantically (or syntactically) equivalent to the FS15. At this point, I would like to speculate that in non-FS- and FS-dialects informationally-specified elements (e.g. pero and dizque) might occur TP-internally (like the FS), but would occupy their own discourse-specific phrase. So, the syntactic

14 Sentences (90)-(91) are extrapolations from sentences that I have personally uttered or that I have heard in naturally-occurring conversations. 15 Notice that in the FS establishes a systematic morphological agreement with certain sentential arguments, whereas pero and dizque are fixed expressions. Hence, the FS is syntactically not equivalent to pero and dizque.

MURIEL/OROZCO

31/5/12

14:24

Página 124

124

CATALINA MÉNDEZ VALLEJO

structure that I proposed in (39) is always present in all Spanish dialects, and particular discourse-related structures can be added within the TP mid-area. Hence, in dialects in which pero is used contrastively (88), a Contrast Phrase would occur (94); in dialects in which dizque is used for evidentiality (90), an Evidentiality Phrase would occur (95); and in dialects in which both FS and pero (92), or FS and dizque (93) occur, a Contrast Phrase and a Focus Phrase (96), or an Evidentiality Phrase and a FocP would occur (97): (94)

TP

(95)

TP

T’ T tengo

T’ ContrastP

pero

T EvidentialityP compró dizque EvidentialityP’

ContrastP’

Contrast

vP

Evidentiality

dos gatos

(96)

un BMW

(97)

TP

TP

T’ tengo

T’ FocP

es

vP

compró

ContrastP’ pero

vP dos gatos

FocP fue

EvidentialityP’

dizque

vP un BMW

Once all the morpho-syntactic processes have taken place (e.g. agreement, movement, etc.), the TP-internal functional projections (between TP and vP) may be filled (or not) at Spell-Out. In FS-dialects, FS is inserted within FocP (98), and in non-FS-dialects other expressions (e.g. pero, dizque) are inserted within other functional projections, above or below FocP (99):

MURIEL/OROZCO

31/5/12

14:24

Página 125

125

ON THE SYNTAX OF THE FOCALIZING SER (‘TO BE’)

(98) Morpho-syntax:

Spell-out/PF: TP

TP

T’

T’ T

T salí

FocP

FocP

vP

Foc

vP

Foc

FocP’

fui

FocP’

yo

XP

(99) Morpho-syntax:

Spell-out/PF: TP

TP

T’

T’ T

T

ContrastP/EvidentialityP FocP

Contr/Evid Foc

ContrastP/EvidentialityP FocP’

(pero) ContrastP/ EvidentialityP

Contr/Evid

vP

ContrastP/ EvidentialityP (dizque)

XP

vP XP

Although this is an interesting proposal, it is imperative to conduct extensive research in order to explore the relation between FS and other functionallymarked expressions in FS-dialects. I suspect that, at least in Bucaramanga Spanish, the FS is forced to occupy a higher syntactic position over other functionally-marked expressions (e.g. dizque): (100) Fernanda llegó (*dizque ) fue (ok dizque ) acompañada

Furthermore, it is necessary to investigate the semantic and pragmatic functions of expressions such as pero and dizque. This will certainly broaden our

MURIEL/OROZCO

31/5/12

126

14:24

Página 126

CATALINA MÉNDEZ VALLEJO

understanding of TP-internal functional projections in Spanish and would allow us to establish a more coherent and accurate comparison with FS.

7. Conclusions The FS has been reported and described in the literature (Albor 1986, Sedano 1990), but its formal properties have not been completely clarified despite several efforts (Bosque 1999, Toribio 1992, 2002, Curnow /Travis 2004, Camacho 2006, Kato 2009). This dialectally-marked form had been previously examined as an incomplete form of a cleft structure (Albor 1986, Sedano 1990, Toribio 1992, 2002), but later research shows that it is syntactically unrelated to clefts (Bosque 1999, Curnow/Travis 2004, Camacho 2006, Kato 2009). The study presented in this chapter maintains this latter claim, according to which the FS and the pseudo-cleft are different syntactic structures. Based on data presented in sections 3, 4, and 5 I draw some important generalizations regarding the syntactic configuration of FS: 1) the FS may occur in both contrastive and non-contrastive contexts; 2) FS serves as a link between the focused constituent and previous context; 3) FS may focus any type of constituent as long as it is post-verbal (it originates below TP and remains below TP); 4) FS agrees with the verb in T (a matrix or an auxiliary verb) in both tense and aspect; 5) FS agrees with post-verbal subjects in both person and number, and with DOs (lacking Differential Object Marking) only in number; 6) the agreement variability found between FS and DOs directly correlates with contrastive focus; and 7) FS does not establish any kind of agreement with IOs and with DOs showing Differential Object Marking. Given these empirical observations, and taking into account certain problems observed in other syntactic accounts (Bosque 1999, Camacho 2006, Kato 2009), I have proposed that the FS should be examined as a functional projection placed outside vP. Specifically, following Belletti’s research (2004) according to which both topic and focus projections may be found within the internal periphery of IP, I claim that FS is generated inside a Focus Phrase (FocP), placed below T and above vP. Finally, I have discussed the role of the FS as a dialectal manifestation of sentence-internal focus. Looking at other functionally-related expressions such as pero and dizque, I argue that the proposed syntactic structure for FS is present in all Spanish dialects. Thus, discourse-related elements (i.e. elements bearing informationally-related features such as topic, focus, contrast, etc.) may be placed within various kinds of Phrases above or below FocP. Needless to say, it is imperative to conduct more research in order to reach more definite conclusions regarding the relation between the FS and other discourse-related elements.

MURIEL/OROZCO

31/5/12

14:24

Página 127

A L A B O R AT O R Y A P P R O A C H T O S - L E N I T I O N I N T H E S PA N I S H O F B A R R A N Q U I L L A , C O L O M B I A RICHARD J. FILE-MURIEL

1. Introduction: Previous studies and their methodologies Most sociophonetic studies in Spanish deal with segmental lenition or weakening, which Lewis (2001) defines as a process by which a consonant becomes reduced (articulatorily) and/or more sonorous (i.e. vocalic) in some sense. The variation and change of a particular sound is important because it allows one to assess the influence of the different factors (linguistic and extra linguistic) that largely determine speech patterns. Furthermore, it enables us to understand how human beings represent the sound system in their brains and how this shared knowledge changes over time. The majority of the existing quantitative studies that examine s-lenition have relied exclusively on transcription via auditory analysis (see Erker 2010; FileMuriel/Brown 2010; Minnick Fox 2006 for exceptions). Terrell (1979) suggested that a fine-grained phonetic transcription of /s/ is possible, but would hinder the replicability of results for future investigations. For this reason, most sociolinguists and phonologists have adopted a tripartite system for distinguishing between the innumerable phonetic manifestations of /s/ with the three IPA symbols: [s] All phones with some sibilance (i.e. a high-frequency hissing noise caused by the turbulence of air passing between the alveolar ridge and the tongue tip; [h] Phones normally aspirated, sometimes very weak, often voiced or nasalized and possibly assimilated resulting in a geminate consonant cluster; Ø Complete absence of a phone representing /s/. Based on this categorization of /s/, two varieties of Spanish have been identified in the Hispanic Linguistics literature: 1. Retention: Speakers tend to produce /s/ as [s] in all contexts, whereas aspiration (i.e. [h]) and deletion (i.e. Ø) are almost non-existent. This is the standard variety spoken in the capital of Colombia (Bogotá). 2. Aspiration/deletion: Speakers tend to produce /s/ as [h] or Ø. Although [s] appears in some formal registers, it is not very frequent in informal speech. This is the standard variety spoken in the Atlantic coastal region of Colombia.

MURIEL/OROZCO

128

31/5/12

14:24

Página 128

RICHARD J. FILE-MURIEL

S-lenition in Spanish is perhaps the most studied phenomenon in Romance Linguistics. Ferguson (1990:64) goes so far as to state that “the aspiration and deletion of /s/ in dialects of Spanish may be the most extensively treated of all sound changes being investigated from an empirical, variationist perspective.” A likely reason for the plethora of studies detailing this phenomenon is that it is fairly wide-spread in the Spanish-speaking world, occurring in the southern part of Spain (Andalusia), the Canary Islands, and throughout Latin America, excepting the Mexican highlands and the Andean regions of South America. Referring specifically to Spanish-speaking Latin America, Lipski (1994) reports that s-lenition is found in the Caribbean zone, including coastal Mexico, Colombia, and Panama and throughout Central America (excepting Costa Rica and Guatemala), the Pacific coast of Colombia and Ecuador, coastal Peru, Chile, Paraguay, Eastern Bolivia, Uruguay, and most of Argentina. In addition to geographic origin, it has been shown that a multitude of other factors influence the production of /s/, including linguistic, sociolinguistic, stylistic, and functional factors. The following merely samples the abundance of work on s-lenition in Spanish. Variation in the distribution of /s/ has been explained in terms of varying sociolinguistic characteristics of the speakers, such as socioeconomic status and education level, age, gender, and whether the speaker resides in an urban or rural location. Aspiration and deletion are generally considered markers of social class, with upper-class and more-educated speakers tending towards less weakening, while lower-socioeconomic class and less-educated speakers favor more lenition. Terrell (1981) examined the distribution of /s/ in Santo Domingo, finding that education level, task, and gender were significant in explaining the realization of /s/. Cedergren (1973) and Poplack (1979) found that the age of the speaker can also affect the rate of /s/ lenition, with younger speakers displaying more weakening than older speakers. Furthermore, in general, male speakers show higher levels of lenition than female speakers, as the latter are more likely to retain /s/ when lenition is stigmatized (Fontanella de Weinberg 1973). Interestingly, Rodriguez-Castellano/Palacio (1948) observed that rural speakers tend to conserve the word-final /s/ as a voiced aspirate, while urban speakers favor complete deletion. The variants of /s/ (i.e. [s], [h], and Ø) co-vary with register such that the frequency of retention increases in formal speech styles and in reading tasks (Alba 2004; File-Muriel 2009). Increasing the rate of speech has profound effects on the duration of all segments and, not surprisingly, tends to occasion higher incidences of reduction in the form of lenition and assimilatory processes. Segment weakening results because the need for efficiency in coordinating articulatory gestures outweighs the need for precision. Bybee (2002) claims that this is universally true for all segments that undergo lenition. Accelerated speech rates will

MURIEL/OROZCO

31/5/12

14:24

Página 129

A LABORATORY APPROACH TO S-LENITION

129

increase the frequency of stop lenition in Spanish, vowel harmony in Brazilian Portuguese, and flapping in American English, among other phenomena. Lipski (1994) found that style or speed of delivery figure into explaining /s/ variation. By comparing sports commentaries with other forms of broadcasting, Lipski argues that increased speed of delivery favors aspiration and deletion, while more deliberate speed favors retention. It is unclear, however, whether speed of delivery, style, or a combination of both factors contributed to the variation. The linguistic factors that have been found to influence s-realization can be categorized as both lexical and phonological. The position of /s/ within the word has been found to be a reliable predictor of s-realization. Specifically, wordinternal position favors weakening, and word-final position favors retention (Samper Padilla 1990; Terrell 1978; Terrell 1979; Hammond 1981; Vida 2004). With regard to stress, Alba (1982) and Poplack (1979) report that /s/ is more likely to be retained in tonic syllables, whereas a weakening effect is observed in atonic syllables. Poplack (1979) reports that longer words tend to be reduced more than shorter words, a finding also reported in other languages (cf. Johnson 2004 for massive reduction in English). Both the preceding and following phonological contexts have been found to significantly condition s-realization. For example, Terrell reports that a following voiced sound significantly weakens /s/; similarly, Esther Brown/Torres Cacoullos (2003) report that a preceding highvowel favors retention, while non-high vowels promote lenition. The role of lexical frequency has recently received deserved attention. Perhaps the first linguist to articulate the importance of lexical frequency in language change was Schuchardt (1885), who observed that high-frequency words undergo sound change at an accelerated rate compared to low-frequency words, which tend to “behave” more conservatively. Many of the effects of frequency were noted by Zipf (1935), who is recognized for Zipf’s Law. The principal idea behind the law is that given some corpus of natural language utterances, the length of a word is inversely proportional to its frequency. Following the rise of functionalism, which abandons the notion of independence between structure and usage, Bybee (1985) embarked on a line of research investigating the question of how frequency of utterances could affect cognitive representation and grammar. This line of inquiry has developed into what is now referred to as usage-based models of language or exemplar theory. Bybee argues that the reduction of high-frequency words results from the automation of linguistic production as a highly-practiced neuromotor activity (Bybee 2001), while others (c.f. Gregory/Raymond/Bell/Fosler-Lussier/Jurafsky 1999) suggest that high-frequency words are more often reduced than low-frequency words because the former are more predictable in context, so that reduction takes place without interfering with communication. There is a growing

MURIEL/OROZCO

130

31/5/12

14:24

Página 130

RICHARD J. FILE-MURIEL

body of literature aimed at understanding the relationship between frequency and phonological reduction. Multiple cases of this type of reduction have been identified in quantitative studies of different languages: English Schwa deletion (Hooper/Bybee 1976), English t/d deletion (Bybee 2000), Spanish /d/ deletion (Bybee 2002), and s-lenition in several different Spanish varieties (Esther Brown/Torres Cacoullos 2003; Earl K. Brown 2009; Bybee 2002; File-Muriel 2009; File-Muriel 2010; File-Muriel/Brown 2010; Minnick Fox 2006). The role that lexical frequency plays in sound change is not uncontroversial. In much of Labov’s work (Labov 2003), which examines an enormous amount of data on the fronting of the nuclei of the back upgliding diphthongs /uw/, /ow/, and /aw/, almost all variation could be accounted for purely by phonetic constraints. In other words, word frequency played no role at all in the change examined by Labov.

2. Justification for this study The inherently subjective nature of audio transcription, which to date has been the preferred approach to Spanish s-lenition, has not gone unnoticed in the literature. Poplack adequately described what she saw as a methodological concern: “Researchers looking at similar Caribbean dialects have reported grossly different proportions of the same variants. These discrepancies are more likely due to [one researcher] counting the assimilated variants as instances of deletion, while [another researcher] considered them aspiration” (1979: 66). In short, researchers make subjective decisions about what they hear (or think they hear), which becomes highly arbitrary due to the gradient nature of lenition. The fact that transcription is subject to bias based on the expectations and experience of the transcriber is well documented in the literature (c.f. Boucher 1994; Erker 2010; Mann/Repp 1980; Pouplier/Goldstein 2005). Several studies show that the surrounding phonetic context can influence the perception and categorization of segmental phenomena. For example, vowel durations affect the perception of syllable-final stop consonants, which are easier to perceive and categorize when the preceding consonant is shorter in duration (Repp/Williams 1985). Specific to /s/, the surrounding phonetic context has also been shown to influence perception and categorization. Poplack’s early assertion was in fact verified in a perceptual study (File-Muriel/Díaz-Campos 2003), which showed that trained linguists varied significantly in their categorization of controlled, synthesized stimuli, erring especially between different degrees of aspiration and deletion. Examining the perception and categorization of different variants of /s/ using synthesized speech, they found that in pre-pausal position listeners categorized the

MURIEL/OROZCO

31/5/12

14:24

Página 131

A LABORATORY APPROACH TO S-LENITION

131

aspirated and deleted variants at only 46-percent accuracy, which indicates the great difficulty of distinguishing between these two phonetic realizations. Secondly, capturing the subtle (but vital) acoustic variation that exists in different manifestations of /s/ is not possible using the transcription approach, as it limits the representation of this gradient phenomenon to the symbolic units of the IPA. Widdison (1991; 1994; 1995) has shown conclusively that subtle acoustic properties influence perception in lexical decision tasks, even if the listeners are not conscious of the material in the signal that influences such decisions. Erker (2010) demonstrates that, in the Spanish of Dominicans in New York, a strictly segmental description of coda /s/ productions groups together tokens that are significantly different from one another acoustically, thus concealing important patterns present in speech. He reports that, within the class of tokens that were coded as [s], significant differences existed in both of the acoustic measurements used in the study. Furthermore, these differences were correlated with several of the independent variables in the study and several of the conditioning factors only promote weakening in one dimension of the subsegmental description. Parallel to this, File-Muriel/Brown (2010) examine slenition in the Spanish of Cali, Colombia, and report that certain factors, such as following high-vowel, promote retention in one dimension (i.e. s-duration), while advancing lenition in another (i.e. center of gravity). The findings in these studies suggest that the weakening process is more nuanced than what is portrayed by the symbolic representation of /s/ as [s], [h], and Ø. In light of these methodological concerns, the present study looks at s-realization in gradient terms in order to capture the subtle acoustic details that could be relevant to our understanding of the lenition process. I intend to address the following questions: (1) Which independent variables significantly condition srealization in the Spanish of Barranquilla, Colombia when viewed as gradient? (2) Are these independent variables the same ones that have been reported in the previous studies, which are based on IPA transcription? (3) Do the independent variables have the same magnitude of effect in each of the three dependent variables? (4) What advantages, if any, are there to considering lenition in gradient terms as opposed to categorical ones?

3. Methods 3.1. DATA COLLECTION/CORPUS The data used in the present study come from sociolinguistic interviews conducted in Barranquilla, Colombia. The participant pool consisted of 13 native residents of

MURIEL/OROZCO

31/5/12

14:24

Página 132

132

RICHARD J. FILE-MURIEL

Barranquilla, including seven females and six males between 18 and 55 years of age; no participant reported any speech or hearing disorder. The participants in this study were categorized according to socioeconomic class based on their answers to a questionnaire, which assessed income, occupation, education level, housing location, as well as the same information about parents. The interviews (approx. 30 minutes each) were conducted by a native of Barranquilla, Colombia, who met the same criteria as the participants she interviewed. The interviews adhered to standard sociolinguistic methodology; topics included vacation and travel plans, pastimes, schooling, food, dangerous situations, local shopping, among other things. The interviews took place in a quiet setting and were recorded using a solid-state Marantz PMD 670 compact flash recorder with a head-mounted unidirectional mic.

3.2. DEPENDENT VARIABLES The coding of dependent variables follows File-Muriel/Brown (2010). A sequential set of 75 words that contained a syllable-final lexical /s/ was selected, starting at least ten minutes1 into each interview, producing a total of 1,000 tokens for analysis. In order to code the three dependent variables (s-duration, center of gravity, and percent voiceless), the researcher used Praat (v. 5.1.15 for Mac). The first of the three dependent variables, s-duration, was coded by manual delimitation. Attending to both the waveform and the spectrogram, the researcher delimited the left and right boundaries of the visible aperiodic, high-frequency (4,000-11,000 Hz) noise. Ladefoged suggests that “spectrograms cannot give such precise information in the time domain as expanded scale waveforms, which readily permit measurements in milliseconds” (Ladefoged 2003). The boundaries were placed at the zero-intercept in the waveform. Specifically, the onset of /s/ sequences was delineated at the zero-intercept point closest to the first sign of aperiodic noise in the waveform. Similarly, the offset was set at the zero-intercept point closest to the cessation of the aperiodicity associated with the /s/ sound (during the closure of a following voiceless stop there is often aperiodic ambient noise, which is visible in the waveform, but which shows no high-frequency noise in the spectrogram, hence the utility of viewing both the waveform and the spectrogram)2. See Figure 1.

1 The decision to exclude tokens from the first ten minutes of each interview is based on the fact that speakers tend to be most conscious of their speech at the beginning of a recorded conversation, and are therefore more likely to display hypercorrection and style-shifting at the beginning. 2 The methodology used in the present study is based on File-Muriel/Brown (2010), where inter-rater reliability was tested by the independent coding of 25 tokens (selected at ran-

MURIEL/OROZCO

31/5/12

14:24

Página 133

A LABORATORY APPROACH TO S-LENITION

133

FIGURE 1 Waveform and spectrogram of sibilance in que eStá moviendo ‘that’s moving’

Tokens of /s/ that displayed no visible high-frequency noise in the spectrogram nor aperiodicity in the waveform were coded as having a duration of 0 milliseconds. See Figure 2. The band of low peaks and valleys in the waveform and the lighter shades of gray in the spectrogram represent the closure for the wordinitial approximant in “bares.” Due to the fact that fricative consonants produce highly variable noise, statistical techniques are quite useful in investigating the variation in fricative spectra. The second dependent variable in this study includes one of the spectral moments, namely, center of gravity (henceforth centroid), which measures the central tendency (i.e. the mean) of the spectrum (for more on the four spectral moments, the reader is directed to Forrest/Weismer/Milenkovic/Dougall (1988). The centroid measurement was taken from the middle 60-percent of the duration measurement in order to avoid the transitional boundaries, which were intentionally excluded in order to reduce the influence of surrounding segments on the measurements3. Following Silbert/de Jong (2008: 2772), the centroid was limited to frequencies above 750 Hz with a pass Hann band filter, as this was intenddom) by both researchers; independent measurements all fell within 10 ms, indicating a highdegree of precision, reliability, and replicability of the stated procedure. 3 This decision is not meant to imply that the transitional boundaries do not contain important information (cf. Widdison 1994); the transitional boundaries were included in the duration measurements.

MURIEL/OROZCO

31/5/12

14:24

Página 134

134

RICHARD J. FILE-MURIEL

FIGURE 2 Waveform and spectrogram of the deleted /s/ in dos bares ‘two bars’

ed to capture the noise component of the fricatives; energy produced by glottal pulsing is located predominantly in the frequency range below 750 Hz. Finally, the third dependent variable, percent voicelessness, was taken from the “Voice Report” in Praat. This measurement was taken when the marked duration of /s/ filled exactly two-thirds of the Editor Window. Aside from the initial manual delimitation of the duration boundaries, the procedures described above were performed automatically using a Praat script. There are several advantages of using a script. First, in terms of efficiency, Praat takes all of the measurements automatically in a matter of minutes. Secondly, and more importantly, using a script creates uniformity in the coding process. For example, the script pulled exactly 60 percent of each token to measure the centroid and zoomed into the token so that it filled exactly two-thirds of the Editor Window before obtaining the voicing report, regardless of whether the duration was 20 or 120 milliseconds.

3.3. INDEPENDENT VARIABLES This study analyzes the conditioning effect of ten independent variables, including: (1) Local speaking rate, (2) Word length, (3) Preceding phonological con-

MURIEL/OROZCO

31/5/12

14:24

Página 135

A LABORATORY APPROACH TO S-LENITION

135

text, (4) Following phonological context, (5) Voicing of following segment, (6) Word position, (7) Prosodic stress, (8) gender, (9) socioeconomic class, and (10) Lexical frequency (including bigram frequency). Local speaking rate was calculated based on the duration of the word (measured in seconds) containing /s/ divided by the number of phonemes in the word (assuming full articulation), which produced a scalar measurement of phonemesper-second. Inclusion of this variable was based on a study by Lipski (1985a) which reports a significant correlation between registers of different speaking rates and s-realization; fast registers (such as sports broadcasting) favor higher rates of reduction than more deliberate registers. Word length was counted as the number of phonemes in the word (assuming its full articulation), which created an ordinal variable with a range between three (e.g. más ‘more’) and seventeen (responsabilidades ‘responsibilities’) phonemes. Méndez Dosuna (1985), examining a variety of peninsular Spanish, reports that as word length increases, the duration of individual units decreases. Similar results have been reported in other languages. For example, Johnson (2004) notes that longer words are subject to massive reduction in English. Both the preceding and following phonological contexts were considered in light of several studies that report the assimilatory effects of surrounding sounds. For example, Esther Brown (2006) notes significant differences in the realization of syllable-initial /s/ based on the preceding vowel, reporting that a preceding non-high vowel (a, e, o) favors reduction, while a preceding high vowel (i, u) favors retention. Parallel to this, studies in laboratory phonetics show that the duration of “s” in consonant clusters (e.g. sp, sk, st) is not uniform and in fact varies according to the obstruent that follows (c.f. Borden/Gay 1979; Méndez Dosuna 1985; Schwartz 1970). For the present study, the preceding context was coded according to three categories: High vowel, non-high vowel, and consonant. The following context was categorized according to five: Pause, vowel, or consonant (including coronal, velar, and labial consonants). Given that Poplack (1979) found that a following voiced segment promotes higher rates of s-lenition than a voiceless segment, the sound following /s/ was also coded as voiced or voiceless. Word position was coded dichotomously as medial (e.g. hasta ‘until’) or final (e.g. país ‘country’). Past studies (Hammond 1981; Samper Padilla 1990; Terrell, 1979b, 1982) have found more reduction word-internally, where /s/ is often followed by a consonant and the environment is invariable, whereas higher rates of retention are reported word-finally, where the environment varies according to the word that follows. Prosodic stress was coded with respect to the syllable containing /s/. In other words, whether /s/ occurs in a tonic (e.g., semestre ‘semester’) or atonic (testigo

MURIEL/OROZCO

31/5/12

136

14:24

Página 136

RICHARD J. FILE-MURIEL

‘witness’) syllable. In several previous studies (Alba 1982; Poplack 1979), stress was found to significantly influence s-realization: When tonic stress falls on the syllable containing /s/, retention is preferred; whereas atonic stress promotes lenition. Almost all previous studies that have examined Spanish s-weakening varieties quantitatively have found that both gender and socioeconomic status of the speaker are reliable predictors of s-realization. Regarding gender, higher rates of lenition tend to characterize male-speech, while s-retention is favored by female speakers (Stewart 1999). With respect to socioeconomic class, the most prevalent trend is that speakers of higher socioeconomic status tend to retain /s/ more than speakers of lower status. Lexical frequency was measured by calculating the number of occurrences each token word occurs in a specific corpus. Three different corpora were used to calculate three independent measures of lexical frequency using the open source programming language R (R Development Core Team). The first corpus is the Corpus Oral del Español Barranquillero (CODEB), which consists of 212,000 words from sociolinguistic interviews collected between 2004-9. The second corpus is the Habla Culta, consisting of 2.5 million words and is made up of interviews from 12 major cities in the Spanish-speaking world, including Bogotá, Buenos Aires, Caracas, Havana, La Paz, Lima, Madrid, Mexico City, San José, San Juan, Santiago, and Seville. The third corpus that was considered was the Corpus Oral de Referencia del Español Contemporáneo, which consists of 2.1 million words (Marcos-Marín 1992). These corpora were used to calculate the number of occurrences of individual words. Bigram-one frequency measures the number of occurrences of the two-word string in which the s-word is preceded by another word (e.g., no estábamos ‘we were not’). Bigram-two frequency measures the number of occurrences of the two-word string in which the s-word is followed by another word (e.g., estábamos preparados ‘we were prepared’).

4. Results Linear regressions were run with SPSS (v. 16.0 for Mac) for each of the three dependent variables (s-duration, centroid, and voicelessness) using a forward stepwise selection procedure in which the independent variables are sequentially entered into the model based on which ones have the largest correlation with the dependent variable. The results in Table 1 exclude the bigram frequencies, as their inclusion excluded the several hundred tokens in which the s-word was either preceded or followed by a pause, as a pause followed by a word or a word followed by a pause are not bigrams. Therefore, the bigram frequencies were not

MURIEL/OROZCO

31/5/12

14:24

Página 137

137

A LABORATORY APPROACH TO S-LENITION

included in the linear regressions reported in Table 1, but their influence was assessed in a second set of regressions that will be discussed shortly. The standardized coefficients given in Table 1 indicate that the corresponding independent variable significantly contributes (at an alpha level of p < 0.05) to the dependent variable under investigation. These numbers can be used to interpret the strength and direction of influence. Standardized coefficients furthest away from zero, as either a positive or a negative number, indicate a strong influence on the dependent variable. The order of inclusion of each variable in the model precedes the standardized coefficient, and indicates the degree of correlation with the dependent variables. For example, the independent variable “following pause” was selected as significant in all three regressions and has coefficients relatively far from zero, which seems to indicate that it is a very important predictor of s-realization. On the other hand, “socioeconomic status” was selected only for centroid, indicating significant influence, but less than that attested for “following pause.” TABLE 1 Summary of standardized coefficients taken from three linear regressions (bigram frequencies excluded) Variables

S-duration

Centroid

Voicelessness

Following pause

1) 0.577

3) 0.175

3) 0.169

Following labial consonant

3) –0.093

1) –0.304

4) –0.147

Following voiced sound

5) –0.118

4) –0.216

1) –0.196

Gender (female)

4) 0.110

2) 0.243

2) 0.246

Speaking rate

2) –0.194

6) –0.151

5) –0.158

Following vowel

6) 0.067

ns

ns

Following velar consonant

ns

5) –0.203

ns

Socioeconomic status (3-5)

ns

7) 0.111

ns

Lexical frequency (CODEB)

ns

ns

6) –0.058

0.509

0.365

0.240

R-Square

Not selected for any dependent variable: Lexical frequency (COREC), Lexical frequency (Habla Culta), preceding phonological context, word position, stress, and word length.

MURIEL/OROZCO

138

31/5/12

14:24

Página 138

RICHARD J. FILE-MURIEL

The results in Table 1 allow us to make several fundamental observations. First, several independent variables stand out as powerful predictors of s-realization. They are: following phonological context, speaking rate, and gender. Each of these variables was selected as significant for all three dependent variables at an alpha level of p < 0.001 and show the same direction of influence for all three measures. For example, as local speaking rate increases, s-duration, centroid, and voicelessness decrease, which indicate a strong weakening tendency. Likewise, when /s/ is followed by a labial or a voiced sound, the acoustic measures of /s/ decline dramatically. On the other hand, when /s/ is followed by a pause, the opposite effect holds: s-duration, centroid, and voicelessness increase, indicating that these factors strongly favor retention. Gender is also a reliable predictor of s-realization, with female speakers producing stronger s-correlates than male speakers across all three measures. Several of the independent variables were significant predictors of only one acoustic measure (i.e. dependent variable). This is important because such information is concealed within a categorical account of s-realization. A following vowel appears to significantly condition duration, while not affecting centroid or voicelessness. A following velar consonant seems to lower the centroid, while not affecting duration or voicelessness. Similarly, socioeconomic status of the speaker also appears to impact the centroid only, with speakers of higher socioeconomic status producing /s/ with a higher centroid than speakers of lower status. Finally, lexical frequency appears to affect only the degree of voicelessness, with high-frequency words favoring more voicing assimilation than low-frequency words. The results for the following phonological context can be contrasted with the preceding phonological context, which was not selected as significant in any of the three regressions. In addition to the preceding phonological context, the results of the present study suggest that word position, stress, and word length are not significant predictors of s-realization. These findings contradict results reported in the previous research that applied the traditional transcription approach. With regard to lexical frequency, it’s important to highlight that only the local corpus (CODEB) was selected as significant, albeit for only one independent variable (voicelessness). The much larger global corpora, which included the COREC and the Habla Culta, were not selected for any of the dependent variables. This is not surprising, given that the latter corpora sample usage patterns from different varieties of Spanish, to which the participants in this study have had (at most) limited exposure during their lifetime. As mentioned above, the linear regressions reported in Table 1 exclude the bigram frequencies to allow the inclusion of all the tokens in this study (N=1,000). In order to address the

MURIEL/OROZCO

31/5/12

14:24

Página 139

A LABORATORY APPROACH TO S-LENITION

139

influence of the bigram frequencies, three more linear regressions were run in which these frequency measures were included, even though it meant the exclusion of several hundred tokens. Neither measure of bigram frequency was found to be significant.

5. Discussion When viewed in gradient terms, the results suggest that s-realization in the Spanish of Barranquilla is influenced by a variety of factors, including following phonological context, gender, local speaking rate, socioeconomic status, and to some extent, lexical frequency. Of all the factors examined, following phonological context, gender, and local speaking rate appear to be the most significant predictors of the phenomenon, as they were selected in all three regressions and the coefficients indicate a strong and consistent pattern of influence across all three measures in terms of directionality. Perhaps the least surprising factor reported in this study is gender. This factor seems to exert the same influence as reported in numerous other varieties. With regard to speaking rate, these findings show the importance of including speaking rate as a control variable in any type of sociophonetic study, providing more evidence that an increased rate of speech tends to occasion higher incidences of lenition and assimilatory processes (c.f. Bybee 2002; Lipski 1985a). In accelerated speech, the need for efficiency in coordinating articulatory gestures outweighs the need for precision. Concerning the influence of neighboring sounds, the following phonological context was found to be a very strong predictor of s-realization, whereas the preceding context does not appear to have significant influence. These findings support the broad general claim that Spanish is a language characterized more by regressive assimilatory processes than progressive ones (Schwegler/Kempff/Ameal-Guerra 2010: 243-246). This study highlights a few of the advantages of using instrumental acoustic measurements in lieu of traditional auditory analysis. The gradient approach allows us to make detailed observations regarding how temporal, spectral, and energy properties of /s/ are modeled across linguistic, social, and stylistic factors. It was pointed out that the acoustic correlates associated with the sound /s/ are not affected equally by all of the factors examined, which, to my knowledge, is a novel finding regarding s-lenition in Spanish. For example, the socioeconomic class of the speaker significantly conditions the centroid, while not affecting the other measures (duration and voicelessness). Such information is impossible to capture using traditional IPA categories, which collapse all relevant acoustic cues into one or more categorical labels (i.e., s, h, Ø, etc.).

MURIEL/OROZCO

140

31/5/12

14:24

Página 140

RICHARD J. FILE-MURIEL

Similarly, by employing gradient measurements of both dependent and independent variables we gain a more accurate account of the conditioning factors previously reported in the literature. For example, Terrell (1979) proposes that word length significantly conditions s-realization in Cuban Spanish. Specifically, he employs a binary distinction of word length and shows that polysyllabic words are more likely than monosyllabic words to have a lenited /s/. However, when analyzed as a gradient variable by the number of phonemes (which in these data range from three phonemes (e.g., más ‘more’) to seventeen (e.g., responsabilidades ‘responsibilities’), word length was not selected as a significant predictor of any of the three dependent variables reported in Table 1 above. In conclusion, I have presented evidence supportive of the claim that continuous acoustic measures (when available) are preferred to symbolic representation. First, continuous variables allow the researcher to address the actual acoustic parameters that are immeasurable with categories, making possible the detailed study of gradient phenomena. Second, by analyzing s-realization in terms of three dependent variables, we are able to observe the variable influence of the independent factors that have long been studied in the literature. Third, scalar measures analyzed with software (such as Praat) are less vulnerable to transcriber bias, as transcription encourages the researcher to impose segmentation on inherently gradient phenomena.

MURIEL/OROZCO

31/5/12

14:24

Página 141

L A VA R I A B L E E X P R E S I Ó N D E L A I M P E R S O N A L I D A D E N E L E S PA Ñ O L DE LOS COLOMBIANOS DE BOGOTÁ Y MIAMI LUZ MARCELA HURTADO CUBILLOS

1. Introducción La expresión de la impersonalidad en español se realiza mediante formas como uno, se, usted, y tú, pronombres singulares que no identifican a un sujeto concreto: la acción del verbo es susceptible de ser desarrollada por un agente, pero éste queda encubierto (Muñiz Cachón 1998: 55-61). La utilización de estos pronombres impersonales depende tanto de condicionamientos semántico-pragmáticos como gramaticales, “[…] de las conexiones que se establezcan o dejen de establecerse entre los participantes de la acción y los del discurso” (Maldonado 1999: 206). En cuanto a los condicionamientos gramaticales, se ha determinado que el aspecto y la clase semántica del verbo contribuyen a la lectura genérica: los verbos modales y el aspecto imperfectivo (el presente habitual) son clave para el empleo impersonal de la segunda persona singular (Hernanz 1990: 157) y la lectura cuasi-universal de se (Maldonado 1999: 221). La mayoría de los análisis concuerdan en que se trata de una impersonalidad semántica y pragmática, cuya interpretación depende del contexto (Gómez Torrego 1998). El hablante utiliza una forma u otra conforme al grado de generalización que intente lograr. Se alcanza el máximo grado de indeterminación del sujeto: el hablante puede no sólo ocultar su identidad sino también referir al oyente o a una(s) tercera(s) persona(s) (Vera Lujan 1990: 88). Evita ponerse a sí mismo en “primer plano de interés” (García González 1996: 389) y referir su propia identidad como sujeto único de la predicación o como parte de un grupo determinado: se reduce la responsabilidad del emisor (Maldonado 1999: 215). En cambio, el pronombre indefinido uno no resulta tan impersonal, pues existe diferencia de género (uno/a). Incluye al emisor, al receptor o a una tercera persona, aunque no los identifica explícitamente (Muñiz Cachón 1998: 87-91). El carácter genérico también se logra con la segunda persona singular (tú o usted), cuando se emplea sin referir exclusivamente a un interlocutor (Muñiz Cachón 1998: 67). A pesar de ser una estrategia desfocalizadora (Hidalgo Navarro 1996: 173), el tú impersonal connota familiaridad, pues no se elimina totalmente el contenido gramatical propio de esta forma (Carrasco 1986: 393). Los

MURIEL/OROZCO

142

31/5/12

14:24

Página 142

LUZ MARCELA HURTADO CUBILLOS

estudiosos coinciden en que el uso impersonal de tú es un recurso estilístico, propio del habla coloquial y familiar (Gómez Torrego 1998; Hidalgo Navarro 1996), muy común en España. Gorosch (1967) evidencia dicho empleo tanto en el habla como en la forma escrita (en diarios y revistas semanales que imitan la conversación cotidiana, en textos literarios). Con base en sus observaciones durante los años cuarenta, Lorenzo (1984: 278) se arriesga a plantear un posible origen levantino de este uso y a caracterizarlo como típico de las generaciones más jóvenes. Además, menciona la posibilidad de la influencia de “fuerzas extralingüísticas que actúan constantemente en el lenguaje” (Lorenzo 1989: 228). DeMello (2000: 365) proporciona algunos porcentajes de uso de los impersonales. Es un estudio comparativo del habla de Buenos Aires, Caracas, La Habana, La Paz, Lima, Madrid, México, San Juan, Santiago de Chile y Sevilla, cuyos datos fueron recogidos a finales de los años setenta y comienzos de los ochenta. DeMello evidencia la conexión entre el impersonal y las formas de tratamiento típicas del hablante: el usted impersonal se encuentra casi exclusivamente en los hablantes que se tratan de usted, y el tú impersonal ocurre principalmente en las variedades que tutean. No obstante, es más frecuente el empleo impersonal de tú en hablantes cuya variedad ustedea, que la utilización de usted impersonal entre los que tutean. Los estudios hasta aquí referidos analizan el influjo del estilo o registro, de la edad y de la variedad de español en la utilización de los pronombres impersonales. Faltaría mencionar el posible efecto de factores externos como el contacto con una segunda lengua (por ej., el inglés) o el contacto dialectal.

1.1. INFLUENCIA DE LA SITUACIÓN DE CONTACTO Los estudios sobre el influjo del contacto interlingüístico e interdialectal en el uso de los pronombres impersonales son escasos y contradictorios. En España, por ejemplo, Lorenzo (1989: 229) sugiere –mas no lo comprueba– que el predominio de tú impersonal es un calco del inglés, lengua cuya segunda persona del singular se utiliza considerablemente para generalizar. Por el contrario, Vila (1987) observa la tendencia a usar tú en vez de uno entre los jóvenes españoles de comunidades que no estaban en contacto con el inglés. En situaciones de intenso contacto de lenguas, es probable que se produzcan cambios directos e indirectos. El grado de incorporación de la lengua en contacto depende tanto del grado de bilingüismo de los hablantes como de la duración de dicha coexistencia (Palacios 2007a: 266-267). Morales (1995), en su estudio del español hablado en Puerto Rico, revela la influencia de factores sociales como el nivel de bilingüismo y la edad del hablante: los más jóvenes y bilingües

MURIEL/OROZCO

31/5/12

14:24

Página 143

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

143

favorecen uno y tú impersonal en detrimento de se (155-156). Éste es uno de los pocos estudios empíricos que proveen evidencia cuantitativa de la influencia. Por su parte, Lipski (1987) explica la utilización del tú impersonal como un anglicismo sintáctico presente en el español isleño de Luisiana. Otheguy/Zentella (2007), Hurtado (2001, 2005) y Orozco (2007a) son estudios que incluyen a los colombianos residentes en Estados Unidos. Otheguy/Zentella (2007: 293) evidencian la influencia del factor de contacto interlingüístico en la frecuencia de expresión de pronombres de sujeto por parte de los hablantes bilingües de segunda generación que viven en Nueva York. En contraste, en el análisis de la expresión de pronombres de sujeto de los colombianos residentes de Miami, Hurtado (2001: 177) no encuentra diferencias significativas en las tasas pronominales de los que nacieron en Colombia y llegaron a EEUU después de los 11 años de edad (51%) y de los que nacieron en Estados Unidos o llegaron antes de cumplir los 11 años (47%). Sin embargo, en este trabajo tampoco se descarta la posible influencia de la intensidad del contacto con el inglés, ya que el grupo que estaba cursando estudios universitarios favoreció más el uso pronominal (peso Varbrul: .59). En cuanto a la expresión de futuro, Orozco (2007a: 311) sugiere que la situación de contacto no genera diferencias entre los colombianos residentes de Nueva York y los de Barranquilla, Colombia. Sin embargo, el autor plantea que el cambio del futuro morfológico por el perifrástico, detectado en Barranquilla, se ha acelerado en el contexto de NY. Según Palacios (2007a: 267), éste sería un caso de cambio indirecto, en el que la situación de contacto con el inglés genera variaciones significativas, “[…] generalmente en el registro oral coloquial de la lengua B, que aprovechan la propia evolución interna de esa lengua B”. En una situación de contacto de lenguas, también ocurre que aunque los inmigrantes se adaptan a la situación anglohablante, continúan con núcleos de vida social hispanohablante. Esto posibilita el contacto entre diversos grupos hispanos y promueve la escogencia de algunas formas, como sucede con la preferencia por ciertos pronombres de tratamiento. Hernández (2002) y Schreffler (1994), por ejemplo, evidenciaron que la comunidad hispana minoritaria de salvadoreños residente en Houston se acomoda al uso pronominal del grupo mayoritario mexicano en el momento de interacción con miembros de ese grupo. Por el contrario, el estudio de Lamanna (este volumen) sobre empleo de la segunda persona por parte de los colombianos de Carolina del Norte, no comprobó que éstos se acomodaran a los usos de los mexicanos. Este trabajo de Lamanna, como los de Hurtado en Miami y Orozco en Nueva York, vislumbra la posibilidad de que la situación de contacto con anglohablantes o con otros grupos hispanos no influya en el español de los colombianos de la misma manera en que ocurre con otros grupos hispanos.

MURIEL/OROZCO

144

31/5/12

14:24

Página 144

LUZ MARCELA HURTADO CUBILLOS

Por último, Palacios (2007b) muestra el efecto del tiempo de permanencia en un contexto de interacción de variedades de español. La autora analiza la influencia del español madrileño en el español de los ecuatorianos residentes en la ciudad de Madrid y determina que cuanto “mayor tiempo de residencia en España mayor es el abandono de la norma ecuatoriana aunque no haya adscripción a la norma madrileña” (121).

1.2. ACERCAMIENTO DE ESTE ESTUDIO Ya que la mayoría de los artículos sobre la expresión de la impersonalidad han demostrado la relevancia de factores pragmáticos (la situación de comunicación, el hablante, el tipo de discurso, etc.) en la utilización de los impersonales, nuestro objetivo general es identificar las causas externas que determinan el predominio de cierta forma impersonal en situaciones de contacto de lenguas y dialectos. La escasez de estudios empíricos sobre la incidencia de los factores extralingüísticos motiva el presente trabajo, realizado desde la perspectiva variacionista. Para demostrar que la expresión de impersonalidad obedece también a una conjunción de variables, no sólo incluimos las variedades principales del español colombiano (costeña y andina), sino también dos contextos de interacción en los que se plantean situaciones de contacto diferentes. La situación sociolingüística en Bogotá es una de contacto interdialectal, pues los sucesos políticos y económicos del país han forzado la inmigración de colombianos de diversas zonas del país a esta ciudad. En Miami, los inmigrantes colombianos se adaptan socio-culturalmente a un contexto en el que coexisten diversas variedades de español con la lengua inglesa. Los aproximadamente 102 mil colombianos que residen allí, interactúan no sólo con anglohablantes, sino con alrededor de 800 mil cubanos, 100 mil nicaragüenses, 88 mil puertorriqueños, 49 mil dominicanos, 48 mil hondureños, 42 mil mexicanos, 41 mil venezolanos y 34 mil peruanos (datos del censo de 2006 citados en Klee/Lynch, 2009: 201). En esta ciudad, el español ocupa un lugar privilegiado y sus hablantes lo reconocen como un “vehículo lingüístico de una gran cultura internacional” (López Morales 2003: 92). Es una ciudad de mayoría hispanoamericana, “donde se habla español en todas partes” (Lynch 2009: 387). Según Hurtado (2001: 4276), los colombianos se comunican en español en la casa, con los amigos y, casi la mayoría, en el trabajo. Es posible que los de Miami sufran un proceso de acomodación dialectal dada la interacción con colombianos de diferentes regiones e hispanohablantes de distintos orígenes. Sin embargo, este proceso puede ser de acomodación temporal, por las exigencias del momento específico de comunicación con otros hispanohablantes (Hurtado 2002: 168; Lynch 2009: 397).

MURIEL/OROZCO

31/5/12

14:24

Página 145

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

145

Basados en los hallazgos sobre la incidencia de situaciones de contacto en el favorecimiento de ciertas formas lingüísticas (Palacios 2005, 2007a, 2007b; Otheguy/Zentella, 2007; Silva-Corvalán 1994) y en las tazas pronominales de los impersonales en los colombianos de Miami (Hurtado 2001: 156; 2005: 193), la hipótesis de partida es que, aunque en el habla de los colombianos predomina el empleo de uno, las diferentes situaciones de contacto generan variabilidad en el sistema pronominal impersonal. Para comprobar esta hipótesis nos planteamos las siguientes preguntas de investigación: ¿Hay diferencias de uso de los pronombres impersonales en Bogotá y Miami? ¿Qué factores sociales y dialectales propician estas diferencias? ¿Qué tipo de influencia (directa o indirecta) ejerce la situación de contacto en el sistema pronominal impersonal de los colombianos de Miami? Esta investigación intenta proporcionar apoyo estadístico a los estudios sobre la impersonalización en español y esclarecer un poco más la posibilidad de influencia de las situaciones de contacto interlingüístico e interdialectal. Nos interesa la variabilidad del sistema pronominal impersonal, en especial, el caso de tú, cuyo empleo en la expresión de un sujeto indeterminado se remonta al latín (Batllori 1998: 385); amerita demostrar si las situaciones de contacto han ocasionado su reactivación en los dos últimos siglos y qué variables contribuyen a su avance.

2. Metodología 2.1. POBLACIÓN Y MUESTRA Ya que el objetivo primordial de este estudio es averiguar la influencia de la situación de contacto en el contexto de Miami, se analizaron dos muestras. Una de las muestras consiste en 43 entrevistas/conversaciones con colombianos y colombo-americanos residentes del área metropolitana de Miami, grabadas a comienzos del año 2000 en Kendall (38) y Cutler Ridge (5). Kendall, sector suburbano al suroeste de Miami, es una zona pujante que ha experimentado un gran crecimiento y desarrollo desde los años ochenta, cuyos promotores han sido los jóvenes profesionales hispanos de distintos orígenes (Pérez 1992: 90). Las edades de los participantes oscilaban entre los 20 y los 70 años. Procedían principalmente de la ciudad de Barranquilla (15), de Cartagena, Lorica y Santa Marta (6), de los departamentos de Antioquia, Caldas, Quindío y Risaralda (14) y de la capital (8). La segunda muestra consiste en 77 entrevistas recogidas en Bogotá. Se incluyeron tanto 30 relatos semilibres reunidos y publicados por el Instituto Caro y

MURIEL/OROZCO

31/5/12

14:24

146

Página 146

LUZ MARCELA HURTADO CUBILLOS

Cuervo (Montes Giraldo et al. 1997), como 47 entrevistas/conversaciones recogidas de 2005 a 2008 mediante el método de redes sociales de interacción. En esta muestra, las edades de los participantes fluctúan entre los 18 y 70 años. Debido a que no existen trabajos cuantitativos sobre estos usos en Colombia comparables al análisis cuantitativo que nos proponemos aquí, esta muestra constituye nuestro punto de partida. TABLA 1 Lugar de entrevista y población Lugar

Mujeres

Hombres

Edad

# Entrevistados

Miami, EEUU

23

20

20-70

43

Bogotá, Colombia

37

40

18-70

77

Total

60

60

120

Se seleccionaron y codificaron en Excel 5,369 enunciados con pronombres impersonales singulares. Una vez codificados los datos, se sometieron al programa estadístico de análisis de regla variable Varbrul 2005. Con este programa es posible analizar y comparar de manera simultánea todas las variables, para determinar cómo se relacionan con la variable dependiente (con determinado pronombre impersonal). El programa elimina las variables que no son significativas y escoge aquéllas que condicionan la variable dependiente. 2.2. VARIABLE DEPENDIENTE: PRONOMBRE IMPERSONAL TÚ, USTED, UNO Y SE Para interpretar los pronombres como impersonales, se tuvo en cuenta los criterios semántico-pragmáticos que expusimos en la introducción, es decir, se realizó con base en el análisis del contexto del enunciado. En el ejemplo 1, consideramos el empleo de tú como impersonal, ya que el entrevistado se refiere a su situación personal con empleo de uno y tú. (1) En mi caso, uno está muy predispuesto... pues si tú no conoces muy bien a esa persona, se tiene que fijar uno mucho, donde tú te vas a meter (entrevista 15, 2000, Miami).

En el fragmento 2, el hablante emplea un usted impersonal cuando explica una situación hipotética que incluye la experiencia del hablante y posiblemente la de un grupo general.

MURIEL/OROZCO

31/5/12

14:24

Página 147

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

147

(2) Entonces pues digo que ya es como la cuestión de cómo uno lo está viviendo. Si uno no… como dice, aquí utilizamos nosotros, no da papaya, pues no hay problema. No dar oportunidad, pues hay sitios… que usted puede andar normal y no le va a pasar nada (entrevista 6, 2005, Bogotá).

Además de enunciados que contenían verbos conjugados, se incluyeron aquellos con verbos en infinitivo (ejemplo 3). Contamos con numerosos casos de pronombres impersonales como sujeto de infinitivo: (3) En Bogotá es raro uno encontrar gente abierta (entrevista 20, 2006, Bogotá).

En cuanto a se, examinamos tanto los usos generalizadores –cuya referencia es una tercera persona– como los no generalizadores– los que en realidad están refiriendo a la primera o segunda persona. Así pues, en el ejemplo 4 el referente de se puede identificarse con el hablante: (4) Y yo creo… uno se da cuenta siempre si alguien es de fuera de la ciudad. Entonces se puede decir que un bogotano tiene… (entrevista 10, 2005, Bogotá).

2.3. VARIABLES SOCIOLINGÜÍSTICAS PARA LAS DOS MUESTRAS Consideramos las variables de dialecto, ocupación, educación, edad y sexo para las muestras de Bogotá y Miami. Incluimos la variable dialectal, ya que algunos estudios previos sobre la influencia del contacto interlingüístico e interdialectal han encontrado influencia del dialecto de origen en la expresión de los pronombres se sujeto (Otheguy/Zentella 2007: 283) y de la selección del impersonal (DeMello 2000: 365). De igual manera, exploramos cómo los usos pronominales en cada ciudad están condicionados por factores socioeconómicos (variables de ocupación y nivel educativo), la edad y sexo del hablante. 2.3.1. Variable dialectal Nuestra división se fundamenta en la clasificación dialectal que Montes Giraldo (1982: 23-27; 1996) propone con base en el criterio de la realización alofónica de la /-s/ implosiva y el empleo de los pronombres tú, vos, usted y su merced. Por consiguiente, incluimos un superdialecto costeño y un superdialecto andino. En la muestra, tanto de Bogotá como de Miami, no contamos con participantes de la variedad costeña pacífica, sólo de la atlántica. En esta zona predomina el tuteo, uso que se ha extendido desde la costa hacia el interior y, en los últimos años, diastráticamente de arriba abajo, desde las clases dominantes.

MURIEL/OROZCO

31/5/12

14:24

148

Página 148

LUZ MARCELA HURTADO CUBILLOS

En la variedad del interior, el superdialecto andino se caracteriza por el predominio de usted tanto en situaciones formales como informales. Nuestra clasificación comprende la subdivisión occidental y oriental, ya que los pronombres de tratamiento varían en las dos zonas: en la andina occidental (Antioquia, Caldas, Quindío, Risaralda y partes de Tolima y del Valle del Cauca), vos se utiliza en todas las clases sociales como índice de confianza, pero alterna con usted en ciertos contextos y situaciones (Montes Giraldo 1985: 242-249). Según Hurtado Idárraga (1994), en las relaciones familiares prevalece usted, aunque se favorece vos entre hermanos y tú con la familia política y los hijos. En la zona andina oriental, usted es la forma preferida tanto en la expresión de poder como de solidaridad, en especial en Bogotá (Uber 1985); sin embargo, en algunas situaciones de confianza también se utiliza tú. En cuanto a los usos de los impersonales, los estudios del español de Colombia tienen pocas referencias. En las muestras recogidas entre 1958 y 1959 para el Atlas Lingüístico-Etnográfico de Colombia, Montes Giraldo registró la utilización impersonal de tú en oraciones como: “El caimán como […] y que tú no le hagas ruido, se deja rascar” (1985: 144), en el departamento de Bolívar, zona que adoptó el tuteo desde la época colonial por intensos contactos económicos y administrativos con España.

2.3.2. Variables sociales La organización de las variables de ocupación, nivel educativo (sin escolaridad; primaria, secundaria, técnico, universitario) y edad (18 a 34; 35 a 54; 55+) se apoya en la clasificación de Labov (2001: 61) y del proyecto Pressea de la Asociación de Lingüística y Filología de América Latina (ALFAL). Las variables de ocupación y nivel educativo son nuestros indicadores de clase social. Es importante considerar estas dos variables conjuntamente tanto en la muestra de Bogotá como en Miami, ya que en ocasiones el nivel educativo alcanzado por los participantes no está necesariamente ligado a la profesión u oficio que desempeñan. En cuanto a la variable de ocupación, la clasificación de las profesiones y oficios revela el grado de independencia y poder económico. Esta distribución también intenta ubicar a los participantes en cuanto a las relaciones de poder que ejercen o a las que están sometidos; los de mano de obra por lo general están subordinados, y los tres grupos de profesionales están clasificados de acuerdo al grado de autoridad que tienen o ejercen sobre otros: a) Mano de obra no especializada: vendedores ambulantes, obreros no especializados, servicio doméstico, hogar y servicios no especializados.

MURIEL/OROZCO

31/5/12

14:24

Página 149

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

149

b) Mano de obra especializada: pequeños comerciantes, secretarios y oficinistas, obreros especializados, artesanos, mecánicos, vendedores en tiendas, cobradores, ayudantes técnicos, policías, guardias y soldados. c) Estudiantes: individuos que cursaban los últimos años de bachillerato o estudios universitarios de pre-grado. d) Profesionales y/o individuos de mando bajo: profesionales universitarios, docentes de primaria y media, pequeños empresarios y productores, mandos intermedios, técnicos y supervisores. e) Profesionales y/o individuos de mando medio: universitarios de libre ejercicio, gerentes medios, militares con graduación, medianos empresarios y productores, docentes universitarios. f) Altos funcionarios del poder ejecutivo, legislativo y judicial, altos oficiales del ejército, grandes empresarios privados, grandes hacendados, altos ejecutivos.

2.4. ÍNDICES DE LA INFLUENCIA DE LA SITUACIÓN DE CONTACTO EN MIAMI Con el fin de analizar los efectos del contacto interlingüístico e interdialectal, la muestra de Miami también fue codificada según las variables que aportan información sobre el nivel de conocimiento del inglés, el tiempo de contacto con esta lengua y la edad de llegada del hablante (grado de bilingüismo). La variable de tiempo de residencia proporciona datos sobre el efecto de la duración del contacto en la ciudad de Miami. El tiempo de estadía de nuestros participantes fluctuaba entre 6 y 42 años, por lo que dividimos esta variable así: 6-10 años; 11-20; 2130; 31+. Entre los colombianos de Miami, al igual que en otros grupos de inmigrantes, se evidencian diversos grados de conocimiento y uso del inglés: encontramos desde los individuos completamente monolingües hasta los bilingües simétricos. Por eso, con base en los criterios de Morales (1995: 148) y la escala de bilingüismo de Palacios (2005: 87-88), la clasificación del grado de bilingüismo incluye la forma de aprendizaje del inglés (formal o informal), su nivel de instrucción en esta lengua y el ámbito bilingüe o monolingüe en el que se desenvuelven los participantes: a) Bilingüe simétrico: individuos con competencia completa en ambas lenguas, ya que las aprendieron desde su infancia. Realizaron estudios de primaria y/o secundaria en inglés en EEUU. Este grupo está conformado por los que nacieron o llegaron antes de los 13 años a Miami, y cuyas redes sociales de interacción son predominantemente anglohablantes.

MURIEL/OROZCO

150

31/5/12

14:24

Página 150

LUZ MARCELA HURTADO CUBILLOS

b) Bilingüe consecutivo: individuos con dominio del inglés, pero que por haberlo aprendido en su juventud o en una edad adulta no adquirieron una pronunciación más nativa. En nuestra muestra, la mayoría de los participantes llegó después de los 13 años de edad y algunos de ellos realizaron estudios de pregrado en inglés. Aunque sus redes sociales son en su mayoría hispanohablantes, también se relacionan con algunos anglohablantes. c) Bilingüe incipiente: individuos que aprendieron inglés en su edad adulta, en centros comunitarios o de manera informal. Su competencia del inglés se limita a las demandas de su trabajo. En general, llegaron a Miami después de los 20 años de edad y sus redes sociales de interacción son hispanas. d) Monolingües: declaran no hablar, leer, escribir, ni entender inglés. Se trata de amas de casa o individuos cuyo trabajo no les ha exigido aprender inglés, y cuyas redes sociales son únicamente hispanohablantes. Los miembros de este grupo llegaron a Miami en una edad adulta.

3. Resultados y análisis Una vez codificadas, las 5,369 oraciones se sometieron al programa de análisis estadístico de regla variable Varbrul. Este programa proporciona evidencia en tres niveles: la relevancia estadística del efecto (significance), qué factores son o no son significativos al nivel .05; el alcance del efecto (range), qué variable es más o menos determinante; la jerarquía de la influencia, determinada por el orden de los pesos Varbrul dentro de cada grupo de variables (Tagliamonte 2006: 235). El programa provee los pesos de relevancia para cada factor de las variables independientes. Cuanto más influyente el factor, más cercano estará el número a .99. Los porcentajes de la última columna de la tabla 2 indican que, tanto en Miami como en Bogotá, predominan los pronombres impersonales uno (52%) y se (33%), y se utilizan menos tú (11%) y usted (4%). Estos primeros resultados van en la misma dirección que los porcentajes de DeMello (2000: 365), ya que tanto las variedades tuteadoras como ustedeadoras preferían uno (40% y 64%) y usaban en menor proporción se (12% y 20%). También observó una taza pronominal de tú impersonal del 47% en los participantes tuteadores y en los ustedeadores, una taza pronominal de usted impersonal del 14% (excepto en Sevilla). Los pesos Varbrul de la tabla 2 indican que el lugar de residencia tiene efecto en el uso de uno y tú. Los residentes en Miami favorecen el tú impersonal (.84) y los de Bogotá, uno (.57). En el caso de usted y se, ya que los pesos están cerca de .50, podemos apreciar una falta de efecto en las dos ciudades. En general, los

MURIEL/OROZCO

31/5/12

14:24

Página 151

151

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

TABLA 2 Porcentajes y pesos Varbrul del uso de pronombres impersonales con relación al lugar de residencia del hablante Miami

Bogotá

Total

N

%

Peso

N

%

Peso

N

%

Uno

415

15

.43

2349

85

.57

2764

52%

Tú

334

55

.84

276

45

.16

610

11%

65

32

.50

141

68

.50

206

4%

273

15

.47

1516

85

.53

1789

33%

1087

20

4282

80

5369

100%

Usted se Total

Relevancia (significance): .0001 Números en negrita: pesos más influyentes

colombianos de Miami y Bogotá utilizaron el se impersonal en un 67% de los casos para referirse a un sujeto general, en un 12% para referirse a sí mismos y en el 21% a un nosotros; uno, se empleó más para incluir al hablante (70%). Los resultados de se y uno confirman la idea de los grados de impersonalización que se alcanzan con cada pronombre, grados que mencionamos en la introducción. Al contrario que en los estudios anteriores prevaleció el uso de tú y usted para referir a un sujeto general (60% y 71% de los casos, respectivamente). Hasta aquí, confirmaríamos que aunque existe un predominio de uno en la expresión de impersonalidad en las dos ciudades, tú es el segundo impersonal más usado en Miami, compitiendo e incluso desplazando a se. Sin embargo, todavía es necesario indagar si es el factor dialectal o el de contacto con el inglés lo que determina este aumento. Luego de dividir la muestra conforme a los lugares de residencia, realizamos cómputos separados para cada uno, lo cual facilitó tanto el análisis y comparación simultánea de todas las variables, como la delimitación de su grado de influencia en cada variante de impersonalidad. Para determinar la relevancia de las variables al nivel .05, el alcance de su efecto y la jerarquía de la influencia de los factores de cada variable (pesos Varbrul), también se pasó el programa a cada pronombre impersonal por separado. En la medida que los resultados de Bogotá serán el punto de referencia para determinar el posible efecto de la situación de contacto en Miami, presentaremos primero estos resultados. Con el fin de incluir los diferentes cálculos, los resultados están organizados en las tablas 3, 4, 6 y 7 de la siguiente manera:

MURIEL/OROZCO

152

31/5/12

14:24

Página 152

LUZ MARCELA HURTADO CUBILLOS

1. Cada columna representa un cómputo independiente para el pronombre impersonal que la encabeza. Por eso, el número de relevancia (significance) de cada uno se encuentra al final de la columna. 2. Se incluye el alcance del efecto de cada variable independiente (diferencia numérica entre el peso Varbrul más bajo y el más alto). También se especifica qué variable fue eliminada por el programa estadístico. Con base en el alcance, organizamos los resultados en dos tablas para Bogotá y tres para Miami, y situamos en la primera tabla los grupos de variables que tuvieron más efecto (ocupación y educación). 3. El porcentaje de ocurrencias del impersonal para cada factor de una variable debe leerse de manera horizontal. Por el contrario, los pesos Varbrul que le siguen (un número entre .00 y .99) deben interpretarse verticalmente, para observar la jerarquía de la influencia de cada factor: recordemos que cuanto más influyente el factor, más cerca está el número de .99.

3.1. LA EXPRESIÓN DE IMPERSONALIDAD EN BOGOTÁ, COLOMBIA El porcentaje total de cada pronombre con respecto al total de la muestra de Bogotá revela que en esta ciudad predominan uno y se como formas generalizadoras (54.9% y 35.4%) y escasea el empleo de los impersonales tú (6.4%) y usted (3.3%). Según los resultados de las tablas 3 y 5, las variables de dialecto, edad, sexo, ocupación y nivel educativo son estadísticamente significativas (al nivel .05) en la expresión de impersonalidad con uno, se y tú. Sin embargo, el programa eliminó las variables de edad y sexo en el cómputo de usted, por no tener efecto en la escogencia de esta forma. Observamos que las variables de ocupación y nivel educativo ejercieron mayor influjo en uno, se y usted que las variables de dialecto, edad y sexo (ver alcance). En el caso de tú, las variables de educación y dialecto fueron las más determinantes (alcances: 82 y 62). Según la jerarquía de la influencia (orden de los pesos Varbrul dentro de cada grupo de variables), observamos que el impersonal tú fue favorecido considerablemente por los individuos de mando medio (.86) y aquellos que habían alcanzado un nivel educativo universitario (.92); uno por los de mano de obra no especializada (.76); se por los estudiantes universitarios (.64) y por los de mando bajo (.67) y, también, por los de menor nivel de escolaridad (primaria: .60). En el caso de usted los pesos Varbrul no proporcionan una clara dirección de la influencia, ya que, sumados a los de mano de obra (.74) y a los altos funcionarios (.81), los que no tenían escolarización (.76) y los que cursaron estudios técnicos (.72) y universitarios (.69) favorecieron este pronombre. Esta falta de claridad en el efecto puede estar ligada a que sólo se registró un 3.3% de usted en el total de la muestra.

MURIEL/OROZCO

31/5/12

14:24

Página 153

153

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

TABLA 3 Análisis de regla variable (Varbrul) sobre la influencia de la ocupación y nivel educativo en la selección de los pronombres impersonales. Muestra de Bogotá, Colombia tú

Ocupación Mano no espec. Mano espec. Estud. Mando bajo Mando medio Altos funci. Alcance (range) Nivel educativo alcanzado Sin escolar. Primaria Secundaria Esc. técnica Universidad Alcance (range) Porcentaje del total de la muestra de Bogotá Significance:

uno

se

Ud.

%

Peso

%

Peso

%

Peso

%

Peso

.2 1.9 3.6 10.5 14.5 —

.32 .45 .25 .44 .86

74.9 55.7 51.2 43.9 49.4 46.3

.76 .49 .42 .36 .44 .48 [40]

20.1 37.5 42.4 44.0 33.7 36.6

.25 .50 .64 .67 .48 .35 [42]

4.7 4.8 2.8 1.6 1.7 17.1

.74 .74 .48 .25 .27 .81 [56]

36.8 68.2 64.0 46.3 47.7

.08 .47 .52 .49 .52 [47]

24.6 29.2 33.7 38.8 38.2

.58 .60 .57 .40 .47 [20]

38.6 2.4 2.1 6.8 2.4

.76 .27 .20 .72 .69 [56]

[54]

— .2 .2 8.1 11.7

.32 .10 .69 .92 [82]

6.4

54.9 .0001

35.4 .005

3.3 .002

.033

Para evidenciar mejor la relación de la variable de ocupación con la de nivel de educación presentamos una tabulación cruzada (Tabla 4). En cuanto al predominio de uno en los niveles de ocupación más bajos, los porcentajes de la tabla 4 muestran que los usos de uno y usted compiten (37% y 38%) en los individuos de mano de obra no especializada que no tienen nivel de escolaridad. Esclarecen también que uno prevalece entre los de nivel de primaria y secundaria (74% y 78%). En un nivel de ocupación más alto, de trabajadores de mano de obra especializada con nivel de primaria, se favorece tanto uno como se y, a medida que aumenta el nivel educativo (de secundaria a técnico) domina uno y disminuye se. Por último, los que cursaron estudios universitarios son los que utilizan considerablemente el se impersonal (67%).

MURIEL/OROZCO

31/5/12

14:24

154

Página 154

LUZ MARCELA HURTADO CUBILLOS

TABLA 4 Porcentajes de uso de los pronombres impersonales con respecto a la variable de ocupación y el nivel de educación. Tabla cruzada Sin esc. Mano no espec. Tú Uno Se Ud. Mano espec. Tú Uno Se Ud. Estud. Tú Uno Se Ud. Mando bajo Tú Uno Se Ud.

57

Primaria 575

Secundaria

Universidad

222 .3 74 23 2

37 25 38

Técnico

190

86 11 3 600

198 .3 57 40 2

52 45 3

3 9 57 20 14

170

33 67

365 5 47 44 4

60 40

286

714 8 38 52 2

12 46 41 1

Mando medio Tú Uno Se Ud.

861

Altos funci. Tú Uno Se Ud.

41

15 49 34 2

46 37 17

MURIEL/OROZCO

31/5/12

14:24

Página 155

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

155

El aumento de se (40%) ocurre en los individuos de mano de obra especializada con un nivel educativo técnico. Al parecer, los técnicos que no se desempeñan en una labor acorde a su nivel de educación comienzan a promover este uso; es el mismo caso de aquellos con estudios universitarios pero que no han tenido oportunidades laborales en su carrera (52%). Es posible que estos dos grupos intenten marcar su falta de pertenencia a la ocupación que desempeñan con el empleo considerable de se. Ya que entre los estudiantes universitarios tanto uno como se tienen porcentajes similares, y esta tendencia se mantiene entre los diversos grupos de profesionales y de diferente grado de mando, es probable que el empleo de se impersonal esté asociado con un habla más formal y/o educada. Este aumento en el uso de la variante que perciben de prestigio se explica con el concepto de “movilidad social” (Labov 2001: 67). El grupo de individuos de educación técnica que ha alcanzado una ocupación con un grado de poder ha experimentado un ascenso social; el grupo con estudios universitarios que desempeña una ocupación de mano de obra ha sufrido un descenso social. Es interesante resaltar que cuando los de nivel técnico hablaban de temas personales, obtuvieron la tasa pronominal de se más alta en relación con el resto de los grupos de ocupación (42%). Puesto que es posible ubicar estos dos grupos en las clases sociales trabajadoraalta y media-baja, podemos afirmar que este sector de la sociedad está liderando el aumento de se. Asimismo, el impersonal tú comienza a ser utilizado por los de nivel de estudios técnicos de mano de obra (9%), de mandos bajos (8%), y su porcentaje de uso se incrementa en los profesionales de mando bajo (12%) y medio (15%). Sin embargo, no predomina tanto este empleo de tú en los grupos que experimentan menos presión social en cuanto a las relaciones de poder (los estudiantes: 5%) y los que ejercen poder sobre los demás grupos (los de alto mando: 0%). Esta tendencia de las clases sociales trabajadora-alta y media-baja también se detecta también en los porcentajes de usted de los individuos de educación técnica dedicados a trabajos de mano de obra especializada (14%). Como hemos observado, los de nivel técnico es un grupo innovador, que posiblemente utiliza las formas asociadas a los grupos con más prestigio. Sin embargo, los porcentajes de utilización de usted impersonal por parte de los altos funcionarios (17%) y el gran porcentaje en los de mano de obra no especializada (38%) encuentran explicación en la variable dialectal: el 100% de los casos de usted impersonal de los altos funcionarios pertenece a hablantes de la variedad andina oriental, es decir, el grupo con mayor uso del pronombre de sujeto usted. No obstante, el 100% de las ocurrencias de usted en los trabajadores de mano de obra no especializada y sin escolaridad corresponde a hablantes de la variedad andina occidental. Este dato nos conduce a los resultados de las variables sociales de dialecto, edad y sexo, tabla 5.

MURIEL/OROZCO

31/5/12

14:24

156

Página 156

LUZ MARCELA HURTADO CUBILLOS

TABLA 5 Análisis de regla variable (Varbrul) de la influencia de las variables de dialecto, edad y sexo en la selección de los pronombres impersonales. Muestra de Bogotá, Colombia tú

uno

se

Ud.

%

Peso

%

Peso

%

Peso

%

Peso

2.4 .3 14.5

.32 .21 .83 [62]

54.7 63.8 52.8

.47 .62 .51 [15]

39.7 25.8 30.9

.58 .39 .40 [19]

3.2 10.1 1.7

.50 .78 .42 [36]

Edad 18-34 35-54 55+ Alcance

6.3 5.2 8.4

.71 .29 .40 [42]

58.8 50.2 53.9

.51 .46 .53 [7]

32.1 40.9 34.2

.44 .56 .52 [12]

2.9 3.7 3.5 [Eliminado]

Sexo Femenino Masculino Alcance

6.6 6.3

.58 .42 [16]

59 50.8

.53 .47 [6]

31.7 39.0

.43 .56 [13]

2.7 3.8 [Eliminado]

Dialecto A. Oriental A. Occidental Costeño Alcance

Porcentaje total de la muestra de Bogotá Significance:

6.4

54.9 .0001

35.4 .005

3.3 .002

.033

En los resultados del cómputo que se realizó para el tú impersonal (primera columna), el alcance del efecto indica que de estas tres variables, la variable dialecto ejerce la influencia más fuerte en la escogencia de todos los impersonales (alcance: 62, 15, 19 y 36). Según la jerarquía de la influencia (pesos Varbrul), observamos que los hablantes del dialecto costeño favorecen tú (peso .83), del andino oriental se (.58) y del andino occidental uno (.62) y usted (.78). Así, confirmamos que los andinos occidentales lideran el uso impersonal de usted: uno de los grupos más conservadores del país que, a pesar de encontrarse fuera de su contexto de origen, mantiene las características de su variedad dialectal. Al comparar nuestros resultados con el estudio de los pronombres de sujeto en el dialecto occidental realizado por Hurtado Idárraga (1994), observamos

MURIEL/OROZCO

31/5/12

14:24

Página 157

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

157

algunas similitudes y diferencias con los usos pronominales de la zona occidental: en Medellín, usted predomina en la clase media-alta, tú en la media-media y vos en la media baja; en Manizales, vos casi no se usa, mientras que usted prevalece en todos los niveles sin importar el grado de confianza de los hablantes. En nuestro estudio, los individuos de la variedad occidental de estrato y nivel educativo más bajo favorecen el usted impersonal, tal vez debido a la inconveniencia de usar el vos en una ciudad no voseante como Bogotá. En cuanto a las variables de edad y sexo, éstas tienen poco efecto en el empleo de uno (pesos Varbrul muy cercanos a .5) y ningún efecto en usted (eliminadas). Sin embargo, se predomina en los hablantes cuyas edades oscilan entre 35 y 54 años (.56) y los hombres (.56), y tú, tanto en el grupo de 18 a 34 años (.71) como en las mujeres (.58). Estos resultados concuerdan con Hurtado Idárraga (1994), pues en Medellín las mujeres utilizan más los pronombres personales tú y usted, y los hombres, vos. Además, la creciente tendencia al uso del pronombre de sujeto tú entre los más jóvenes está contribuyendo al declive del voseo. En la zona oriental, las preferencias por los pronombres personales también están vinculadas con distinciones de sexo; los hombres utilizan usted tanto en situaciones formales como de confianza, y las mujeres, para marcar distancia. Ellas también emplean tú como forma de confianza y solidaridad (Bartens 2004). En situaciones de confianza, a la vez que usted predomina entre los hombres, los inmigrantes y los jóvenes, compite con tú entre los hablantes de nivel educativo superior. En situaciones formales, usted prevalece entre los hombres, los jóvenes y los de educación superior, mientras que tú entre los mayores y los de nivel educativo más bajo (Montes Giraldo et al. 1998: 131-132). Rey Castillo (2005: 122123) detectó usos similares en Funza, Cundinamarca, es decir, preferir usted en situaciones de respeto y cercanía, y tú para el trato entre mujeres, jóvenes y la clase alta. En la región del Tolima, Rey (1994a) encontró relación con la clase social y el grado de educación de los hablantes: menor uso de usted si los ingresos económicos eran mayores y se interactuaba con compañeros de trabajo; mayor uso de tú si los hablantes de clase alta se dirigían a un hombre joven de rango superior. En síntesis, los resultados de Bogotá demuestran, en primer lugar, que la ocupación del hablante vinculada al nivel de educación son índices de cómo la case social, en especial el factor de movilidad social, es el más influyente en la elección de los pronombres impersonales. En segundo lugar, revelan un fuerte efecto de la variedad dialectal: aunque los impersonales que más se utilizan en Bogotá son uno y se, los hablantes de origen costeño tienden al tú impersonal y sólo los andinos occidentales al usted impersonal. Este aspecto corresponde a los usos pronominales propios de cada región, ya que, como se explicó, los costeños son

MURIEL/OROZCO

158

31/5/12

14:24

Página 158

LUZ MARCELA HURTADO CUBILLOS

predominantemente tuteadores y los andinos occidentales, ustedeadores. Dado que en esta ciudad predominan los hablantes de variedad andina oriental y esta variedad también utiliza tú en situaciones de confianza y condescendencia, el tercer efecto en el uso impersonal de tú lo ejercen más dos grupos sociales: los jóvenes y, con menos fuerza, las mujeres. Estos dos grupos sociales están impulsando esta forma en Bogotá gracias a que la variedad andina oriental no es exclusivamente ustedeadora.

3.2. EFECTOS DEL CONTACTO INTERLINGÜÍSTICO E INTERDIALECTAL EN MIAMI, EEUU De acuerdo con los porcentajes totales de la tabla 6 y en relación con los de Bogotá, observamos que en Miami, los porcentajes de uno (38.2%) y se (25.1%) disminuyen, tú (30.7%) aumenta y usted (6%) se mantiene bajo. La tasa pronominal de tú corrobora lo evidenciado en la tabla 2, esto es, un porcentaje de ocurrencia más alto que el de Bogotá (3.3%). Como se aprecia, no hay una simplificación del sistema pronominal impersonal sino nivelación de los porcentajes. Este equilibrio sugiere una neutralización de los usos que probablemente resultan marcados (Palacios 2007b: 112) en el contexto de Miami, como el sobreuso de uno. Con base en los resultados de las tablas 6 y 9 se aprecia que la única variable significativa para todos los impersonales fue la de ocupación: en Miami, continua siendo la variable de mayor alcance para se y usted (63 y 76), y la segunda para tú y uno (63 y 38). Por el contrario, la variable de educación no tuvo la misma importancia que en Bogotá, así que fue eliminada por el programa por no tener efecto en usted y uno. El hecho que sea la quinta variable en importancia para tú plantea una gran diferencia, pues el nivel educativo era la variable con mayor efecto para el impersonal tú en Bogotá: los que alcanzaron un nivel educativo técnico y universitario favorecían más este uso. De acuerdo a la jerarquía de la influencia que nos proporcionan los pesos Varbrul, los resultados de tú, uno y se difieren con los de Bogotá: los dos grupos de ocupación de mano de obra favorecen considerablemente el uso impersonal de tú (.79 y .78), los de mando bajo prefieren uno (.74) y los de mando medio, se (.74). En cuanto a usted este pronombre sigue predominando entre los de mano de obra (.96 y .74). Es necesario analizar la tabla cruzada 7 con el fin de determinar si existe la tendencia que habíamos observado en Bogotá, es decir, un desplazamiento de uno a se entre los de nivel educativo técnico y universitario que se desempeñaban en oficios de mano de obra especializada. Esta tabla nos permitirá observar de manera más clara la relación entre las variables de ocupación y nivel de educación.

MURIEL/OROZCO

31/5/12

14:24

Página 159

159

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

TABLA 6 Análisis de regla variable (Varbrul) de la influencia de las variables de ocupación y nivel educativo en la selección de los pronombres impersonales. Muestra de Miami, EEUU tú

Ocupación Mano no espc. Mano espec. Estud. Mando bajo Mando medio Alcance Nivel educativo alcanzado Primaria Secundaria Esc. técnica Universidad Alcance (range) Porcentaje del total de la muestra de Miami Significance:

uno

se

Ud.

%

Peso

%

Peso

%

Peso

%

Peso

25.5 28.6 27.9 34.7 35.2

.79 .78 .33 .16 .51 [63]

52.9 36.1 49.0 43.4 21.2

.61 .50 .44 .74 .36 [38]

9.8 25.0 22.1 13.3 40.7

.13 .23 .76 .37 .74 [63]

11.8 10.2 1.0 8.7 2.9

.96 .74 .20 .68 .26 [76]

— 26.1 30.6 32.8

58.3 .26 43.8 .27 29.4 .64 38.0 [38] [Eliminado]

41.7 20.8 28.1 25.5

.85 .71 .74 .36 [49]

30.7

38.2 .008

25.1 .0001

— 9.3 11.9 3.6 [Eliminado]

6.0 .002

.007

Una diferencia importante en relación con los porcentajes de Bogotá es que si bien predomina todavía uno, tú aparece desde los individuos de nivel de secundaria dedicados a labores de mano de obra no especializada y especializada (27% y 26%). No obstante, el porcentaje de uso de tú es más alto entre los técnicos dedicados a labores de mano de obra no especializada (34%) y los de nivel universitario de mando bajo (38%) y medio (35%). Este aumento de tú va en detrimento de se y uno en los de nivel técnico de mano de obra especializada, de uno en los universitarios de mando medio y de se en los universitarios de mando bajo. Como en Bogotá, los de nivel técnico que trabajan en oficios que requieren menos educación, aquellos que experimentaron un descenso social, son los propiciadores de los cambios al adoptar las variantes de las clases de mayor poder

MURIEL/OROZCO

31/5/12

14:24

160

Página 160

LUZ MARCELA HURTADO CUBILLOS

TABLA 7 Porcentajes de uso de los pronombres impersonales con respecto a la variable de ocupación y el nivel de educación. Tabla cruzada, Miami Primaria

Secundaria

Mano no espec. Tú Uno Se Ud.

2

49

Mano espec. Tú Uno Se Ud.

10

177

145 26 41 25 8

34 28 25 13

Estud. Tú Uno Se Ud. Mando bajo Tú Uno Se Ud. Mando medio Tú Uno Se Ud.

Universidad

27 55 6 12

100

70 30

Técnico

290 28 49 22 1 15

158 38 44 9 9

40 60 241

35 21 41 3

económico. Sin embargo, favorecen tú, un impersonal distinto al que favorecían en Bogotá. De igual manera, son los colombianos de posición socioeconómica más alta (mando medio) los que disminuyen considerablemente el uso de uno y en los que aumenta tú. Este aumento de tú entre los grupos más educados y de mandos bajos y medios también ocurrió en Bogotá, aunque en menor proporción. El segundo grupo de variables que contribuyen a explicar el aumento del impersonal tú son las variables de dialecto (alcance: 40) y edad (alcance: 69). En

MURIEL/OROZCO

31/5/12

14:24

Página 161

161

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

relación con otros impersonales, tanto estas variables como la de sexo fueron eliminadas por carecer de efecto en el empleo de ciertos impersonales: dialecto en el cómputo de uno, edad en el cómputo de se y sexo en el cómputo de tú y se. TABLA 8 Análisis de regla variable (Varbrul) de la influencia de las variables de dialecto, edad y sexo en la selección de los pronombres impersonales. Muestra de Miami, EEUU tú

uno

%

Peso

Dialecto A. Oriental A. Occidental Costeño Alcance

40.1 13.4 42.8

.68 41.3 .28 38.0 .64 37.3 [40] [Eliminado]

Edad 18-34 35-54 55+ Alcance

30.9 37.6 3.3

.76 .50 .07 [69]

Sexo Femenino Masculino Alcance Porcentaje del total la muestra de Miami Significance:

33.8 26.9 [Eliminado]

30.7

%

Peso

Ud.

%

Peso

%

Peso

18.0 34.5 19.3

.43 .63 .41 [22]

.6 14.1 .6

.15 .80 .35 [65]

48.7 32.7 36.4

.41 .47 .73 [36]

19.0 22.0 51.0 [Eliminado]

1.5 7.7 9.3

.32 .54 .73 [41]

44.6 30.4

.57 .42 [15]

19.1 32.4 [Eliminado]

2.5 10.2

.29 .75 [46]

38.2 .008

se

25.1 .0001

6 .002

.007

En cuanto a la variable dialectal, observamos que la variedad costeña nuevamente lidera el empleo de tú (.68) seguida por los andinos orientales (.64), y la variedad andina occidental favorece se y usted (.63 y .80). Puesto que la variable de dialecto es la segunda en importancia en cuanto al empleo de usted en los andinos occidentales, es posible que los usos pronominales de su variedad de español continúen influyendo el uso de los impersonales. Para el caso de uno,

MURIEL/OROZCO

162

31/5/12

14:24

Página 162

LUZ MARCELA HURTADO CUBILLOS

debido a que el programa Varbrul eliminó la variable de dialecto y ésta obtuvo uno de los alcances más bajos en Bogotá, es factible concluir que no influye en la selección de uno ni en Bogotá ni en Miami. De igual importancia es que las diferencias dialectales de los andinos orientales frente a los costeños, que tenían relevancia en Bogotá, desaparecen con los usos impersonales de tú y se. Esto nos permitiría hablar de un primer signo de nivelación dialectal en Miami, nivelación de los andinos orientales, variedad ustedeadora y tuteadora; ya que los jóvenes y las mujeres tutean con frecuencia, posibilitaría una relación de esta variable con las de edad y sexo. Al contrario, la variable de sexo no tuvo efecto en la escogencia de tú y se y obtuvo un alcance bajo para uno. Usted fue el único impersonal en que la variable de sexo fue determinante: mayor efecto en los hombres (.75) En el cómputo que se realizó para tú, el alcance indica que la edad ejerce la influencia más alta (alcance: 69). Como en Bogotá, se advierte que el grupo de 18 a 34 años de edad favorece el impersonal tú (.76), mientras que los mayores de 55 años lo desfavorecen de manera sustancial (.07), siendo este último el grupo con mayor efecto en uno y usted (.73 y .73). En cuanto al grupo más joven (18-34 años), es posible que generalice el uso de tú impersonal por influencia del inglés. Su bilingüismo los convierte en agentes del cambio (Thomason/Kaufman 1988). Morales (1995: 155) ya había demostrado la influencia de las variables de edad y grado de bilingüismo en los usos impersonales de los puertorriqueños residentes en la isla: tanto los jóvenes (17-25 años) como los que habían estudiado su primaria, secundaria y/o universidad en Estados Unidos, desfavorecieron el uso de se y utilizaron tú y uno (este último con el porcentaje más alto). Esta posibilidad también fue planteada en cuanto al empleo de los pronombres personales en el estudio de Jaramillo (1995) realizado en Nuevo México (sólo los individuos de 17 a 30 años utilizaban tú en todos los contextos). Es factible que los mayores sigan empleando las formas que usaban en Colombia antes de emigrar, y que los jóvenes estén trasplantando los usos pronominales del inglés, su lengua de interacción diaria y de estudios. Por esto, los resultados de la tabla 9 son fundamentales para determinar la influencia de la situación de contacto con el inglés y en el contexto de Miami. Presentamos tanto la variable que combina el grado de bilingüismo, la edad de llegada y las interacciones de los individuos, como la del tiempo de estadía en Miami. El grado de bilingüismo es para el caso de tú y usted la variable con el alcance en tercer grado de importancia (50 y 63 respectivamente), para uno, la primera (42) y para se, la cuarta (23). Los pesos bajos de tú, tanto de los monolingües como de los bilingües simétricos (.49 y .27), contrastan con el peso alto de los bilingües consecutivos y los insipientes (.66 y .77). Es decir, la utilización de tú impersonal se vincula a un menor grado de bilingüismo, de hablantes que lo aprendieron durante su juventud o en la edad adulta, y cuyas redes sociales son

MURIEL/OROZCO

31/5/12

14:24

Página 163

163

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

TABLA 9 Análisis de regla variable (Varbrul) de la influencia del contexto interlingüístico e interdialectal de Miami en la selección de los pronombres impersonales tú % Grado de bilingüismo y edad de llegada Simétrico Consecut. Insipiente Monolin. Alcance Tiempo en Miami 6-10 11-20 21-30 31+ Alcance Porcentaje del total la muestra de Miami Significance:

uno

se

Ud.

Peso

%

Peso

%

Peso

27 33 50 15

.27 .66 .77 .49 [50]

48 26 31 53

.72 .35 .30 .39 [42]

19 34 16 21

.37 .60 .57 .55 [23]

39 29 36 6

.39 .43 .73 .42 [34]

32 46 36 24

.59 .61 .31 .35 [30]

26 18 21 59

.48 .41 .59 .68 [27]

30.7

38.2

.008

25.1

.0001

%

5 7 2 11

Peso

.72 .43 .09 .34 [63]

2 6 7 10 Eliminado 6

.002

.007

principalmente hispanohablantes. También observamos que un alto grado de conocimiento del inglés e interacción con anglohablantes no es determinante en el uso de tú. Estos resultados se acogerían a la idea de Palacios (2005: 88-89) de que los rasgos que se alejan de la norma estarían más presentes en los grupos bilingües insipientes y consecutivos que en los simétricos y los monolingües. En relación con la variable de tiempo en Miami, observamos que los que llevan de 21 a 30 años favorecen el uso de tú, es decir, el segundo grupo con más tiempo de contacto. Según nuestros datos, todos son bilingües insipientes y llegaron a EEUU después de los 20 años de edad. Deducimos que a pesar de que este grupo ha permanecido en Miami por un tiempo considerable, la intensidad del contacto con el inglés no ha sido fuerte. Esto aminora la posibilidad del efecto del contacto de lenguas en el empleo de tú. Por el contrario, se apunta a que es la intensa interacción hispanohablante la que está causando el aumento de tú. Es

MURIEL/OROZCO

164

31/5/12

14:24

Página 164

LUZ MARCELA HURTADO CUBILLOS

decir, es la cantidad de presión cultural que ejercen otros grupos hispanoamericanos sumada a factores sociales como la duración del periodo de contacto, el tamaño de cada grupo y el dominio socioeconómico (Thomason 2001: 66). Es interesante, sin embargo, que el grado de bilingüismo e interacción en inglés muestra más efecto en la utilización de uno (.72) y usted (.72) por parte de los bilingües simétricos. Este hecho apuntaría a la posibilidad de influencia del impersonal one del inglés y, tal vez, de la segunda persona singular. También se consideraría como un índice del dominio de las dos lenguas, de este grupo que nació en Miami o llegó antes de cumplir los 13 años de edad, y cuyas redes de interacción son predominantemente anglohablantes. No obstante, debido a que los casos de uno ocurrieron principalmente en el grupo de estudiantes (67%), consideramos que el factor decisivo es la falta de la presión social que los grupos de interacción hispanohablante ejercen sobre los otros. A pesar de que existe la posibilidad del efecto del bilingüismo, el alto peso de usted impersonal en los bilingües simétricos nos dirige hacia la influencia de la variedad dialectal. Esto resulta cierto tanto para uno como para usted, pues de las ocurrencias de uno en los bilingües simétricos, un 61% eran hablantes de la variedad andina oriental, un 30% de la andina occidental y el 9% de la costeña. En cuanto a los casos de usted, el 91% pertenecía a hablantes andinos occidentales. El que tanto los bilingües consecutivos y los insipientes como los que llevan más de 31 años en Miami hayan favorecido la variante se (.60 y .57; .68) siga cuestionando la idea de la influencia del contacto interlingüístico. Según nuestros datos, el 100% de las ocurrencias son de bilingües consecutivos y el 82% de los que llegaron a EEUU después de los 20 años. El 69% de estos bilingües consecutivos son profesionales de mando medio y bajo. Tienen un buen grado de manejo del inglés y lo utilizan en el trabajo. Es decir, son colombianos más establecidos y con mejor nivel socioeconómico y educativo, que llegaron a EEUU cuando las características de su primera lengua ya estaban plenamente establecidas. A pesar de llevar un periodo de tiempo más prolongado en Miami, este grupo posiblemente no sigue la tendencia hacia el aumento del uso de tú por no experimentar el mismo tipo de presión social a la que están sometidos los otros grupos: tienen solvencia socioeconómica. En resumen, los resultados de Miami revelan que el conocimiento e interacción en inglés no tiene influencia directa en el empleo de los pronombres impersonales. Por el contrario, se demuestra que son más fuertes los factores socioculturales, en especial la situación de contacto con la misma comunidad colombiana de otras regiones y con otros grupos hispanos. Con respecto a la muestra de Bogotá, existen diferencias tanto en el porcentaje de usos como en la dirección de la influencia: variables como la de educación, que mostraron un fuerte efecto en la muestra de Bogotá, no lo ejercieron de la misma manera en todos los impersonales; la diferen-

MURIEL/OROZCO

31/5/12

14:24

Página 165

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

165

ciación entre los dialectos se borra cuando los andinos orientales usan los impersonales tú, uno y se en la misma proporción y dirección que los costeños. Hay una tendencia general al aumento de tú y a la disminución de uno y se. La edad es la variable que influye fuertemente en este aumento (liderado por los del grupo de edad de 18-34 años), y la interacción entre las variables de educación y nivel educativo revela que aquellos que han experimentado descenso social favorecen más este impersonal. En cuanto a uno y se, se demostró que el factor más importante entre los que los utilizan es la falta de presión social, observable principalmente en la interacción de las variables de ocupación y grado de bilingüismo. Por último, usted sigue siendo la variante menos utilizada y la más determinada por el origen dialectal del hablante.

4. Conclusiones Con este estudio hemos realizado una evaluación de las variables sociales que afectan la expresión de la impersonalidad por parte de los colombianos de Miami y Bogotá. El objetivo era, primero, esclarecer si existían diferencias de uso en cada contexto y, segundo, determinar la procedencia de la influencia más fuerte: del contacto con el inglés o del contacto con otras variedades de español. Por eso, comenzamos el análisis con el corpus de Bogotá con el fin de obtener los patrones de uso de los impersonales que nos permitieran evidenciar los cambios en el contexto de Miami. Además de incluir las mismas variables de Bogotá (dialecto, ocupación, nivel educativo, edad, y sexo), en el análisis de Miami incorporamos dos variables sociales índices de la influencia del contacto interlingüístico y/o interdialectal (el tiempo de permanencia en Miami y el grado de bilingüismo, el cual abarcaba no sólo información sobre el conocimiento del inglés sino también la edad de llegada del hablante y sus interacciones con anglo/hispanohablantes). Los porcentajes totales de empleo de estos pronombres en Miami y en Bogotá ratifican que el pronombre impersonal más utilizado por los colombianos es uno (52% del total de las dos muestras, 54.9% en Bogotá y 38.2% en Miami). El análisis de las dos muestras reveló la complejidad del efecto y de las relaciones de las variables de ocupación, educación, variedad dialectal y edad. Esto indica que no hay un solo factor social que determine la expresión de la impersonalidad en Bogotá y en Miami, sino que es la interacción entre variables la que explica el favorecimiento de un impersonal sobre otro. De igual manera, los resultados de la muestra de Bogotá y Miami sustentan que los grupos que experimentan descenso y presión social son los que promueven o son más susceptibles a los cambios. Nuestros resultados sugieren que la dirección de la innovación en el uso de los impersonales proviene de la clase trabajadora-alta y media-baja.

MURIEL/OROZCO

166

31/5/12

14:24

Página 166

LUZ MARCELA HURTADO CUBILLOS

No comprobamos que el grado de conocimiento del inglés fuera el causante directo de los cambios en los porcentajes. Más bien, se trata de un cambio indirecto producido por el contacto interdialectal, en el que la frecuencia del impersonal tú aumenta, y uno y se disminuyen. Esta desviación de los usos que detectamos en el español hablado en Bogotá puede considerarse, de acuerdo con Palacios (2007b: 122), como un caso de convergencia lingüística en el que se desarrollan estructuras gramaticales comunes. En lugar de haber ocurrido una simplificación del sistema pronominal impersonal, se ampliaron las variantes con el incremento del impersonal tú. Esta variación en la frecuencia de los impersonales puede implicar a su vez la neutralización de usos marcados: en el caso de uno, su sobreuso asociado a los niveles de ocupación más bajo en Bogotá. Además de esta intensa situación de contacto con otras variedades de español en Miami, podemos sugerir que un segundo factor importante es el compartir características dialectales; tal vez por el hecho que los colombianos de la variedad costeña tienen afinidad lingüística con algunos grupos caribeños, tienden más al uso del tú impersonal que los de la variedad andina occidental. Asimismo, influye el que los andinos orientales sean hablantes usteadores y tuteadores, pues los abre a la innovación o adaptación de los rasgos del nuevo contexto, incluso a su nivelación con la variedad costeña, como lo observamos en los porcentajes de las variantes impersonales. En cambio, los usos pronominales de la variable andina occidental demuestran que este grupo sigue siendo el más conservador. Así, comprobamos que la similitud de las características dialectales funciona de la misma manera que la de las lenguas en situaciones de contacto intenso (Palacios 2005: 79). No obstante, Klee/Lynch (2009) plantean que por cuestión de las orientaciones políticas entre los cubanos, compartidas sólo por algunos grupos hispanos, se ha ocasionado una falta de solidaridad entre la comunidad hispana de Miami. En el examen que Alfaraz (2002) realiza sobre las percepciones de los cubanos de Miami en cuanto a otras variedades de español, se revelan datos importantes: el español de los colombianos está considerado como uno de los mejores, junto con el de los españoles, de los cubanos que llegaron antes de la Revolución y de los argentinos; el español caribeño y el centroamericano están menos valorados. Por lo tanto, es posible que el colombiano andino occidental comparta la misma percepción frente a otras variedades y, por ello, conserve los usos propios de su norma originaria en Colombia, en especial el impersonal usted, e incluso se resista a emplear elementos propios de la variedad que domina en materia económica y política (Hurtado 2002: 167). Todavía queda por resolver si los andinos orientales y los costeños se están acercando a la norma del grupo mayoritario, o a una norma Caribe, o panhispánica. Infortunadamente, aún no hay estudios sobre el sistema pronominal impersonal de otros grupos hispanos de Miami que nos per-

MURIEL/OROZCO

31/5/12

14:24

Página 167

LA VARIABLE EXPRESIÓN DE LA IMPERSONALIDAD EN EL ESPAÑOL

167

mitan resolver estos interrogantes. Un estudio futuro que analice el grado de influencia que ejerce un grupo hispano sobre otros y las percepciones de una comunidad frente a las otras, ayudaría a determinar el grado de influencia interdialectal. Guiados por Kuteva/Heine (2008: 219), podríamos también plantear la posibilidad de que este uso de tú impersonal tenga un nuevo significado en el contexto de Miami. Como se mencionó en la introducción, es una variante de lengua española proveniente del latín, que en los dos últimos siglos ha adquirido una mayor frecuencia y su empleo se ha extendido a contextos nuevos. Esta situación constituiría un ejemplo para la idea planteada en Company (2008), relativa a que el contacto lingüístico y la convergencia comunicativa reactivan y difunden “[…] viejas estructuras de una lengua, ya perdidas en la norma general, como consecuencia de la convivencia interétnica y el contacto cultural en una determinada comunidad” (2008: 37). Los resultados de este trabajo constituyen uno de los pocos estudios variacionistas sobre la expresión de la impersonalidad que encara este tema con la comparación de la comunidad en dos contextos diferentes, y que utiliza los mismos procedimientos de análisis. Un estudio futuro puede considerar las variables extralingüísticas más relevantes de esta investigación (por ej., dominio socioeconómico, redes de interacción social, variedad dialectal, edad y duración del contacto) e incorporar algunos factores lingüísticos como el tipo de discurso. Ya que las situaciones de contacto interdialectal son diferentes en cada región de Estados Unidos, sería interesante seguir explorando el sistema pronominal impersonal de los colombianos en ciudades como Nueva York y Chicago, lugares en los que se encuentra un gran número de colombianos.

MURIEL/OROZCO

31/5/12

14:24

Página 168

MURIEL/OROZCO

31/5/12

14:24

Página 169

SECOND PERSON SINGULAR PRONOUN USE D U R I N G C L O S E D R O L E P L AY S B Y C O L O M B I A N S I N T H E N O RT H C A R O L I N A P I E D M O N T T R I A D * SCOTT LAMANNA

1. Introduction The varieties of Spanish spoken in the United States have been the subject of numerous books, dissertations, and collections of scholarly articles in recent decades (Amastae/Elías-Olivares 1982, Bergen 1990, Chaston 1987, Durán 1981, Elías-Olivares 1983, Elías-Olivares/Leone/Cisneros/Gutiérrez 1985, Fishman/Cooper/Ma 1971, Fishman/Keller 1982, Hurtado 2001, Lacorte/Leeman 2009, Lipski 2008b, López Morales 2009, Lynch 1999, Ornstein-Galicia/Green/Bixler-Márquez 1988, Orozco 2004, Ortiz López/Lacorte 2005, Porcel 2003, Roca 2000, Roca/Colombi 2003, Roca/Lipski 1993, Silva-Corvalán 1994, 1995, Valdés/Lozano/García-Moya 1981, Wherritt/García 1989, Zentella 1997b). Several authors have commented on theoretical and methodological issues and challenges related to the study of these varieties (Lipski 2000, Otheguy/Zentella/Livert 2007, Silva-Corvalán 2004, Torres 1991, Zentella 1990, 2004). These include the relationship of such dialects to monolingual Spanish varieties (Torres 1991) and the role of dialect contact in the formation of Spanish-speaking communities in the US (Otheguy/Zentella/Livert 2007, Zentella 1990). These authors have pointed out the need to use a variety of data collection methods to access “a greater range of the total linguistic repertoire of individual speakers” (Torres 1991: 264) and to expand the scope of research beyond the more heavily-studied Mexican, Puerto Rican, and Cuban varieties to include the Spanish of other groups such as those of Central and South American origin (Lipski 2000, Silva-Corvalán 2004, Zentella 2004). One of these less-studied varieties is Colombian Spanish, which represents a promising avenue for the expansion of research on Spanish in the United States. Colombians make up a significant component of the US Hispanic population, as * The author would like to thank Julie Auger, J. Clancy Clements, Manuel Díaz-Campos, Stephen Fafulas, Richard J. File-Muriel, Kimberly Geeslin, Rafael Orozco, and two anonymous reviewers for their assistance with various aspects of this research project, including the design and carrying out of the study and helpful comments on earlier drafts. All errors remain my own.

MURIEL/OROZCO

170

31/5/12

14:24

Página 170

SCOTT LAMANNA

the largest group of South American origin in the country and the seventh largest group overall (US Census Bureau 2010). The status of Colombian Spanish as a minority variety within a larger minority language community makes it appealing as a focus of empirical investigation. Unlike Mexicans, Puerto Ricans, and Cubans, Colombians do not represent the largest group in any of the major Hispanic communities in the US. Thus they serve as an ideal case for examining dialect contact, as presumably the influence of more widely-spoken varieties of Spanish would be exerted on the Colombian variety. Relatively little empirical research has been carried out to date on the Spanish of Colombian communities in the United States, and what has been done has been limited to the states of Florida and New York. Hurtado (2001) studied variable subject expression in the Spanish of Colombians in Miami-Dade County. Orozco (2004) carried out a comparative study of the expressions of futurity and possession in Colombian Spanish as spoken in Barranquilla, Colombia, and New York City. A series of articles based on and/or going beyond these dissertations was subsequently published (Hurtado 2005a, 2005b, this volume, Orozco 2007a, 2007b, 2009b, this volume), and at least two other studies have been carried out on Colombian immigrants in New York State (Ramírez 2003, 2007b). The current investigation seeks to expand the range of Colombian communities studied in the US by examining second person singular pronoun use during closed role plays by Colombians residing in the North Carolina Piedmont Triad (the cities of Winston-Salem, Greensboro, and High Point, and the surrounding counties), where the Hispanic community is predominantly Mexican. The Piedmont Triad has a recently established and rapidly growing Colombian community in the midst of what was the third-fastest growing Hispanic population in the United States between 1990 and 2000 (Bailey 2005). This growth has continued over the past decade as well, with the Colombian population tripling in size over ten years (US Census Bureau 2010). As a region composed of three medium- to small-sized cities, the Triad offers an opportunity to study Colombian Spanish not only outside of areas where Colombians are most numerous (e.g. New York and Miami), but also in an environment other than a major metropolis. In addition, studies of Colombian Spanish in the US have not yet been carried out in a state in which Mexicans are the largest Hispanic group, since speakers of Caribbean dialects dominate in both New York and Florida. In the present study, Colombians in North Carolina are compared both to a corresponding group of Colombians in Bogotá (bogotanos) and to Mexicans from the same North Carolina community in order to determine whether there is evidence of dialect contact shaping their use of second person singular pronouns. The variable of second person singular pronominal address was chosen because no research has previously been done on this variable among Colom-

MURIEL/OROZCO

31/5/12

14:24

Página 171

SECOND PERSON SINGULAR PRONOUN USE

171

bians in the US1. This issue is worth exploring since, according to Lipski (1994: 213), “[t]he most outstanding morphosyntactic variable in Colombian Spanish is the choice of familiar pronouns, and the corresponding verb morphology”. This is particularly true for the Bogotá dialect, where tú, usted, and su mercé/su merced are all available options (Montes Giraldo et al. 1998), although not all speakers use all three forms. Also, Cashman (2007) argues that research on politeness in US Spanish needs to be expanded beyond the Mexican-American majority to include other less-frequently studied varieties. The current study addresses this gap in the literature by focusing on one specific area of linguistic politeness among Colombians: the use of pronominal address forms in interpersonal relations.

2. Second person singular pronominal address: Theory and research Brown/Gilman (1960) defined two principles at work in their seminal paper analyzing pronominal address forms in a number of European languages: the power semantic and the solidarity semantic. The power semantic refers to situations where one speaker has power over another, and in which pronominal address is characterized by a non-reciprocal pattern, with the superior using the T (taken from the Latin tu) form to address the inferior, and the inferior giving the V (taken from the Latin vos) form in return. The solidarity semantic refers to situations where the interlocutors have equal power and which are characterized by a reciprocal pattern of address. A mutual T form is used between interlocutors who share a high degree of solidarity, while those who share a low degree of solidarity use a mutual V form. The authors argued that the solidarity semantic was becoming dominant in modern European societies due to their increasingly egalitarian nature. Brown/Levinson (1987) discussed the use of T and V forms in relation to the concepts of positive and negative politeness respectively. They defined positive politeness as “the expression of solidarity” and negative politeness as “the expression of restraint” (Brown/Levinson 1987: 2). In their discussion of positive politeness strategies, they mentioned forms of address under Strategy 4 (Use in-group identity markers) and claimed that in languages with T/V systems, “the use of a T (singular non-honorific pronoun) to a non-familiar alter2 can claim solidarity” (Brown/Levinson 1987: 107). When commenting on negative politeness strate-

1

Hurtado (2005a) deals with variable expression of tú and usted as subject pronouns, but does not examine the choice between the two when addressing an interlocutor. 2 This term means ‘other’ in Latin.

MURIEL/OROZCO

172

31/5/12

14:24

Página 172

SCOTT LAMANNA

gies, they discussed forms of address under Strategy 5 (Give deference). Specifically, T/V forms were presented as an example of referent honorifics, in which the V form is used to show respect to the hearer. Forms of address were also discussed under Strategy 7 (Impersonalize S and H)3. The authors pointed out a general pattern observed in the T/V systems of many languages in which a second person plural pronoun is used as a respectful singular pronoun. They mentioned not singling out the addressee and referring to the addressee’s status derived from group membership as possible motives for such usage. According to Brown/Gilman (1960: 254), the T and V forms in Spanish are tú and usted, respectively. In standard Spanish and in many spoken varieties, speakers must choose between these two pronouns, traditionally defined as informal and formal, respectively (Rossomondo 2002: 115), when directly addressing an interlocutor. In some dialects of Latin American Spanish, including parts of Colombia, a different pronoun, vos, is used in familiar domains (Lipski 1994: 141). And in parts of the eastern Andean region of Colombia, including Bogotá, su mercé/su merced exists as another option, and can either communicate deference and respect or function as a marker of shared identity or group membership (Ruiz Morales 1987). A number of studies have investigated pronominal address forms, principally tú and usted, in different varieties of Spanish. These include the Spanish of Colombia (Bartens 2004, Bayona 2006, Montes Giraldo 1967, Rey 1994a, Rincón 2006, Ruiz Morales 1987, Uber 1984, 1985), Ecuador (Placencia 1997), Honduras (Castro-Mitchell 2000), Mexico (Covarrubias 2002, Lastra de Suárez 1972), Puerto Rico (Uber 2000), and Spain (Alba de Diego/Sánchez Lobato 1980, Blas Arroyo 1995, Molina 1993, Rossomondo 2002). Other studies have dealt with more than one national variety (Rey 1994b, Schwenter 1993, Solé 1970, Uber 1999), or have discussed the phenomenon in general (Marín 1972, Solé 1978). Factors that were found to influence choice of pronoun in the above studies can be divided into participant (subject) and interlocutor (addressee) characteristics. Participant characteristics include gender (Bartens 2004, Blas Arroyo 1995) and age (Bayona 2006, Blas Arroyo 1995, Molina 1993, Rossomondo 2002). Interlocutor characteristics include social class (Schwenter 1993), age (Bayona 2006, Schwenter 1993), gender (Bartens 2004, Bayona 2006, Molina 1993, Schwenter 1993), whether the interlocutor is a family member or not (Alba de Diego/Sánchez Lobato 1980, Molina 1993, Rossomondo 2002), the participant’s degree of acquaintance with the interlocutor (Bartens 2004, Molina 1993, Rossomondo 2002), and the relative status of the interlocutor compared with the participant (Bartens 2004, Rossomondo 2002). 3

S and H refer here to the speaker and hearer respectively.

MURIEL/OROZCO

31/5/12

14:24

Página 173

SECOND PERSON SINGULAR PRONOUN USE

173

Pronominal address forms in US Spanish have also been the subject of a number of studies (Brown 1975, Hernández 2002, Jaramillo 1990, 1995, 1996, Schreffler 1994). Regarding participant characteristics, Jaramillo (1995, 1996) found that age, but not gender, influenced choice of address form, whereas Schreffler (1994) found that gender was a factor. Hernández (2002) found that age of arrival in the US influenced the choice of pronoun. As far as interlocutor characteristics are concerned, Jaramillo (1990) found that the participants’ degree of acquaintance with the interlocutor, as well as the relative age and status of the interlocutor, played a role in address form choice. Schreffler (1994) found the interlocutor’s gender to influence pronoun choice for males but not for females. Hernández (2002) and Schreffler (1994) examined address forms in order to specifically investigate the influence of dialect contact on second person singular pronoun use. Both investigators researched Salvadoran Spanish in contact with Mexican Spanish in Houston, Texas. Salvadoran Spanish is characterized by its use of vos as a familiar pronoun, while in Mexican Spanish tú predominates (Lipski 1994). Both studies presented evidence that Salvadorans in Houston were accommodating to their Mexican interlocutors by using more tú and less vos than what was found for El Salvador. The current study is the first on US Spanish to examine the effect of contact between a dialect where usted is used frequently by many speakers in a wide range of contexts, including familiar or informal ones (the eastern Andean region of Colombia, including Bogotá) (Lipski 1994, Travis 2002, Uber 1985, 2011) and one in which tú is clearly preferred in familiar domains (Mexico) (Covarrubias 2002, Lastra de Suárez 1972, Schwenter 1993).

3. Methodology 3.1. RESEARCH QUESTIONS The current study investigates whether second person singular pronoun use in the Spanish of Colombians in the North Carolina Piedmont Triad, when compared with use of the same forms by Colombians in Bogotá (bogotanos) and Mexicans from the same North Carolina community, shows evidence of an effect of dialect contact with Mexican Spanish. The following three research questions guided the study: 1. What is the frequency of use of second person singular pronouns during closed role plays in the Spanish of Colombians in North Carolina and

MURIEL/OROZCO

31/5/12

174

14:24

Página 174

SCOTT LAMANNA

Bogotá and Mexicans in North Carolina? Are there statistically significant differences among the three groups? 2. What factors determine the use of second person singular pronouns during closed role plays overall and in each of the three participant groups individually? Are these factors the same or different for the three groups? 3. Does the use of second person singular pronouns by Colombians in North Carolina during closed role plays show evidence of accommodation to the use of these forms by Mexicans from the same community?

3.2. HYPOTHESES The following hypotheses address each of the research questions in turn: 1. Colombians in North Carolina will use tú more frequently than Colombians in Bogotá but less frequently than Mexicans in North Carolina during closed role plays. There will be a statistically significant difference between Colombians in Bogotá and both Colombians and Mexicans in North Carolina, but not between the two North Carolina groups. 2. Factors involving both characteristics of the participant and his or her interlocutor will determine the use of second person singular pronouns overall during closed role plays. Participant gender will determine the use of the pronouns for Colombians in Bogotá (Bartens 2004), but not for Colombians and Mexicans in North Carolina (Jaramillo 1995, 1996). Interlocutor characteristics determining the use of the pronouns will be identical for all three groups. These include the participant’s degree of acquaintance with the interlocutor and the relative age and status of the interlocutor (Bartens 2004, Bayona 2006, Jaramillo 1990). 3. The use of second person singular pronouns by Colombians in North Carolina during closed role plays will show evidence of accommodation to the use of these forms by Mexicans from the same community. First, the data will be consistent with the notion that Colombians in North Carolina have moved towards, but are not yet totally replicating, the frequency of tú use of their Mexican counterparts in the community, who form the majority of the North Carolina Hispanic population (US Census Bureau 2010). Second, the two groups in North Carolina will be similar to each other and different from the Bogotá participants in the lack of significance of participant gender in determining the use of the pronouns.

MURIEL/OROZCO

31/5/12

14:24

Página 175

175

SECOND PERSON SINGULAR PRONOUN USE

3.3. PARTICIPANTS Table 1 displays the characteristics of the participants in this research project, who were nearly equally distributed among the three participant groups and both genders: TABLE 1 Participant characteristics (group and gender) Males

Females

Total

NC Colombians

13

11

24

bogotanos

12

13

25

NC Mexicans

12

11

23

Total

37

35

72

The Colombian participants in North Carolina were all either originally from Bogotá or had lived there the greater part of their lives, and the Bogotá participants had lived in that city for all or most of their lives. It was necessary to limit participants in both groups to those representing the Bogotá dialect since the use of pronouns of address varies greatly in Colombia from region to region (Lipski 1994: 213-214). Specific steps were taken to ensure that the two groups of Colombians were comparable in terms of sociodemographic factors such as age and educational level. The NC Colombians participated in the study first, and as one component of the process, subjects completed a background questionnaire in which they provided sociodemographic information. The researcher utilized their responses to determine the characteristics of the NC Colombian participant group as a whole, and subsequent recruitment in Bogotá focused on obtaining informants with similar characteristics whenever possible. The NC Colombians ranged in age from 18 to 65 (mean age = 44.4) and the bogotanos from 15 to 57 (mean age = 31.6). In addition, all participants in both groups had either attained or begun their secondary or post-secondary education, with the majority in both locations having attained or begun the latter. It was especially important for the groups to be comparable regarding this variable since in the first investigation of Bogotá Spanish using sociolinguistic methodology, Montes Giraldo et al. (1998) found that level of education had the greatest effect of all variables considered on the choice of second person singular pronoun. The majority of the Mexican participants in the study were born in the south-central region of Mexico. Both

MURIEL/OROZCO

176

31/5/12

14:24

Página 176

SCOTT LAMANNA

Colombian and Mexican participants in North Carolina were limited to those who had resided in the Piedmont Triad area for at least two years, with length of time in the community ranging from 2 to 9.5 years for the Colombians (mean length of residence = 5.92 years) and 2 to 14.5 years for the Mexicans (mean length of residence = 8.64 years). The investigator recruited participants in both North Carolina and Colombia using a social network approach (Milroy 1987), through personal contacts in both locations.

3.4. DATA COLLECTION INSTRUMENT Participants completed a series of closed role plays in which they were presented with 27 situations likely to occur between two people in daily life. (See the Appendix and also the sample data coding in section 3.6.) The relationships between the specific people involved in each situation represented all possible combinations of the values of three variables: relative age, relative status, and degree of acquaintance (explained in further detail below). Participants had a written version of the situations in front of them and heard a recording of a native speaker of Colombian Spanish reading the same information. At the end of each situation, they were asked what one of the two people would say to the other in response to the situation4. They saw a photo of their imaginary interlocutor (a female5 with physical features typically associated with Hispanics from a variety of countries) to help them to visualize the situation. The instructions directed them to imagine in each case that they were the person indicated, and that they were participating in a real conversation6. This instrument is a modified version

4

The situations, including the final questions, were presented in the third rather than the second person to avoid prompting the informants with any pronoun of address immediately prior to their responses. 5 The gender of the interlocutor remained constant throughout the closed role plays. Each situation involved a male speaking to a female. The decision to hold gender constant was made for the following reasons: (1) To limit the total number of situations in order to avoid participant burnout as a possible influence in their responses; (2) Rossomondo (2002), who also performed a logistic regression analysis of second person singular pronoun choice, did not find either gender of the interlocutor or whether the interlocutor was of the same or opposite gender to be a significant factor. 6 Although both male and female participants were instructed to imagine that they were males for purposes of this investigation, it was still considered reasonable to hypothesize that participant gender would be a significant factor for the bogotanos in Hypothesis (2) above, as this variable has been significant in previous research examining a similar situation. Lauzen/Dozier (2002) studied appearance comments made and received by both male and

MURIEL/OROZCO

31/5/12

14:24

Página 177

SECOND PERSON SINGULAR PRONOUN USE

177

of one used by Schreffler (1994) to elicit second person singular address forms, in which she asked participants to create both sides of a dialogue based on a series of pictures. As is apparent from the description above, the task used in the current study only required participants to create one side of the dialogue. This type of data collection method is classified in the pragmatics literature as a closed role play because there is no response from an interlocutor and thus no interaction, in contrast with an open role play, which includes these features (Kasper/ Dahl 1991). The intent was that the closed role plays would more closely reflect the way second person singular pronouns are actually used in real life contexts than do questionnaires, the data collection method most commonly employed in previous studies investigating this morphosyntactic variable (Medina Morales 2010: 29). Kasper/Dahl (1991) indicated this may be the case by placing closed role plays closer than questionnaires to authentic discourse on a continuum of data collection methods in interlanguage pragmatics research, and the same relative positioning of these methods would presumably apply to native speaker data as well. An additional advantage of this method is highlighted by Yuan (2001), who compared written discourse completion tests (DCTs) to closed role plays (which she termed oral DCTs). She found that the oral DCTs generated more characteristics of natural speech than the written DCTs, including repetitions, omissions, and longer responses.

3.5. PROCEDURE The investigator digitally recorded participants’ responses on a laptop computer using the sound-recording software Audacity and a microphone. Data collection took place in participants’ homes whenever possible. When this was not feasible, data were collected at a mutually convenient location. Participants took approximately 20 to 30 minutes to complete the task, and crucially were unaware that it was intended to elicit second person singular pronominal address forms. The responses of 60 of the informants were subsequently transcribed by the researcher, a near-native speaker of Spanish with extensive familiarity with

female characters during the 1999-2000 prime-time television season and compared episodes with all-male writing teams to episodes written by teams including at least one female. They found a significant increase in appearance comments made by male characters when females helped write the scripts, and also found a significant decrease in the number of occasions where men insulted other men (from 36% of all appearance comments exchanged between men with all-male writing teams to only 12% of appearance comments between men for mixed-gender writing teams).

MURIEL/OROZCO

31/5/12

178

14:24

Página 178

SCOTT LAMANNA

Bogotá speech and a good level of familiarity with the varieties of Mexican Spanish spoken by immigrants in North Carolina. Research assistants who were native speakers of Bogotá Spanish transcribed the responses of the remaining 12 informants (all bogotanos). The method of transcription employed was that described in Bentivoglio/Sedano (1993), with a few slight modifications. Cases of uncertainty regarding what the informants said in particular portions of their responses were checked by an additional research assistant who is also a native speaker of Bogotá Spanish, and the uncertainty was resolved in the majority of cases.

3.6. DATA CODING The dependent variable for the analysis of the closed role plays was choice of second person singular pronoun (tú or usted)7. The transcriptions of the responses were examined to identify all instances in which the speakers indicated their choice of pronoun. In cases of morphological ambiguity (for example, where le was used and it was unclear whether the intended referent was usted or ella), the researcher consulted with the same native speaker who helped resolve uncertain portions of the transcriptions and successfully determined the most likely referent given the context in the majority of cases. For each of the 27 situations comprising the closed role plays, the speaker’s response was counted as only one instance of tú or usted, no matter how many linguistic elements (one, two, three, etc.) he or she produced indicating his or her choice of pronoun. For example, necesito que me ayude and necesito que usted me ayude would both be counted as one instance of usted. Responses containing both pronouns were grouped with tú as cases where tú is allowed. This approach has proven useful in the analysis of at least one other case of morphosyntactic variation in Spanish, that of copula choice (ser vs. estar). Geeslin (2003) grouped two types of responses together under the label ‘estar allowed:’ those where native speakers indicated both copulas were grammatical and those where they indicated only estar was grammatical. There was not a single token of vos produced during the closed role plays,

7 The words tú and usted refer here not only to these pronouns themselves, but also to “any number of co-occurring linguistic elements such as oblique pronouns as well as possessive and verbal forms”. (Jaramillo 1990: 21) During the closed role plays, a speaker in a given situation may have produced tú or usted and/or a co-occurring linguistic element, but all such productions, whether or not tú or usted was overtly realized, were counted as indicating a choice of the pronoun in question.

MURIEL/OROZCO

31/5/12

14:24

Página 179

SECOND PERSON SINGULAR PRONOUN USE

179

and only three cases of su mercé/su merced8, once along with tú and twice with co-occurring linguistic elements indistinguishable from those used with usted. The decision was made to group the first case with tú (as a case where tú is allowed) and the other two with usted. The instrument successfully elicited use of a pronoun and/or a co-occurring element in 1561 out of 1944 responses (80.3%), and the 383 cases (19.7%) where the speaker did not indicate his or her choice of pronoun were excluded from the analysis. The independent extralinguistic variables used in the study can be divided into two categories: participant variables and interlocutor variables. The participant variables are group (NC Colombians, bogotanos, and NC Mexicans) and gender (male and female). Interlocutor variables include relative status (interlocutor of higher, equal, or lower status), relative age (older, same-age, or younger interlocutor), and degree of acquaintance (stranger, acquaintance, or intimate)9. An example of the coding procedure appears below. First, one of the situations, involving a teacher and his student, is presented. Then a sample response is given in (1), along with how it was coded. Situación 19 Un maestro de tercer grado está sentado detrás de su escritorio calificando tareas mientras sus estudiantes toman una prueba de matemáticas. Se le acerca una de sus estudiantes, y él levanta la mirada. Ella le dice que uno de sus compañeros le acaba de pegar. El maestro duda que sea verdad, porque la niña ya le ha mentido varias veces antes. ¿Cómo responde a la queja de su estudiante? (1) Participant 6M1C47: 47-year old NC Colombian male Pues, eh, mira, si el muchacho te vuelve a pegar, me dices, pero por ahora, preferiría que mm te sientes aquí al lado mío un rato y me, me ayudas con lo que estoy haciendo.

8 It is interesting to note that su mercé/su merced was only used by bogotanos. While caution must be exercised in interpreting this finding due to the small number of cases involved, the fact that the NC Colombians did not use this form at all may be indicative of accommodation to their Mexican counterparts in the community. 9 Following Rossomondo (2002), intimate is defined as anyone with whom the interlocutor has a personal relationship. Thus workplace relationships (boss, employee, co-worker) are included in this category along with family and friends. While recognizing that this method of defining intimate may be controversial, it seemed appropriate due to the fact that people often spend large amounts of time with members of all three categories. Also, Rossomondo (2002) obtained significant results for degree of acquaintance with intimate defined in these terms.

MURIEL/OROZCO

31/5/12

14:24

180

Página 180

SCOTT LAMANNA

This response was coded as follows: tú (dependent variable); NC Colombian, male (participant variables); lower status, younger, acquaintance10 (interlocutor variables).

4. Results and discussion Results are organized according to the three research questions and hypotheses that guided the study.

4.1. RESEARCH QUESTION/HYPOTHESIS 1: FREQUENCY OF USE OF SECOND PERSON SINGULAR PRONOUNS

Table 2 below displays the number of situations in which each pronominal form was chosen and percentages of use of each pronoun per informant group and overall during the closed role plays. The NC Colombians appear in the middle of the table to facilitate comparison with the other two groups. TABLE 2 Second person singular pronoun use by participant group Bogotanos

NC Colombians

NC Mexicans

Total

usted

298 (56.4%)

363 (67.6%)

247 (49.8%)

908 (58.2%)

tú

230 (43.6%)

174 (32.4%)

249 (50.2%)

653 (41.8%)

Total

528 (100%)

537 (100%)

496 (100%)

1561 (100%)

Pronoun

10

The teacher-student relationship was categorized as an acquaintance relationship. The rationale for this decision was that teachers tend to interact more frequently with groups of students than with individual students on a one-on-one basis, and also, that the teacher-student relationship tends to last for a relatively short period of time (e.g. an academic year) rather than for longer periods of time. Both of these limitations on the amount of personal contact between a teacher and each of his or her students made it seem more appropriate to categorize this relationship as belonging to the acquaintance rather than to the intimate category. The percentage of tú use by participants in Situation 19 (81.7%) was similar to that observed for some of the other acquaintance relationships, including Situation 7 (76.5% tú use) and Situation 18 (74.5% tú use). (See the Appendix for specific details regarding the relationships included in these situations.)

MURIEL/OROZCO

31/5/12

14:24

Página 181

SECOND PERSON SINGULAR PRONOUN USE

181

The answer to the first part of Research Question 1 is that tú is used most frequently by the NC Mexicans, and usted by the NC Colombians, with the bogotanos falling in between these two groups. Although chi-square tests of independence are frequently used in linguistic research to test for statistical significance in these types of distributions, they are not included here because there is a crucial assumption of this test that does not apply to the data in this study. This assumption is that all the observations are independent of one another (i.e. they each come from a different participant). However, since each participant responded to 27 situations, the assumption is not valid in this case. In order to avoid this problem, the investigator converted the data to only one observation per speaker by calculating an overall percentage of tú use during the 27 closed role plays for each participant. A one-way ANOVA was subsequently run to test for statistically significant differences in the mean percentage of tú use among the three participant groups. Table 3 below shows the means calculated for the three groups, which correspond closely to the percentages reported in Table 2: TABLE 3 Mean percent tú use by participant group N

Mean% tú use

bogotanos

25

.4323

NC Colombians

24

.3256

NC Mexicans

23

.5015

Total

72

.4188

Group

The ANOVA on mean percentage of tú use indicated that the means differ significantly across the three groups: F(2, 69) = 6.314, p = .003. This tells us that at least two of the sample means are significantly different from each other, but does not indicate which ones. Tukey post-hoc comparisons of the three groups were performed to identify the relevant difference(s). The results indicated that the NC Mexicans (M = .5015, 95% CI [.4380, .5649] had a significantly higher percentage of tú use than the NC Colombians (M= .3256, 95% CI [.2710, .3801]), p. = .002. Comparisons between the bogotanos (M = .4323, 95% CI [.3413, .5233]) and the other two groups were not statistically significant at p < .05. Thus the answer to the second part of Research Question 1 is that there is a statistically significant difference between the two North Carolina groups only.

MURIEL/OROZCO

182

31/5/12

14:24

Página 182

SCOTT LAMANNA

Hypothesis 1 predicted that the NC Colombians would use tú more frequently than Colombians in Bogotá but less frequently than Mexicans in North Carolina. It also predicted that there would be a statistically significant difference between Colombians in Bogotá and both Colombians and Mexicans in North Carolina, but not between the two North Carolina groups. Neither part of the hypothesis was supported by the data. Colombians in North Carolina use tú less frequently than the other two groups, and the difference between the two North Carolina groups is the only significant one. In attempting to account for this unanticipated result, the possibility was considered that the two Colombian groups may differ from each other in one or more relevant characteristics (despite efforts to ensure their comparability), and that this may explain why there is a statistically significant difference in percentage of tú use between the NC Colombians and the NC Mexicans but not between the bogotanos and the NC Mexicans. One possibility is that mean age differences between the groups may have been a confounding factor11. A Pearson product-moment correlation coefficient was computed to evaluate the relationship between each Colombian participant’s age and his or her percentage of tú use, but since no correlation was found between the two variables [r = .050, n = 49, p = .734], the mean age difference between the groups was rejected as an explanatory factor. Another possible explanation is that differences in the level of education of the informants could have influenced their use of the pronouns (Montes Giraldo et al. 1998). Of the NC Colombians (n = 24), 21 (87.5%) had attained or begun a post-secondary degree, while 3 (12.5%) had completed or begun secondary school. The bogotanos (n = 25) as a group were somewhat less educated, since only 16 (64%) had attained or begun a post-secondary degree, and 9 (36%) had either completed or begun secondary school. In order to determine whether there was a relationship between level of education and second person singular pronoun use for the Colombians, a Mann-Whitney U test was run on the mean percentage of tú use by those with a secondary vs. post-secondary level of education (34.67% vs. 39.08% tú use respectively). Since the results of this test were not significant (Mann-Whitney U = 189.000, p = .443), this explanation was also rejected. Assuming the two Colombian groups truly are comparable, a third possibility is that greater use of usted by the NC Colombians reflects the fact that second person singular pronoun use functions as an indicator within the Hispanic community of the Piedmont Triad, since it varies according to group membership, in

11

See section 3.3 above for the mean age of each group.

MURIEL/OROZCO

31/5/12

14:24

Página 183

SECOND PERSON SINGULAR PRONOUN USE

183

this case the speaker’s national origin (Labov 1972: 178). There would need to be evidence that stylistic variation occurs as well to label it a marker in the Labovian sense, but in either case, Colombians may be using usted more frequently in the Piedmont Triad than they would in Bogotá in order to distinguish themselves from their Mexican counterparts in the community.

4.2. RESEARCH QUESTION/HYPOTHESIS 2: FACTORS DETERMINING SECOND PERSON SINGULAR PRONOUN USE

A binary logistic regression identified the factors that determine the use of second person singular pronouns when all factors are considered together in a single model. This statistical analysis was first conducted for the data from all three participant groups together to determine which factors are significant overall. The results are reported in Table 4 below: TABLE 4 Logistic regression (overall) Wald

Sig

acquaintance

181.079

.000

status

179.571

.000

age

67.195

.000

age*group

20.941

.000

participant group

20.091

.000

acquaintance*group

20.005

.000

status*group

14.673

.005

participant gender

2.556

.110

gender*group

1.036

.596

The results of the logistic regression analysis answer the first part of Research Question 2 by showing that for the data as a whole, all of the participant and interlocutor variables except for participant gender are statistically significant in determining the informants’ choice of second person singular pronouns during

MURIEL/OROZCO

31/5/12

184

14:24

Página 184

SCOTT LAMANNA

the closed role plays. Thus the results support the first part of Hypothesis 2, which predicted that both participant and interlocutor variables would be significant overall. In addition, the interactions between age and group, acquaintance and group, and status and group are significant. The variables are ranked according to their relative importance or strength in determining the choice of tú or usted based on the value of the Wald statistic (Hosmer/Lemeshow 2000). Degree of acquaintance has the most influence on the use of second person singular pronouns, followed closely by relative status and then by relative age. The interaction between age and group, participant group, the interaction between acquaintance and group, and the interaction between status and group are next in importance. Participant gender and the interaction between gender and group have the least influence and in fact do not make a significant contribution to the model at all12. The results show that the interlocutor variables have a greater influence on the choice between tú and usted than the participant variables, which indicates that, at least for these participants in this task, pronominal address behavior is more determined by the characteristics of the addressee than by the speaker’s own characteristics (specifically, nationality and gender). Additional logistic regression analyses were conducted for each participant group individually, in order to answer the second part of Research Question 2 by identifying the factors that determine the use of second person singular pronouns during closed role plays in each of the three participant groups and noting whether these factors are the same or different for the three groups. Results appear first for the bogotanos, then for the NC Colombians, and finally for the NC Mexicans.

4.2.1. Bogotanos Table 5 displays the results of the logistic regression analysis for the bogotanos. The results indicate that all of the independent extralinguistic variables except for participant gender are statistically significant in determining the choice of second person singular pronoun. Relative status has the strongest influence on pronoun choice for this group, followed by degree of acquaintance, relative age, and participant gender. The results from the three significant variables are examined in greater detail below from the most to the least significant. Table 6 dis12

It is worth noting that results for participant gender might have been significant if participants had been asked to respond to the situations as themselves, rather than always assuming the persona of a male participant in the situations presented to them and, thus, in the case of the females, speaking as someone of the gender opposite their own.

MURIEL/OROZCO

31/5/12

14:24

Página 185

185

SECOND PERSON SINGULAR PRONOUN USE

TABLE 5 Logistic regression (bogotanos) Wald

Sig

relative status

61.202

.000

degree of acquaintance

43.950

.000

relative age

7.673

.022

participant gender

0.128

.721

plays a crosstabulation of second person singular pronoun use by relative status of the interlocutor for the bogotanos: TABLE 6 Second person singular pronoun use by relative status of interlocutor (bogotanos) Pronoun usted tú Total

Higher status

Equal status

Lower status

Total

135 (78.9%)

52 (31.0%)

111 (58.7%)

298 (56.4%)

36 (21.1%)

116 (69.0%)

78 (41.3%)

230 (43.6%)

171 (100%)

168 (100%)

189 (100%)

528 (100%)

The greatest percentage of usted use is with an interlocutor of higher status, while the greatest percentage of tú use is with an interlocutor of equal status. When addressing someone of lower status, usted is used less frequently than with someone of higher status, but still in a majority of cases. Similar results have been obtained in previous research on pronominal address forms in Bogotá (Montes Giraldo et al. 1998) as well as in Madrid (Rossomondo 2002). The findings make sense in light of Brown/Gilman’s (1960) solidarity semantic, if we assume that those of equal status share a higher degree of solidarity (and thus use tú in most cases), while those of different status (represented here by both higher and lower status interlocutors) share a lower degree of solidarity (and thus tend to use usted). Table 7 displays a crosstabulation of second person singular pronoun use by degree of acquaintance with the interlocutor for the bogotanos:

MURIEL/OROZCO

31/5/12

14:24

186

Página 186

SCOTT LAMANNA

TABLE 7 Second person singular pronoun use by degree of acquaintance with interlocutor (bogotanos) Pronoun usted tú Total

Stranger

Acquaintance

Intimate

Total

143 (78.6%)

87 (53.4%)

68 (37.2%)

298 (56.4%)

39 (21.4%)

76 (46.6%)

115 (62.8%)

230 (43.6%)

182 (100%)

163 (100%)

183 (100%)

528 (100%)

The highest percentage of usted use is with strangers, and the highest percentage of tú use is with intimates, while acquaintances fall in between these two extremes with almost equal use of the two pronouns. These results are also similar to those of previous studies. For example, Rossomondo (2002) found that usted was used significantly less often to address an acquaintance than to address a stranger in Madrid. Jaramillo (1990) found that tú was used in a majority of cases in New Mexico in her ‘informal’ domain cluster (nuclear family and friendship, both coded as ‘intimate’ in the current study), while usted was used in a majority of cases in other domains. Finally, Bartens (2004) found more use of tú than usted in Bogotá to address family members, and more use of usted than tú to address strangers. Table 8 displays a crosstabulation of second person singular pronoun use by relative age of the interlocutor for the bogotanos: TABLE 8 Second person singular pronoun use by relative age of interlocutor (bogotanos) Pronoun

Older interlocutor

Same-age interlocutor

Younger interlocutor

Total

usted

110 (60.8%)

111 (60.7%)

77 (47.0%)

298 (56.4%)

71 (39.2%)

72 (39.3%)

87 (53.0%)

230 (43.6%)

181 (100%)

183 (100%)

164 (100%)

528 (100%)

tú Total

For this participant group, the percentage of usted use with an older interlocutor is identical to the percentage with an interlocutor of the same age (60.8% and 60.7% respectively). While usted is used most of the time in both cases, tú is used in a majority of cases with a younger interlocutor, although there is nearly

MURIEL/OROZCO

31/5/12

14:24

Página 187

SECOND PERSON SINGULAR PRONOUN USE

187

equal use of the two pronouns (53.0% tú, 47.0% usted). These results closely parallel those of Bayona (2006), who found a preference for usted with older and same-age interlocutors and nearly equal selection of both pronouns with younger interlocutors among her Bogotá informants. Jaramillo (1990) also found more frequent use of usted with older interlocutors and of tú with younger interlocutors.

4.2.2. NC Colombians Moving on to the second of the two Colombian groups, Table 9 displays the results of the logistic regression analysis for the North Carolina Colombians: TABLE 9 Logistic regression (NC Colombians) Wald

Sig

degree of acquaintance

92.625

.000

relative status

68.605

.000

relative age

24.922

.000

3.467

.063

participant gender

As with the bogotanos, the results of the regression analysis for the NC Colombians indicate that all of the independent extralinguistic variables except for participant gender are statistically significant in determining choice of pronoun during the closed role plays. However, the relative strength of the variables is slightly different than in Bogotá. For the NC Colombians, degree of acquaintance and relative status have switched positions, with the former having more of an influence on the choice of address form than the latter13. For both groups of Colombians, relative age ranks third in strength and participant gender ranks last and is not significant in determining the use of tú and usted. Each of the significant variables will now be examined in more detail in order of significance. Table 10 shows a crosstabulation of second person singular pronoun use by degree of acquaintance with the interlocutor for the NC Colombians:

13

For further discussion of this important finding, see Section 4.3 below.

MURIEL/OROZCO

31/5/12

14:24

188

Página 188

SCOTT LAMANNA

TABLE 10 Second person singular pronoun use by degree of acquaintance with interlocutor (NC Colombians) Pronoun usted tú Total

Stranger

Acquaintance

Intimate

Total

165 (91.2%)

126 (72.0%)

72 (39.8%)

363 (67.6%)

16 (8.8%)

49 (28.0%)

109 (60.2%)

174 (32.4%)

181 (100%)

175 (100%)

181 (100%)

537 (100%)

Once again, as with the bogotanos, the highest percentage of usted use is with strangers, and the highest percentage of tú use is with intimates. There is a noteworthy difference between the two groups, however, in the percentage of use of usted with strangers (91.2% for the NC Colombians versus 78.6% for the bogotanos). A parallel difference is observed with acquaintances. While usted is clearly preferred by the NC Colombians to address this category of interlocutor (used 72% of the time), the bogotanos only used usted in slightly over half the relevant cases (53.4% of the time). These results may be due to the fact that the NC Colombians are more likely to interact with strangers and acquaintances in their community who belong to other Hispanic groups (especially Mexicans), whereas the bogotanos are more likely to interact with strangers and acquaintances who are fellow Colombians. The NC Colombians may use usted more frequently because they share a lower degree of solidarity with these interlocutors from other groups, whereas the bogotanos share a higher degree of solidarity with other Colombians and are thus more likely to use tú. Table 11 displays a crosstabulation of second person singular pronoun use by relative status of the interlocutor for the NC Colombians: TABLE 11 Second person singular pronoun use by relative status of interlocutor (NC Colombians) Pronoun usted tú Total

Higher status

Equal status

Lower status

Total

159 (91.4%)

79 (45.7%)

125 (65.8%)

363 (67.6%)

15 (8.6%)

94 (54.3%)

65 (34.2%)

174 (32.4%)

174 (100%)

173 (100%)

190 (100%)

537 (100%)

MURIEL/OROZCO

31/5/12

14:24

Página 189

189

SECOND PERSON SINGULAR PRONOUN USE

As with the Bogotá participants, the highest percentage of usted use is with those of higher status, and the highest percentage of tú use is with those of equal status, while pronoun use with those of lower status falls in between these two extremes. Again, these results make sense in light of Brown/Gilman’s (1960) solidarity semantic, as discussed above with regard to the bogotanos. Table 12 shows a crosstabulation of second person singular pronoun use by relative age of the interlocutor for the NC Colombians: TABLE 12 Second person singular pronoun use by relative age of interlocutor (NC Colombians) Pronoun

Older interlocutor

Same-age interlocutor

Younger interlocutor

Total

usted

135 (76.3%)

131 (71.6%)

97 (54.8%)

363 (67.6%)

42 (23.7%)

52 (28.4%)

80 (45.2%)

174 (32.4%)

177 (100%)

183 (100%)

177 (100%)

537 (100%)

tú Total

As expected in light of the results for Bogotá, the highest percentage of usted use is with an older interlocutor, while the highest percentage of tú use is with a younger interlocutor. Again, following the same pattern observed with the bogotanos, the percentage of usted use is similar for older and same-age interlocutors, and there is nearly equal use of the two pronouns with a younger interlocutor (with a slight preference for usted in this case).

4.2.3. NC Mexicans Findings for the North Carolina Mexican participants in this study will now be considered. Table 13 displays the results of the logistic regression analysis for this group. These results indicate that, once again, all of the independent extralinguistic variables except for participant gender are statistically significant in determining choice of pronoun during the closed role plays. However, the relative strength of the variables is different than in the case of the two Colombian groups. For the NC Mexicans, relative age has the greatest influence on the use of tú and usted, whereas this variable ranks third for both groups of Colombians. Relative age is followed in importance by degree of acquaintance and relative

MURIEL/OROZCO

31/5/12

14:24

Página 190

190

SCOTT LAMANNA

TABLE 13 Logistic regression (NC Mexicans) Wald

Sig

relative age

68.695

.000

degree of acquaintance

58.629

.000

relative status

55.432

.000

participant gender

0.522

.470

status, with participant gender ranking last. The significant variables are again examined in greater detail in order of significance. Table 14 shows a crosstabulation of second person singular pronoun use by relative age of the interlocutor for the NC Mexicans: TABLE 14 Second person singular pronoun use by relative age of interlocutor (NC Mexicans) Pronoun usted tú Total

Older interlocutor

Same-age interlocutor

Younger interlocutor

Total

115 66.5%

84 46.7%

48 33.6%

247 49.8%

58 33.5%

96 53.3%

95 66.4%

249 50.2%

173 100%

180 100%

143 100%

496 100%

Once again, the highest percentage of usted use is with an older interlocutor, while the highest percentage of tú use is with a younger interlocutor. However, in contrast with both groups of Colombians, where the percentages of usted use with an older and same-age interlocutor were quite similar, there is a large percentage difference (19.8%) for the Mexicans in the use of usted with an older interlocutor (66.5%) versus a same-age interlocutor (46.7%). This difference explains why the interaction between group and age was significant in the logistic regression for the data as a whole. Table 15 shows a crosstabulation of second person singular pronoun use by degree of acquaintance with the interlocutor for the NC Mexicans:

MURIEL/OROZCO

31/5/12

14:24

Página 191

191

SECOND PERSON SINGULAR PRONOUN USE

TABLE 15 Second person singular pronoun use by degree of acquaintance with interlocutor (NC Mexicans) Pronoun

Stranger

Acquaintance

Intimate

Total

usted

121 70.8%

83 55.0%

43 24.7%

247 49.8%

50 29.2%

68 45.0%

131 75.3%

249 50.2%

171 100%

151 100%

174 100%

496 100%

tú Total

As with both groups of Colombians, the highest percentage of usted use is with strangers and the highest percentage of tú use is with intimates. Acquaintances elicit a nearly equal use of the two pronouns, the same as in the case of the bogotanos, which contrasts with the NC Colombians’ use of usted 72% of the time with acquaintances. The statistically significant interaction between group and degree of acquaintance in the overall regression analysis may be due, at least in part, to this difference. Another contributing factor may be the percentage of use of usted with intimates: 39.8% and 37.2% of the time for the NC Colombians and bogotanos respectively but only 24.7% of the time for the NC Mexicans. Table 16 displays a crosstabulation of second person singular pronoun use by relative status of the interlocutor for the NC Mexicans: TABLE 16 Second person singular pronoun use by relative status of interlocutor (NC Mexicans) Pronoun usted tú Total

Higher status

Equal status

Lower status

Total

118 76.6%

43 27.2%

86 46.7%

247 49.8%

36 23.4%

115 72.8%

98 53.3%

249 50.2%

154 100%

158 100%

184 100%

496 100%

The overall pattern of the results is the same as with both groups of Colombians. The highest percentage of usted use is with those of higher status, while the highest percentage of tú use is with those of equal status, although for the Mexicans, tú is used in a (slight) majority of cases with those of lower status as well. Since, however, the percentages in this last context are nearly equal, this

MURIEL/OROZCO

31/5/12

14:24

Página 192

192

SCOTT LAMANNA

may indicate that when the Mexicans address someone of lower status, both Brown/Gilman’s (1960) solidarity semantic (which would lead to the choice of usted) and power semantic (which would lead to the choice of tú) play a role in their decision regarding which form to use. Since the results of the logistic regression for the data as a whole indicated a significant interaction between status and group, the percentages for the three participant groups were examined in greater detail to identify any differences existing within the overall similarity of the pattern. This examination revealed that the percentage difference between the use of usted with a higher and lower status interlocutor was considerably greater for the NC Mexicans (29.9%) than for the bogotanos (20.2%), with the NC Colombians falling in between these two groups (25.6%). To summarize the results of the logistic regression analyses for the three participant groups, the factors determining pronoun use (degree of acquaintance, relative status, and relative age) were exactly the same for all three groups (although with different orders of significance). The second part of Hypothesis 2 correctly predicted that these three interlocutor variables would be significant for all groups, but incorrectly predicted that participant gender would be significant as well for the bogotanos only. Results indicated instead that this variable was not significant for any of the groups.

4.3. RESEARCH QUESTION/HYPOTHESIS 3: ACCOMMODATION IN SECOND PERSON SINGULAR PRONOUN USE

A simple examination of the frequency of use of tú and usted by the three participant groups (bogotanos: tú 43.6%, usted 56.4%; NC Colombians: tú 32.4%, usted 67.6%; NC Mexicans: tú 50.2%, usted 49.8%) does not support the prediction of Hypothesis 3 that there would be accommodation by NC Colombians to NC Mexicans in the use of these pronouns. On the contrary, since the difference between the bogotanos and the NC Mexicans is not statistically significant, while that between the two North Carolina groups is, these results suggest that the NC Colombians may actually be diverging from the NC Mexicans in terms of their frequency of use of tú and usted. Thus the first part of Hypothesis 3, which predicted that the evidence would show that Colombians in North Carolina have moved towards, but are not yet totally replicating, the frequency of tú use of their Mexican counterparts in the community, is not supported by the data. In order to determine whether the NC Colombians are in fact diverging from the NC Mexicans in their use of second person singular pronouns, the NC Colombians were assigned to one of three groups according to their length of

MURIEL/OROZCO

31/5/12

14:24

Página 193

SECOND PERSON SINGULAR PRONOUN USE

193

residence in the Piedmont Triad. A one-way ANOVA was then run to test for differences in the mean percentage of tú use among the three groups. Table 17 below shows the mean calculated for each group: TABLE 17 Mean percent tú use of NC Colombians by length of residence in Piedmont Triad N

Mean% tú use

2-3 years

7

.2965

4 to 6.75 years

8

.3436

7 to 9.5 years

9

.3301

24

.3248

Length of residence

Total

The ANOVA on mean percentage of tú use indicated that the means do not differ significantly across the three groups: F(2, 21) = .243, p = .786. This means that these data do not support the notion that, over time, the NC Colombians are either distancing themselves from or moving towards the NC Mexicans in terms of their frequency of use of second person singular pronouns. Thus, the frequency data alone do not provide support for a dialect contact hypothesis. There are at least two possible interpretations of these results. One is that the NC Colombians are in fact marking their Colombian identity by using more usted in North Carolina than they would in Bogotá, but that this begins soon after their arrival in the community (within the first two years) without varying significantly over time after that. In order to confirm that this is indeed the case, future research would need to include data from newly arrived immigrants (those with less than two years’ residence in the community) and compare these data with the data from the current study. Another possible interpretation (Rafael Orozco, personal communication) is that the NC Colombians exhibit frequencies of use of tú and usted that reflect the norm in Bogotá at the time of their departure, and have not participated in any increase in the use of tú that may have taken place in Bogotá since they left, nor have they yet begun to accommodate to the NC Mexicans in their frequency of use of this pronoun. This may simply indicate that evidence of accommodation in dialect contact situations takes longer than a decade to appear, if it appears at all. Moving on to the second part of Hypothesis 3, it was predicted that the two groups in North Carolina would be similar to each other and different from the Bogotá participants in the lack of significance of participant gender in determin-

MURIEL/OROZCO

31/5/12

14:24

Página 194

194

SCOTT LAMANNA

ing the use of second person singular pronouns. Results of the individual regression analyses showed that participant gender does not provide evidence of accommodation in pronoun use since, as mentioned above, this variable is not significant for any of the three participant groups (including the bogotanos). Although the above results do not support a dialect contact hypothesis, the issue of dialect contact was addressed from another angle. Specifically, the variable hierarchies from the three logistic regression runs were compared to see whether they provide any evidence of convergence in pronoun use between the two North Carolina groups. Table 18 displays the variable hierarchies, including the Wald statistic and significance for each variable, for all three participant groups: TABLE 18 Variable hierarchies from logistic regression runs on each participant group bogotanos Variable

NC Colombians

Wald

Sig.

Wald

Sig.

61.202

.000 1. acquaintance 92.625

.000

2. acquaintance 43.950

.000 2. status

68.605

3. age

7.673

.022 3. age

4. gender

0.128

.721 4. gender

1. status

Variable

NC Mexicans Variable

Wald

Sig.

1. age

68.695

.000

.000

2. acquaintance

58.629

.000

24.922

.000

3. status

55.432

.000

3.467

.063

4. gender

0.522

.470

The variable hierarchies provide some evidence that the NC Colombians, if not converging with their Mexican counterparts in the community in terms of overall frequency of use of each pronoun, may nevertheless be converging in terms of the relative influence of the variables that determine pronoun choice. The first piece of evidence suggesting this is that whereas in Bogotá, status has more of an influence in determining pronoun use than acquaintance, the opposite is true for the NC Colombians. For this participant group, acquaintance is ranked higher than status, the same relative ranking of these variables observed for the NC Mexicans. Another piece of evidence is found in comparing the Wald statistic and significance for the variable of age for both groups of Colombians with the corresponding values for the NC Mexicans. These values are more similar to those of the Mexicans in the case of the NC Colombians. Since age is the most influential variable for the NC Mexicans in determining pronoun use, the NC Colombians appear to be accommodating to their Mexican counterparts in the

MURIEL/OROZCO

31/5/12

14:24

Página 195

SECOND PERSON SINGULAR PRONOUN USE

195

community by assigning this variable a greater influence than they would in Bogotá. These two pieces of evidence lend support to Hypothesis 3, that the use of address forms by Colombians in North Carolina will show evidence of accommodation to the use of these forms by Mexicans from the same community, albeit in a different fashion than predicted.

5. Conclusion The current investigation has expanded the scope of research on Colombian Spanish in the US by examining the use of second person singular pronouns by Colombians in the North Carolina Piedmont Triad. The study also makes a methodological contribution to research on pronominal address forms through the use of closed role plays, a data collection method whose results more closely approximate natural speech than questionnaire data. The NC Colombians were compared both to a corresponding group of Colombians in Bogotá (bogotanos) and to Mexicans from the Piedmont Triad to determine whether there is any evidence of dialect contact shaping their use of tú and usted. Results of the study are mixed, with some of the data providing such evidence, and other data failing to do so. A summary of these results, in light of the three research questions and hypotheses that guided the study, appears below. The first research question asked what the frequency of second person singular pronoun use was in the three participant groups and whether there were statistically significant differences among the groups. Hypothesis 1 predicted that the NC Colombians would use tú more frequently than the bogotanos but less frequently than the NC Mexicans, and that there would be a statistically significant difference between the bogotanos and both Colombians and Mexicans in North Carolina, but not between the two North Carolina groups. Neither part of the hypothesis was supported by the data. The NC Colombians use tú less frequently than the other two groups and the only significant difference is between the two North Carolina groups. The second research question asked what factors determined the use of second person singular pronouns overall and in each of the three participant groups individually, and whether these factors were the same or different for the three groups. The first part of Hypothesis 2 predicted that both participant and interlocutor variables would determine the use of the pronouns overall, and this was supported by the data, since participant group and all three interlocutor variables (degree of acquaintance, relative age, and relative status) were significant for the data as a whole. The second part of Hypothesis 2 predicted that participant gender would be significant for the bogotanos, but not for the two North Carolina

MURIEL/OROZCO

196

31/5/12

14:24

Página 196

SCOTT LAMANNA

groups, and that the interlocutor variables would be significant for all three groups. Participant gender turned out not to be significant for any of the groups, while the interlocutor variables were indeed significant for all three. Thus the results supported most of the second part of this hypothesis by indicating that the factors determining pronoun use were exactly the same for all three groups. The third research question asked whether the use of second person singular pronouns by Colombians in North Carolina showed evidence of accommodation to the use of these forms by Mexicans from the same community, and Hypothesis 3 predicted that this would indeed be the case. More specifically, it predicted that the evidence would show that Colombians in North Carolina have moved towards, but are not yet totally replicating, the frequency of tú use of their Mexican counterparts. The hypothesis also predicted that the two North Carolina groups would be similar to each other and different from the bogotanos in the lack of significance of participant gender in determining the use of the pronouns. Neither of these specific predictions was supported by the data, since the frequency of use of tú and usted did not show the NC Colombians either accommodating to the NC Mexicans or distancing themselves from them, and since participant gender was not significant for any of the three groups and thus provided no evidence of convergence. However, the variable hierarchies that determine pronoun choice for the three participant groups provided support for the general prediction of Hypothesis 3. These results showed that the hierarchy of the NC Colombians was closer than that of the bogotanos to the hierarchy of the NC Mexicans, thus providing some evidence of convergence between the two North Carolina groups in their use of second person singular pronouns. For both of the NC groups, degree of acquaintance was more important than relative status, whereas the reverse was true in Bogotá. Concurrently, relative age, the most influential variable for the Mexicans, was more important to Colombians in North Carolina than to those in Bogotá. Although the NC Colombians are not (or are not yet) accommodating to the NC Mexicans in the more overt linguistic behavior of frequencies of pronoun use, they do appear to be converging in terms of the relative importance assigned to different variables when choosing a pronoun. These findings indicate that dialect contact can influence linguistic phenomena below the level of consciousness of the individual speaker (variable hierarchies) without (or before) influencing behavior of which speakers are more aware (choice of individual pronouns). The findings of the current study also provide evidence that the effects of contact on the choice of second person singular pronoun between two dialects where both tú and usted are available options (albeit with different frequencies of use in specific contexts) are different from the results of contact between a dialect that prefers vos and another that prefers tú in informal domains. Both

MURIEL/OROZCO

31/5/12

14:24

Página 197

SECOND PERSON SINGULAR PRONOUN USE

197

Hernández (2002) and Schreffler (1994) found evidence suggesting that Salvadorans in Houston use less vos and more tú than Salvadorans in El Salvador in accommodation to their Mexican interlocutors, while the opposite result was obtained regarding usted and tú for Colombians in North Carolina. The difference may in part be due to the fact that vos is not a socially prestigious form in most of Latin America (Lipski 1994: 141-143), whereas usted enjoys prestige throughout the Hispanic world, as the norm for expressing formality, politeness, power, and distance with a single interlocutor in both Peninsular and Latin American varieties of Spanish (Carricaburo 1997: 10-13). Another possible explanation is the degree of linguistic (in)security felt by the different groups. Lipski (2008b: 144) mentions the linguistic insecurity that Central Americans (including Salvadorans) experience when they come into contact with speakers of other Spanish dialects. In contrast, Zentella (2002) reported that Colombians enjoyed the greatest degree of linguistic security among various Hispanic groups surveyed in New York City. Regardless of the correct explanation(s), the implications of the above findings are that a “one size fits all” approach to the analysis of pronominal address behavior in situations of dialect contact is not appropriate, since both contributing factors and linguistic outcomes may vary. Nevertheless, before drawing definitive conclusions from these results, we must acknowledge that there are a number of limitations to the present study. One is the issue of the extent of correspondence between oral production during role plays (open or closed) and during naturalistic interaction. In a study examining correlations in various aspects of behavior (including linguistic production) between these two contexts, Bellack/Hersen/Lamparski (1979) found a moderate degree of correlation for females and a minimal degree of correlation for males. However, these authors acknowledged that “(v)ariations in format, item content, target behaviors, and subject populations all could yield different results” (Bellack/Hersen/Lamparski 1979: 342). A more recent study that carried out a similar comparison was Félix-Brasdefer (2007). He examined naturally-occurring and open role-play requests in Mexican Spanish and found some degree of correspondence between the two contexts. These included, among other characteristics, “a variety of precursors used to preface a request and high indices of conventional indirectness as the default strategy” (Félix-Brasdefer 2007: 178). However, there were notable differences as well, with natural speech samples being more varied, more elaborate, and longer, and including some features not present in the role-play data such as elliptical requests, use of the negative particle no and emphatic sí to preface requests, and non-verbal responses to requests. In light of the findings of the above studies, we should keep in mind that while the participants in the current investigation indicated during closed role plays how they thought someone would respond to a given situation, these

MURIEL/OROZCO

198

31/5/12

14:24

Página 198

SCOTT LAMANNA

responses may in fact differ from how the participants themselves would actually respond to a similar situation in a naturalistic context (where there is interaction with an interlocutor), including their choice of a pronoun of address. Indeed, lack of interaction is one of the principal shortcomings of the closed role play mentioned in the pragmatics literature (Márquez Reiter 2000, Yuan 2001). However, even if the responses from the closed role plays differ from naturalistic interaction in terms of the specific linguistic structures utilized, the length and variety of the responses, or other characteristics, it is possible that pronominal address behavior would be consistent across the two contexts. Additional research is needed comparing second person singular pronoun use during natural speech and speech elicited through closed role plays to shed further light on this issue. Another possible confounding factor in the current investigation involves the effects of the methodology (specifically the presence of the investigator and the recording of responses) on informants’ linguistic production (Márquez-Reiter 2000). The participants in the study may have produced a different speech sample than in a naturalistic context, due to their desire to make a positive impression on the investigator and/or on those who would listen to the recordings, by using more formal, polite, or “correct” language than they would in natural speech, and this may have influenced their use of address forms (although all three groups would presumably be affected equally in this way). Again, further research is needed to clarify this issue. In addition to completing the closed role plays, the participants in the current study also filled out questionnaires regarding second person singular pronoun use. Future research will analyze the results of those questionnaires and compare them to the responses from the closed role plays reported here. Investigations of second person singular pronoun use will be carried out in other Colombian immigrant communities as well (both within and outside the United States) to shed more light on the role of dialect contact in shaping the pronominal address behavior of Colombians.

MURIEL/OROZCO

31/5/12

14:24

Página 199

SECOND PERSON SINGULAR PRONOUN USE

199

6. Appendix SITUACIONES Instrucciones: Abajo se describen varias situaciones que podrían ocurrir entre dos personas en la vida diaria. Al final de cada situación, se pregunta qué diría una de las dos personas a la otra como respuesta a la situación. Usted debe imaginarse que usted mismo/a es la persona indicada, y que está participando en una conversación real. Se le va a mostrar una foto de la otra persona, a quien usted le habla, para ayudarle a visualizar la situación. Al mismo tiempo que ve la foto, usted oirá la descripción de la situación que aparece abajo. Después de oír la situación entera, por favor diga lo primero que se le ocurra, y todo lo que se le ocurra. ¡Muchas gracias por su ayuda! 1. Una jefa llama por teléfono a la casa de uno de sus empleados, quien es mayor que ella, para informarle que él tendrá que llegar temprano a la oficina mañana para poder entregar un trabajo a tiempo. El empleado ya tiene planes de llevar a su esposa al médico temprano por la mañana antes de entrar a trabajar. Piensa que a lo mejor la jefa entenderá la situación, porque el abuelo de ella ha tenido también muchos problemas médicos últimamente. Sin esperar una respuesta, la jefa le agradece de antemano su colaboración. ¿Cómo responde él? 2. Un hombre mayor está sentado en un sofá en una librería ojeando un libro. Le parece muy interesante y lo quiere comprar. Llega una mujer joven desconocida a sentarse al lado de él, y se saludan. La joven ve el libro que el hombre está leyendo, y le cuenta que vino a la librería precisamente para comprar ese libro. Le dice que por lo visto, él tiene el último ejemplar porque una empleada que la ayudó a buscar el libro no lo pudo encontrar en los estantes. Le pregunta al hombre si lo piensa comprar, o si sólo lo está mirando. ¿Cómo responde él? 3. Un hombre sale del banco y se da cuenta que le han robado el carro, que estaba estacionado en la calle. Mira a su alrededor y ve a una policía de la misma edad suya parada en la esquina. Él se le acerca para reportar el robo, pero se da cuenta que ella está ocupada escribiendo un reporte. ¿Qué le dice el hombre? 4. Un maestro de primer grado se encuentra por casualidad en el pasillo de la escuela primaria con la madre de una de sus estudiantes, quien tiene la misma edad suya. El comportamiento de la estudiante ha sido horrible últimamente, y el maestro ya tenía pensado ponerse en contacto con la madre para comentarle la situación. Ella le pregunta si su hija se está portando bien en la clase. ¿Cómo responde el maestro? 5. Un hombre joven está desayunando solo en un café. Después de terminar el desayuno, le pide la cuenta a la mesera, una mujer mayor. Ella se la trae, y él

MURIEL/OROZCO

200

31/5/12

14:24

Página 200

SCOTT LAMANNA

pregunta si puede pagar con tarjeta de crédito, porque no tiene dinero en efectivo. La mesera le dice que no, que no reciben ninguna clase de tarjeta ahí. ¿Qué le dice el joven? 6. Una jefa entra a la oficina de uno de sus empleados, quien es de la misma edad suya. Ella le informa que todos tendrán que trabajar este sábado para poder terminar un proyecto a tiempo. El empleado ya tiene planes de salir el viernes en la noche para la ciudad donde vive su papá, porque él tuvo un paro cardiaco ayer y está en el hospital recuperándose. Cree que probablemente la jefa comprenderá la situación, porque hace poco el papá de ella también tuvo un paro cardiaco. Sin esperar una respuesta, la jefa le agradece de antemano su colaboración. ¿Cómo responde él? 7. Un estudiante universitario se pone a charlar con una compañera de la clase de física después de clase. Se conocen pero no son amigos. Empiezan a hablar del examen que va a haber la próxima semana. La compañera le pregunta si él le puede prestar su cuaderno para poder fotocopiar los apuntes. Dice que se lo devolverá para la próxima clase. El estudiante ya tiene pensado empezar a estudiar para el examen esta misma noche. ¿Cómo responde a la pregunta de su compañera? 8. Un hombre joven y una mujer mayor que son maestros de historia en escuelas secundarias diferentes se encuentran por casualidad en una tienda. Se conocen porque hay reuniones de vez en cuando entre todos los maestros de historia del distrito escolar, pero no son amigos. El hombre se siente avergonzado cuando la ve porque a los estudiantes de él no les fue muy bien en el examen estatal de fin de año, y teme que la mujer toque el tema. Efectivamente, ella le pregunta cómo les fue a sus estudiantes, y le cuenta que a los de ella les fue súper bien. ¿Cómo responde el maestro? 9. Una mesera joven en una pizzería está atendiendo a un hombre mayor. Al hombre realmente no le ha gustado mucho la comida, pero como la mesera lo ha atendido muy bien, decide mencionar el excelente servicio si ella le pregunta por su experiencia. Efectivamente, después de que él termina, la mesera le pregunta cómo le pareció la pizza. ¿Cómo responde él a la pregunta? 10. Un hombre que trabaja en finca raíz se encuentra con su hija en la cocina a la hora del desayuno. La hija le recuerda que esta noche tiene una presentación de baile en la escuela. Al padre se le había olvidado por completo, y ya no podrá ir porque tiene una cita para mostrarle una casa a un cliente a la misma hora. La hija le pregunta si va a estar en la presentación. ¿Cómo responde su padre? 11. Una maestra de quinto grado está escribiendo en la pizarra antes de empezar el día de clases. En ese momento, el primer estudiante en llegar ese día entra al salón, y se le acerca. El estudiante tiene las manos escondidas detrás de la espalda, y como es el cumpleaños de la maestra, ella sospecha que le tiene un

MURIEL/OROZCO

31/5/12

14:24

Página 201

SECOND PERSON SINGULAR PRONOUN USE

201

regalo. La maestra sonríe y le pregunta qué lleva en las manos. El estudiante se siente avergonzado porque su familia no tiene mucho dinero, y por eso le tocó regalarle algo muy barato. ¿Cómo responde él a la pregunta de la maestra? 12. Una médica joven está con uno de sus pacientes mayores en el consultorio alrededor del mediodía. Después de examinarlo, la médica le dice que tiene la presión extremadamente alta y que debe ir a la sala de emergencias inmediatamente. El paciente trabaja el segundo turno, y no quiere faltar al trabajo porque le han dicho que si vuelve a faltar por cualquier razón, lo van a echar. ¿Cómo responde a la sugerencia de la médica? 13. Una mujer recoge a su hijo de la escuela en su carro. Él no le quiere hablar mucho, porque resulta que la maestra lo castigó esta mañana por haber insultado a un compañero de clase, y él no quiere que su madre se dé cuenta. Ella le dice que ya sabe qué pasó porque la maestra la llamó y le contó todo. ¿Cómo responde el hijo? 14. Una mujer mayor está comiendo sola en un restaurante italiano. Después de terminar la comida, llama al mesero, un hombre joven, y le pide que le traiga helado de postre. Al mesero le acaban de informar en la cocina que ya se acabó el helado. ¿Cómo responde a su cliente? 15. Un hombre tiene que llamar a su esposa desde el trabajo para decirle que necesita quedarse en la oficina hasta tarde. Tiene que terminar un proyecto para entregárselo hoy mismo a un cliente. El hombre cree que seguramente su esposa se va a enojar porque ya es la tercera vez que él ha tenido que quedarse tarde esta semana, y además, a esta hora ella ya habrá empezado a preparar la comida. Marca el número telefónico y contesta su esposa. ¿Qué le dice él? 16. Un hombre joven está sentado en un banco en el parque. Llega una mujer mayor desconocida a sentarse al lado de él, y se saludan. El hombre tiene un celular en la mano, pero no lo está usando porque ya no le quedan minutos para este mes. La mujer le pregunta si él le puede prestar el celular porque quiere llamar a su esposo para que venga a recogerla, y no tiene ni celular ni monedas para usar un teléfono público. ¿Cómo responde el hombre? 17. Un hombre mayor se encuentra en el centro comercial con una mujer joven que es su amiga. La joven le pregunta si le gustaría ir a un festival de música mañana en la tarde. El hombre realmente preferiría asistir a un evento cultural que va a haber en el museo, y decide sugerirle a su amiga que hagan eso en vez de ir al festival. ¿Qué le dice él a ella? 18. Una mujer joven y un hombre mayor que son directores de escuelas secundarias diferentes se encuentran por casualidad en la calle durante el verano. Se conocen porque hay reuniones de vez en cuando entre todos los directores del distrito escolar, pero no son amigos. El hombre está frustrado en este momento porque tiene cinco puestos vacantes en su escuela, y no ha podido encontrar

MURIEL/OROZCO

202

31/5/12

14:24

Página 202

SCOTT LAMANNA

maestros calificados para llenarlos. La mujer le pregunta si ya tiene todo listo para empezar el nuevo año académico a finales de agosto, y le cuenta que a ella sólo le faltan unos pequeños detalles para estar completamente preparada. ¿Cómo responde él? 19. Un maestro de tercer grado está sentado detrás de su escritorio calificando tareas mientras sus estudiantes toman una prueba de matemáticas. Se le acerca una de sus estudiantes, y él levanta la mirada. Ella le dice que uno de sus compañeros le acaba de pegar. El maestro duda que sea verdad, porque la niña ya le ha mentido varias veces antes. ¿Cómo responde a la queja de su estudiante? 20. Un jefe entra a la oficina de una de sus empleadas, quien es de la misma edad suya. Le informa que un cliente no quedó satisfecho con el trabajo que ella le entregó, y que por eso tendrá que volver a hacerlo. La empleada reconoce que no hizo el trabajo con excelencia. Le dice al jefe que su esposo le había pedido el divorcio el día anterior de la entrega, y obviamente eso la había afectado negativamente. ¿Cómo responde él? 21. Un hombre está esperando el bus en el paradero. Llega una mujer desconocida de la misma edad suya a esperar el bus ahí también. Está lloviendo, y ella no tiene paraguas. El hombre tiene uno lo suficientemente grande para los dos, pero se da cuenta que la mujer ha estado bebiendo, y él se siente muy incómodo con gente borracha. La mujer le pregunta si se puede cubrir con su paraguas. ¿Cómo responde él? 22. Un mesero mayor está atendiendo a una mujer joven en un restaurante de comida típica. La joven pide una coca-cola, pero cuando el mesero se la trae, ella dice que pidió agua mineral, y que quiere hablar con el jefe para quejarse del servicio. ¿Cómo responde el mesero? 23. Un policía está caminando por un parque. De repente ve a una mujer de la misma edad suya en la distancia que parece estar borracha, porque está perdiendo el equilibrio. Camina rápidamente hacia ella y la mujer se voltea a mirarlo. Le pregunta por qué la está persiguiendo y le dice que no está haciendo nada malo. ¿Cómo responde el policía? 24. Una maestra de segundo grado se encuentra en la calle con el padre de una de sus estudiantes, quien es de la misma edad suya. Después de saludarse, el padre le pregunta cómo le va a su hija en la escuela. La maestra le cuenta que ella se duerme en clase casi todos los días. El padre se avergüenza porque sabe que es por la falta de disciplina de él y su esposa que la niña se acuesta tarde muchas veces. La maestra le pregunta si su hija está durmiendo lo suficiente, y si hay algo que ella puede hacer para ayudar la situación. ¿Cómo responde él? 25. Un médico joven está en su consultorio con una de sus pacientes mayores que tiene cáncer. El médico le dice a ella que necesita quimioterapia. La paciente le contesta que le han dicho que en muchos casos el tratamiento es peor

MURIEL/OROZCO

31/5/12

14:24

Página 203

SECOND PERSON SINGULAR PRONOUN USE

203

que la enfermedad, y por eso ella ha decidido no recibir ningún tratamiento. El médico tiene sentimientos encontrados porque quiere respetar los deseos de su paciente, pero sospecha que ella no ha considerado las consecuencias de su decisión. ¿Cómo responde a la preocupación de su paciente? 26. Una empleada entra a la oficina de su jefe, quien es menor que ella. En este momento el jefe se siente bastante estresado, porque hay que entregar varios proyectos a diferentes clientes muy pronto. La empleada le pregunta si sería posible salir un poco más temprano del trabajo durante las próximas dos semanas porque se lastimó el tobillo y necesita ir a sesiones de terapia física. ¿Cómo responde él? 27. Un hombre joven se encuentra en la misa con una mujer mayor que es su amiga. El joven tiene planes para ir a jugar fútbol con sus amigos esta tarde. Su amiga no sabe eso y le propone que vayan juntos a una exposición de arte. Se nota que ella está muy entusiasmada de asistir al evento, pero el hombre realmente preferiría pasar el tiempo con sus otros amigos. ¿Cómo responde a la sugerencia de su amiga?

MURIEL/OROZCO

31/5/12

14:24

Página 204

MURIEL/OROZCO

31/5/12

14:24

Página 205

THE EXPRESSION OF NOMINAL POSSESSION I N T H E S PA N I S H O F C O L O M B I A N S IN NEW YORK CITY* RAFAEL OROZCO

1. Introduction Possession is expressed in Spanish by means of various synthetic and analytical morphosyntactic devices (Orozco 2004: 190-197). In expressing nominal possession, as described in Orozco (2010a: 196), Spanish speakers use a tripartite linguistic variable that consists of possessive adjectives, definite articles, and possessive periphrases, illustrated in (1-3), respectively. (1) Possessive adjectives (PA): María se quedó en su casa. (2) Definite articles (DA): María se quedó en la casa. (3) Possessive Periphrases (PP): María se quedó en la casa de ella.

All three variants of this linguistic variable often occur interchangeably wherever possessives appear within a sentence. Possessive adjectives, which are used universally, have consistently monopolized the attention of Spanish grammar books and textbooks. The use of definite articles (example 2 above) as possessive markers, despite being less widespread crosslinguistically, is a common occurrence in those languages that have these determiners. For instance, possession is expressed by means of definite articles in Catalan, French, German, Italian and Portuguese. Their well-established use with possessive value in Castilian has been attested in 16th century prose (Keniston 1937: 235). However, as Orozco (2010a: 197) indicates, the use of definite articles as possessives has not always been discussed or mentioned in the Spanish grammar books (cf. Alarcos Llorach 1973, Bello/Cuervo 1941, Menéndez Pidal 1968/1904, inter alia). However, the use of definite articles as possessive markers is addressed in works of a * I am grateful to Armin Schwegler, Agnes Ragone, Lotfi Sahayi, Kareen Gervasi, and Richard J. File-Muriel for their helpful comments. I assume sole responsibility for all remaining infelicities.

MURIEL/OROZCO

31/5/12

14:24

206

Página 206

RAFAEL OROZCO

more descriptive nature. For example, Leonetti (1999: 808) asserts that Spanish shows a clear preference for the use of definite articles in contexts calling for the use of a possessive. Picallo/Rigau (1999: 1009), who analyze in detail the various ways to express possession in Castilian, describe the use of definite articles to mark possession with nouns that denote family members and relatives. Alcina/Blecua (2001/1975: 566), Fernández (1951: 291), and Picallo/Rigau (1999: 1006 ff.) concur that definite articles alternate with possessive adjectives especially with certain types of nouns including body parts, garments, actions, and mental faculties. Gili y Gaya (1964: 240) indicates that when definite articles mark possession, they are frequently accompanied by a reflexive pronoun as in (4), taken from our corpus. Criado de Val (1966: 101) and Keniston (1937: 235) note that definite articles often compete with possessive adjectives while de Bruyne (2004/1995: 181) indicates that certain uses of possessive adjectives over definite articles, even in non-contact situations, are considered anglicisms. (4) A ese man se le perdieron loh tacoh. [NM0107].

The newest of the three variants, called possessive periphrasis by Orozco (2004: 193), constitutes a modern development. This possessive marker occurs in all varieties of Spanish and in Portuguese. It is also similar to the preferred way of expressing possession in Italian, which involves a determiner and a possessive pronoun. In the first and second person singular, it has the following morphosyntactic configuration. [Definite article + Possessed Noun + Possessive Pronoun] The example in (5) illustrates this usage: (5) El hermano mío1

For the remaining grammatical persons, the periphrasis obtains as follows. (6) [Definite article + Possessed Noun + de + Subject Pronoun/NP]

This results in constructions such as (7) El hermano de él.

1 Although el hermano suyo is morphosyntactically possible, it rarely occurs and was not attested in the corpus explored here.

MURIEL/OROZCO

31/5/12

14:24

Página 207

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

207

The origin of this analytic paradigm is imputed to the substitution of the circumlocution de ustedes for the possessive vuestro (Orozco 2009b: 38). It appears that the use of the periphrastic innovation progressively generalized in vernacular speech as it spread to all grammatical persons (Gili y Gaya 1964: 241, Penny 2008: 169), and became the most accurate way to express nominal possession in Castilian. Yet, as with definite articles, possessive periphrases have not received much attention in the Spanish grammar books. Kany (1969: 68-70) discusses the use of the possessive periphrasis in all varieties of Spanish and attributes its origin to the need for clarification imposed by the several possible meanings of the possessive adjective su. Moreover, Criado de Val (1966: 100), Fernández (1951: 230 ff.), and Penny (2008: 169) assert that su was plagued by a semantic overload, which contributed to its ambiguity and, consequently, led to the emergence of genitive phrases. Gili y Gaya (1964: 241) also points out the ambiguity posed by the possessive su. He concurs with Kany (1969: 68) and Keniston (1937: 244) in emphasizing that phrases consisting of de plus a personal pronoun were already used in the 16th century to remedy the ambiguity of the possessive adjectives. Alcina/Blecua (2001/1975: 938-939) and Penny (2002: 141) also report the usage of the possessive periphrases categorizing them as prepositional complements. Others who report the usage of possessive periphrases include Butt/Benjamin (2004: 99) and de Bruyne (2004/1995: 180). In contrast, their use is mentioned only marginally by Bosque/Demonte (1999) in their Gramática descriptiva de la lengua española. The present study, one of the first few explorations of this linguistic variable, constitutes a follow-up to the analysis of nominal possession in Colombian Costeño Spanish by Orozco (2010a). As in that study, I will explore the alternation among the three variants of the linguistic variable under study, i.e., possessive adjectives, definite articles marking nominal possession, and possessive periphrases that is illustrated in (7), (8), and (9). (7) Yo no me voy a quedar en mi casa/ la casa/ la casa mía na más esperando. (8) Elloh viven con los pies/ sus pies / los pies de ellos en la tierra, son unidoh. (9) Yo sé que la hermana de él/ la hermana/ su hermana tiene esa plata que se ganó.

This linguistic variable remains virtually unexplored not just in Colombia, but also throughout the Hispanic world. While regional variation in Colombian Spanish has been widely analyzed (Orozco 2004: 50), sociolinguistic variationist studies have gradually appeared in recent years (Orozco 2009a: 96). Still, variationist studies on Costeño Spanish are scarce (Orozco 2010a: 196), and, as stated before (p.16, 170), so are such studies on Colombian Spanish in the US. This paper aims

MURIEL/OROZCO

31/5/12

208

14:24

Página 208

RAFAEL OROZCO

to contribute to the emerging body of sociolinguistic literature on Colombian Costeño Spanish, in general, and specifically on Colombian Spanish in the US.

2. Methodology 2.1. THE SPEECH COMMUNITY AND THE DATA SET Spanish speakers started settling in Manhattan in 1625, a year after New York was founded, so they have been part of that city virtually for as long as it has existed. Today, with more than two million speakers (roughly 28% of the city’s population), Spanish is the de-facto second language in NYC. The first Colombian community in New York, also the first such community in the US, dates back to the late 1910s when a few hundred Colombians settled in Jackson Heights, Queens (Sturner 1995, Zentella 1997a). By 1940, there were almost 2000 Colombians in NYC (Orlov/Ueda 1980: 213). Zentella (1997a: 170) notes that during the 1970s this community continued to expand. Most of the Colombians who settled there had belonged to the middle class in their home country. In the mid 1980s, middle-class Colombians were the majority of residents of Jackson Heights, which several decades earlier had become the hotbed of the New York Colombian community. According to Sturner (1995: 329), in 1994 there were 86,000 legal Colombian residents in New York. Moreover, at the turn of the 21st century, the greater NYC area continued to host the largest concentration of Colombians in the US (Orozco 2007a: 312). This community has a captivating sociolinguistic situation since its language is simultaneously in direct contact with English and with many other Spanish dialects. The direct contact between Colombian Spanish and English, a situation of relatively recent inception, provides a singular opportunity for short-term diachronic analysis, which may provide answers to basic questions also involving longer time spans (Weinreich 1967/1953: 104). The contact between Colombian Spanish and other varieties of Spanish is no less intriguing. Colombian expatriates continue to use Colombian Spanish at home and to interact with their compatriots. As they become acclimated to their new environment, New York Colombians find a common ground in communicating with other Hispanics and incorporate into their linguistic repertoire the Spanish of NYC which, as noted by Zentella (1990a, 1997a), is saturated by Caribbean Spanish. Orozco (2004: 58) points out four different sources of leveling affecting Colombian Spanish in NYC: the coexistence of people from all over Colombia; the heavy influence of Caribbean Spanish compounded by the fact that many New York Colombians come from the Caribbean coast; the influence of NYC Spanish on Colombian

MURIEL/OROZCO

31/5/12

14:24

Página 209

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

209

Spanish due in large part to frequent visits from New York Colombians; and the influence, both direct and indirect, of English. The data set explored in this study was culled from the Corpus del Español Colombiano en Nueva York (CEsCoNY), previously used in Orozco (2004, 2007a, 2007b and 2010b). This corpus consists of 20 hours of sociolinguistic conversations with Colombian residents of the greater New York City area, ten women and ten men. At the time of the data collection, their ages ranged from 16 to 70. Three of these speakers were teenagers, two were in their twenties, seven in their thirties, four in their forties, and four older than fifty. All of them were born in Colombia and had immigrated to the US at various ages. Two of them came to the US as young children, two arrived as teenagers, and sixteen immigrated after the age of twenty. Most of them came to the US from the Caribbean coast (17 from Barranquilla), and the rest from the Pacific coast. According to Colombian social class structure, they were from middle and lower class extraction prior to immigrating to the US. Their education and occupational status place them in the middle and working classes in the US. Most of these informants now have a socioeconomic status lower than what they had in Colombia. Occupationally, they fall into three categories. Five of them retained the white-collar status they had in Colombia. Six retained their blue-collar status, and nine traded their whitecollar status in Colombia for blue-collar status in the US. As has happened traditionally with Colombian immigrants in NYC, almost all of the informants first settled in Queens upon arrival. Their social networks are typical of most Colombian immigrants, initially associating mostly with other Colombians, and gradually expanding their social networks to include people from other Hispanic backgrounds (Orozco 2007a: 312). Younger New York Colombians also associate with English-dominant peers and often communicate in English with those of Latino backgrounds. Except for the youngest speakers, the rest are Spanish-dominant even if they extensively use English in diglossic patterns occupationally. Whereas older New York Colombians speak English with a strong Spanish accent, most of the younger ones do not (Orozco 2004: 80). What Zentella (1997b: 85, 293) describes for most immigrant populations also obtains with Colombians. That is, the only people who ordinarily remain monolingual in Spanish after ten or more years in the US are those who migrate in middle age or later, have little access to English-speaking contacts, and never achieve regular employment.

2.2. RESEARCH QUESTIONS AND HYPOTHESES I designed this study to allow an easily replicable procedure and to find answers to the following main research questions based on previous empirical research in

MURIEL/OROZCO

31/5/12

14:24

Página 210

210

RAFAEL OROZCO

this community (Orozco 2004, 2007a, 2007b) and on the linguistic variable under study (Orozco 2010a): • How are the three variants of the linguistic variable explored here (possessive adjectives, definite articles, and possessive periphrases) distributed in the Spanish of New York Colombians, and how does their distribution compare to that found in Barranquilla? • What internal constraints condition the expression of nominal possession in this community and what are their effects? Are these constraints and their effects the same as those that condition nominal possession in Costeño Spanish? If not, why not? • What external constraints condition the expression of nominal possession in the Spanish of New York Colombians? How do these constraints compare to those that condition this linguistic variable in Colombia, and to those conditioning other linguistic variables? With my first research question, I tested the hypothesis that possessive adjectives would occur more frequently in NYC than in Barranquilla due to convergence with English (this variant is the closest to the English preferred expression of possession). With my second question, I tested the theory of interdialectal parallelism (Guy 2000), according to which, the factors constraining language variation and change are consistent within different segments of a speech community. Concurrently, I tested the hypothesis that the existing variation in the expression of possession results largely from internal rather than external constraints. I based this hypothesis on the findings of sociolinguistic explorations in different parts of the Hispanic world (Blas Arroyo 2007, 2008; Geeslin and Guijarro-Fuentes 2007; Otheguy/Zentella 2007; Otheguy/Zentella/Livert 2007, among others), as well as variationist studies of the expressions of futurity (Orozco 2005, 2007a, 2007b), pronominal expression (Orozco/Guy 2008, Orozco 2009a), and possession (Orozco 2010a) in Costeño Spanish that show variation to be mainly internally-motivated. My last research question is mainly intended to document the status of the expression of possession in this community since this linguistic variable remains virtually unexplored.

2.3. CONSTRAINTS EXAMINED To answer the above questions and test my hypotheses, I explored the effects of ten linguistic and seven social constraints. I based my choice of constraints on Orozco’s (2010a) analysis of the possessive in Colombian Spanish. I explored

MURIEL/OROZCO

31/5/12

14:24

Página 211

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

211

the effects of internal constraints that operate at various morphosyntactic levels, so I tested constraints that operate at the whole clause level, at the subject level, and at the NP level – either within the subject or within an object. I coded the data for each of the following internal variables used in the statistical analysis. • Length of the Clause Containing the Possessive: one to five words long, six to nine words long, longer than nine words. • Type of Statement: declarative, conditional, interrogative, negative • Type and Grammatical Number of Subject: singular overt, plural overt, singular null, plural null • Location of the Possessive: subject, direct object, indirect object, oblique object • Presence of Adjectives in the Genitive NP: Adjective present pre-nominally, Adjective present post-nominally, No adjective present • Grammatical Number and Gender of the Possessee: singular feminine, singular masculine, plural feminine, plural masculine • Semantic Category of the Possessed Noun: body part, parent, non-parent relative or family member (grandparents, aunts, uncles, cousins, etc), nonrelative human, not human (animals, things and all other nouns not included in the previous four categories). • Length of Noun Marked for Possession: one to two syllables, three to five syllables, six or more syllables. • Distance in Words Between Referent and Possessive: no overt referent, one to five words between the referent and the possessive, six to ten words between referent and possessive, more than ten words between referent and possessive • Grammatical Person, Number, and Animacy of the Possessor: – Four factors refer to singular possessors: first and second grammatical person, third person human, and third person nonhuman. – Four factors refer to plural possessors: first and second grammatical person, third person human, and third person nonhuman. Exploring the significance of social forces is an important part of this study which, according to Labov’s (1972: 252) perspective, was conceived with the goal of obtaining a better understanding of the existing variation in the expression of nominal possession. Determining the effects of social constraints on language is especially important in immigrant communities since immigration often entails abrupt changes in the immigrants’ socioeconomic status and family roles. To explore the effect of external constraints, I analyzed the effects of four vari-

MURIEL/OROZCO

212

31/5/12

14:24

Página 212

RAFAEL OROZCO

ables traditionally used in variationist analyses as well as in recent sociolinguistic studies of Costeño Spanish dealing with the expression of futurity (Orozco 2007b) and variable subject pronoun expression (Orozco/Guy 2008). Thus, I tested the following external constraints and factors: • Speaker’s sex: male vs. female • Speaker’s age: teenager, twenties, thirties, forties, fifties, older than fiftynine • Socioeconomic status2: blue collar in Colombia and the US, white collar in Colombia and US, white collar in Colombia but blue collar in the US • Education: incomplete high school education, completed high school, attended college in Colombia, graduate school in the US (speakers under 25 were not included.) Following my studies of the expression of futurity (Orozco 2004, 2007b) and variable subject pronoun usage (2010b), I tested the effects of the following three constraints only applicable to this community that are often explored in studies involving immigrant speakers (cf. Silva-Corvalán 1994, Otheguy/Zentella 2007). • Length of Residence (LOR): less than five years, five to ten years, more than ten years in the US. • Age at the Time of Arrival: pre-teenager, teenager, adult. • Linguistic repertoire: biliterate bilingual, bilingual with limited literacy in English, bilingual with limited literacy in Spanish.

2.4. THE ENVELOPE OF VARIATION AND THE ANALYSIS The envelope of variation used here follows the parameters set forth by Orozco (2010a). That is, I included in the data set only those clauses containing nominal possessives that fall inside what sociolinguists call the envelope of variation (Bayley 2004: 124 ff.). I defined the envelope of variation for this analysis as illustrated in (7) – (9), so a clause was included in the database only if all three variants of the nominal possessive (possessive adjectives, definite articles, and 2

Since a sizeable portion of the New York Colombian community belongs to the working class (see §2.1), I explored socioeconomic status by using factors that help account for the fact that immigrants often hold jobs below their occupational status in their countries of origin. Thus, I included “white-collar in Colombia but blue-collar in the US” as a factor.

MURIEL/OROZCO

31/5/12

14:24

Página 213

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

213

possessive periphrases) were likely to occur. That is, I used the criterion that all tokens would constitute different ways of saying the same thing. In so doing, I meticulously scrutinized those clauses in which definite articles mark possession due to the different syntactic roles played by this variant. Consequently, I only included in the database those statements in which definite articles categorically function as possessives as occurs with the definite article la in (10). (10) ... pero ahora que ha venido y que la trae la mamá y o la lleva.

At the same time, I excluded from analysis all occurrences of definite articles with meanings other than possession. Thus, statements such as (11) were excluded from the data set because the definite article that occurs in the phrase la ehcuela, by not being a possessive marker, falls outside the envelope of variation. (11) …noh criamoh juntah, y fuimoh a la ehcuela juntah.

Using a total of 1,210 tokens, I conducted a series of parallel statistical regression analyses for each variant using Goldvarb X as my statistical tool. My presentation of the results is divided into three parts. I first present the distribution of the three variants. I then discuss the internal constraints whose correlation with a specific variant is statistically significant. I address the significant effects of external constraints thereafter.

3. Distribution of variants The frequency distribution, the answer to our first research question, is presented in Table 1 together with the distribution of variants found in Barranquilla (Orozco 2010a). Definite articles take the largest share of the distribution with a frequency of 46.6%, but their frequency remains virtually the same as in Barranquilla. Possessive adjectives appear slightly less often, taking the second largest share with a frequency of 41.3%. As with the distribution in Barranquilla, possessive periphrases hold the smallest portion of the distribution with a frequency of 12.1%. However, they occur almost twice as frequently as they do in Colombia. Compared to the distribution of forms for Barranquilla, the frequency for possessive adjectives shows a slight decrease in NYC while possessive periphrases register an increase. Contrary to the Barranquilla situation, in New York determiners hold the largest share of the distribution. This contradicts my hypothesis that direct contact with English would cause possessive adjectives to

MURIEL/OROZCO

31/5/12

14:24

214

Página 214

RAFAEL OROZCO

TABLE 1 Distribution of Variants Variant

Barranquilla

New York

Definite Articles (e.g. está en la casa)

585 (45.7%)

564 (46.6%)

Possessive Adjectives (e.g. está en su casa)

613 (47.8%)

500 (41.3%)

83 (6.4%)

146 (12.1%)

1281 (100%)

1210 (100%)

Possessive Periphrases (e.g. está en la casa de él) Total

occur more frequently in the diasporic setting. The more frequent occurrence of possessive periphrases among New York Colombians may support the argument that language contact situations favor analytical forms over synthetic ones. It is also congruent with the premise that changes in progress are accelerated in language contact situations. In general, there seems to be a change in the expression of possession, albeit still in its very early stages compared to the change in progress in the expression of futurity discussed by Orozco (2004, 2005, 2007a, and references therein). In what follows, I address the constraints which condition the three variants under study.

4. Constraints on the expression of possession The expression of possession is significantly conditioned by a complex interaction of internal and external constraints. The rankings and statistical ranges of the fourteen (eight linguistic and six social) constraints that condition this linguistic variable are presented in Table 2, partially answering our second and third research questions. Possessive adjectives are conditioned by ten constraints (four linguistic, six social), definite articles by eight (five linguistic, three social), and the periphrases by nine (seven linguistic, two social). These results also disprove the hypothesis that, as in Barranquilla, the expression of possession is mainly conditioned by linguistic constraints in New York, even if each one of the variants is under different conditioning pressures. Judging by the number of external constraints that condition each variant as well as by the aggregate range values, definite articles and possessive periphrases are under stronger internal conditioning (five and seven internal constraints respectively). In contrast, the external pressures are heavier on possessive adjectives, which are conditioned by six social and four linguistic constraints.

MURIEL/OROZCO

31/5/12

14:24

Página 215

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

215

TABLE 2 Constraint Rankings and Ranges Possessive Adjectives

Definite Articles

Possessive Periphrases

1.

Semantic Category

43

Semantic Category

36

Education

46

2.

Age

39

Age

32

Semantic Category

43

3.

Grammatical Person

36

Length of Residence

25

Length of Residence

42

4.

Education

32

Distance ref- possessive

21

Grammatical Person

30

5.

Length of Residence

29

Grammatical Person

20

Presence of Adjectives

29

6.

Socioeconomic Status

23

Education

20

Distance ref- poss

27

7.

Type of Subject

12

Type of Subject

17

Type of Subject

20

8.

Grammatical Gender

12

Grammatical Gender

7

Location of possessive

14

9.

Speaker’s Sex

11

Clause length

13

10.

Age of Arrival

11

4.1. INTERNAL CONSTRAINTS The eight internal constraints that significantly condition the expression of possession in New York are the same as those doing so in Barranquilla (Orozco 2010a). They include constraints affecting all three variants (semantic category of the possessed noun, grammatical person and number of the possessor, and type of subject), as well as others which only constrain either two (distance between referent and possessive, location of the possessive, and grammatical gender of possessee) or one of them (clause length and presence of adjectives in the genitive NP). Also as in Barranquilla, however, the only two linguistic constraints not affecting any of the variants were clause type and length of the possessed noun. Seen from another perspective, four internal constraints condition possessive adjectives, five definite articles, and seven possessive periphrases. In general, when one constraint conditions possessive adjectives as well as definite articles, the individual factor tendencies appear in opposition to each other. That is, for the four constraints affecting both of these variants, the results for possessive adjectives are largely a mirror image of those for articles. For the most part, the probability values show the same tendencies found in Barranquilla

MURIEL/OROZCO

216

31/5/12

14:24

Página 216

RAFAEL OROZCO

and discussed by Orozco (2010a). I shall now discuss the constraints that most significantly condition the expression of nominal possession comparing, whenever possible, the current results with those found in Barranquilla.

4.1.1. Semantic Category of the Possessed Noun This is the first of three internal constraints conditioning all three variants. The results (see Table 3) show that, as in Barranquilla, Semantic Category of the Possessed Noun has the greatest effect of all internal constraints on the variation in the expression of possession as measured by the range. At the same time, these results reflect some minor differences from what was found in Barranquilla. In general, the effects of this constraint on possessive adjectives and definite articles stand in opposition to each other. A pattern is clearly discernible with respect to possessive adjectives: Nouns naming parents (example 12) strongly favor possessive adjectives with a statistical weight of .66. Nouns naming non-parent relatives (example 12) and nonhumans, respectively, also favor this variant. In contrast, possessive adjectives are disfavored by nouns naming non-relative humans, with a statistical weight of .41, and strongly disfavored by nouns naming body parts with a value of .23. These tendencies, except for the values for nouns naming non-humans, are quite similar to those found in Barranquilla. (12) Lee, se lo dah a tu ehposa, tu mamá, o a alguien. [NF01009]

The statistical tendencies for definite articles also show a pattern, which contrary to what occurs in Barranquilla, is not simultaneous with that for possessive adjectives. Nouns naming body parts (example 13) strongly favor definite articles with a statistical weight of .82. Nouns naming non-relative humans also favor this variant with a value of .57, while nouns naming parents have a rather neutral effect. Contrariwise, definite articles are disfavored by nouns naming non-humans and non-parent relatives, respectively, both with a statistical weight of .46. (13) Ya Eduardo ni que venga aquí a asomarse la nariz porque no consigue trabajo. [NF02002]

These tendencies provide quantitative support to several statements about the use of definite articles as possessive markers. For instance, Gili y Gaya indicates that when definite articles mark possession, as in (12), they frequently occur with reflexives (1964: 240). The results are also congruent with observations by

MURIEL/OROZCO

31/5/12

14:24

Página 217

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

217

Fernández (1951: 291), Picallo/Rigau (1999: 1006 ff.), and Alcina/Blecua (2001/ 1975: 566) who point out that definite articles mark possession especially with nouns denoting body parts. However, the results do not validate Picallo/Rigau’s assertion (1999: 1009) that this variant is also used to mark possession with nouns that denote family members and relatives. Possessive periphrases are most strongly favored by nouns naming non-parent relatives (example 14) with a statistical weight of .57. The periphrases are also under the favorable effect of nouns naming non-humans and non-relative humans, respectively. Contrariwise, nouns naming parents disfavor this variant with a value of .24 and nouns naming body parts strongly disfavor it with a statistical weight of .14, the strongest disfavoring effect registered by a single internal factor on any of the three variants under study. (14) … posiblemente Dios quiera; porque si no, los hijoh de nosotroh van a pasá trabajo. [NM02005]

For the most part, the results for semantic category corroborate the findings in Barranquilla with one noticeable difference: the influence of nouns naming non-humans on possessive adjectives and definite articles. Furthermore, the dominant tendencies show that in both communities each variant is used preferentially with a specific type of noun, i.e., possessive adjectives with nouns naming parents, definite articles with nouns naming body parts, and possessive periphrases with nouns naming humans other than parents. Despite the similarities between these results and the Barranquilla findings, the differences, mainly limited to different statistical weights, appear to result from contact with English. For instance, there is a significant increase in the use of possessive adjectives with nouns naming body parts and non-relative humans, respectively, both at the expense of definite articles.

4.1.2. Grammatical Person and Number of the Possessor3 As this section will show, this constitutes another robust constraint on the nominal possessive (the second strongest internal constraint), conditioning this lin-

3 I originally used this constraint to also test the effect of animacy of the possessor as indicated in §2.3, obtaining different results from those found in Barranquilla. Since animacy does not condition the possessive in NYC, I merged those factors with similar tendencies and conducted all subsequent analyses using four factors: (1) first and second person singular, (2) third person singular, (3) first person plural, (4) second and third person plural.

MURIEL/OROZCO

31/5/12

14:24

Página 218

218

RAFAEL OROZCO

TABLE 3 Effects of Semantic Category, Grammatical Person, and Type of Subject Factor

PA

N

%

DA

N

%

PP

N

%

.24 .57 .56 .55 .14

5/75 34/217 97/766 9/63 1/89

7% 16% 13% 14% 1%

Semantic Category of Possessed Noun Parent Non-parent relative Non-Human Non-relative human Body Part

.66 .53 .52 .41 .23

Range

43

39/75 91/217 333/766 24/63 13/89

52% 42% 43% 38% 15%

.49 .46 .46 .57 .82

31/75 92/217 336/766 30/63 75/89

41% 42% 44% 48% 84%

36

43

Grammatical Person and Number of the Possessor 1st and 2nd Person Singular 3rd Person Singular 2nd and 3rd Person Plural 1st Person Plural Range

.66 .39

249/439 144/467

57% 31%

.40 .60

152/439 276/467

34% 59%

.37 .54

38/439 47/467

9% 10%

.54 .30

76/178 31/126

43% 25%

.44 .57

83/178 53/126

46% 42%

.60 .67

19/178 42/126

11% 33%

102/741 24/399

14% 6%

I=.40 500/1210 41% I=.47 564/1210 47% I=.07 146/1210

12%

36

20

30

Type of Subject Overt Null

.54 .42 Range

I=input

12

327/741 143/399

44% 36%

.44 .61 17

312/741 232/399

42% 58%

.57 .37 20

guistic variable even more strongly than in Barranquilla. Unlike what happens there, where this constraint does not condition possessive periphrases, in NYC it does condition all three variants. However, when the tendencies are compared, further similarities between the two populations emerge. The results (Table 3) reveal that grammatical number does not condition possessors either. First and second person singular possessors (mi, tu) with a statistical weight of .66 favor possessive adjectives at the expense of both definite articles (.40) and possessive periphrases (.37). Third person singular possessors (example 13) favor definite articles (.60) and possessive periphrases (.54), respectively, to the detriment of possessive adjectives. First person plural possessors favor posses-

MURIEL/OROZCO

31/5/12

14:24

Página 219

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

219

sive periphrases (examples 14 and 15) and definite articles, with statistical values of .67 and .57 respectively, while disfavoring possessive adjectives. Finally, second and third person plural possessors favor possessive periphrases, with a statistical weight of .60, and, to a lesser extent, possessive adjectives while exerting a disfavoring effect on the articles. The favorable effect of first (example 15) and second person singular (example 12) possessors on possessive adjectives shows clearly that this is the variant of choice for the first two grammatical persons. These tendencies may arguably result from the unambiguous nature of the possessive adjectives involved when compared to those in all other grammatical persons. The disfavoring effect of third person singular possessors on possessive adjectives may indicate that speakers try to avoid using the possessive adjective su due to its strong ambiguity as discussed by Fernández (1951: 230 ff.), Criado de Val (1966: 100), and Penny (2008: 169), inter alios. Thus su has continued to lose ground to definite articles and possessive periphrases. The results (Table 3) also show that the possessive periphrasis is the most favored possessive marker when plurality is involved. This variant is favored by plural possessors in the first person, with a statistical weight of .67, and by those in the second and third persons (.60). New York Colombians may prefer to use possessive periphrases with possessors in all plural grammatical persons as a consequence of the natural evolution of this analytical paradigm. According to Gili y Gaya (1964: 241) and Penny (2008: 169) the possessive periphrasis originated as the substitution of the possessive adjective vuestro for the genitive construction de ustedes. As the usage of these constructions became widespread, it seems to have overspilled to include all plural possessors. Thus, it seems reasonable to assume that New York Colombians may also prefer to use possessive periphrases with first and third person plural possessors. (15) ...mientrah que la política que tenemoh hoy en día siga así, el paih de nosotroh no progresará. Loh senadoreh, to esoh grandeh se pierden una millonada que la consignan a nombre de loh familiareh de elloh. Investigan y eso queda en nada. [NM06056]

This is illustrated in (15). After using a possessive periphrasis in the first person plural, the speaker does so in the third person plural as well. Thus, the use of the first person plural possessive adjective nuestro/nuestra appears to be inhibited as a result of the ongoing restructuring of the Spanish pronominal and possessive systems. Moreover, as Orozco (2010a: 207) indicates, the use of nuestro/ nuestra is now relegated mainly to formal and religious or patriotic references as in El Padre Nuestro.

MURIEL/OROZCO

31/5/12

14:24

220

Página 220

RAFAEL OROZCO

In general, we learn that the expression of possession is not constrained by animacy or grammatical number of the possessor. The tendencies for first person possessors (singular and plural) clearly go in opposite directions according to grammatical number. Possessive adjectives are clearly the possessive form of choice for first person singular possessors in both Colombia and New York. However, the tendencies for the remaining grammatical persons show variation in both places, arguably as a result of the linguistic contact found in New York.

4.1.3. Type of Subject Besides testing the significance of type of subject, as indicated in §2.3, I used this constraint to test for grammatical number. Based on their similar tendencies in a preliminary analysis, as with the Barranquilla data, I merged singular and plural null subjects into one factor and singular and plural overt subjects into another. I conducted all subsequent analyses exploring overt versus null subjects. The results (Table 3) show that, as in Barranquilla, this constraint conditions the occurrence of all three variants with the same general statistical tendencies. That is, overt subjects promote both possessive adjectives (.54) and the periphrases (.57) while disfavoring definite articles. Conversely, null subjects promote the occurrence of definite articles, with a statistical weight of .61, and disfavor both possessive adjectives and possessive periphrases. An explanation as to why overt subjects promote possessive adjectives may be that, as Picallo/Rigau (1999: 1005) assert, possessive adjectives are favored in situations where a speaker chooses to add emphasis. That is, in contexts such as the one presented in (16) a possessive adjective helps the speaker add emphasis. The favorable effect of overt subjects on possessive adjectives is also congruent with the frequent appearance of the indefinite pronoun uno as a grammatical subject also illustrated in (16). In Spanish uno occurs more often than its counterpart (one) does in English. When uno appears as a possessor, it usually prompts the appearance of a possessive adjective. Apparently, speakers avoid using a definite article to indicate possession when the possessor is an indefinite pronoun. (16) Si uno quiere, si uno planea quedarse en ehte paíh y vivir aquí y ganar uno, o sea, vivir en ehte paíh y trabajar uno toda su vida, necesita uno legalizarse. [NM10011]

The favorable effect of null subjects on definite articles marking possession may result from the frequent occurrence of clitics in clauses with null subjects.

MURIEL/OROZCO

31/5/12

14:24

Página 221

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

221

According to Leonetti (1999: 808), who concurs with Bello (1847: §955) and the Real Academia de la Lengua Española (1973: §3.10.9), “[a] menudo el [adjetivo] posesivo es rechazado si en la construcción aparece un pronombre clítico.” The presence of a clitic, which often helps to identify the subject of a sentence as well as the possessor (Leonetti 1999: 809), appears to motivate the preference for definite articles in clauses with null subjects. Accordingly, in (17), the presence of the clitic le arguably triggers the use of the definite article la to mark possession. (17) ... entonces le caen, le amarran, o le agarran la mano y te ponen que lo fumes. [NF08031]

A comparison of the Goldvarb values for the effect of type of subject on the occurrence of possessive adjectives in Barranquilla and New York reveals close similarities not only in the tendencies that obtain but also in the statistical weights and ranges. In general, we ascertain from these results how type of subject affects each one of the variants under study. We also learn that, while the resulting tendencies are the same for both populations, it can be argued that the impact of type of subject on the possessive has not been greatly affected by direct contact with English.

4.1.4. Distance between the Referent and the Possessive As described in §2.3, this constraint originally consisted of four factors. After preliminary analyses, as with the Barranquilla data, I merged (a) six to ten and (b) more than ten words between referent and possessive due to their similar statistical tendencies: Thus, I conducted all subsequent statistical tests using three factors: no overt referent, one to five words, and six or more words between the referent and the possessive. The resulting statistical tendencies, as well as each factor's frequency of occurence, closely resemble those found in Barranquilla. However, in the New York data, this constraint only reached statistical significance in the occurrence of definite articles and possessive periphrases. The results in Table 4 reveal two patterns: one with regard to definite articles marking possession, and the other regarding possessive periphrases. As in Barranquilla, a gradual increase in the distance between referent and possessive has a favorable effect on definite articles. The absence of a referent disfavors them with a statisitical weight of .36. A referent that appears less than six words away slightly favors them, and a referent located six or more words away (Ya Eduardo ni que venga aquí a asomarse la nariz porque no consigue…) favors the use of

MURIEL/OROZCO

222

31/5/12

14:24

Página 222

RAFAEL OROZCO

definite articles more strongly with a statistical weight of .57. The tendencies for definite articles, as illustrated in (18), may result from the fact that a single referent (ella) may refer to more than one possessive in a narrative sequence (el papá, la mamá, la abuela). (18) Ya ella sabe que no puede salirse de ahí con ninguno si no es el papá, la mamá, o la abuela. [NM01043]

The second pattern shows that an increase in the distance between the referent and the possessive has a disfavoring effect on the periphrases. The absence of an overt referent strongly favors them (Los hijoh de nosotroh van a pasá trabajo.) with a probability of .72. The presence of a referent, regardless of distance, disfavors the occurrence of possessive periphrases. The lack of a clear pattern for the periphrastic possessive in both Barranquilla and New York may be explained in terms of Penny’s contention that linguistic innovations are often characterized by instability as they become established in the host linguistic system (Penny 2000). In general, the patterns for definite articles and possessive adjectives reveal the same tendencies in Colombia and New York. The same occurs with the lack of a pattern for possessive periphrases. The tendencies for the effect of this constraint run in the same direction as those registered in Barranquilla. For instance, the value for articles (.36) is the same for both populations, which seems to suggest that contact with English does not affect the absence of an overt referent on this variant. The remarkable similarity in the statistical values may indicate that the sociolinguistic environment of NYC has not had a significant impact on this constraint.

4.1.5. Location of the Possessive In exploring this constraint, as with the Barranquilla analysis, I initially used the four factors listed in §2.3. Due to their similar statistical tendencies, I merged those categories pertaining to an object (direct object, indirect object, and object of a preposition) into a single factor. Thus, I conducted all subsequent analyses exploring the effects of possessives in the subject or the object. As shown in Table 4, this constraint significantly conditions the occurrence of possessive adjectives and possessive periphrases but not that of definite articles. Possessives located in the subject (… el paih de nosotroh no progresará.) favor possessive periphrases with a statistical weight of .61 while possessives located elsewhere inhibit them. Contrariwise, possessives in a grammatical object (.53)

MURIEL/OROZCO

31/5/12

14:24

Página 223

223

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

favor possessive adjectives and disfavor the other two variants. These results are parallel to those in Barranquilla. TABLE 4 Effects of Distance, Location and Grammatical Gender Factor

PA

N

%

DA

N

%

PP

N

%

.72 .40 .45

75/303 40/531 31/375

25% 8% 8%

59/259 85/948

23% 9%

51/549 95/661

9% 15%

Distance Between Referent & Possessive No Overt Referent Less than 6 Words 6 or more Words

[.52] [.51] [.46]

130/303 226/531 144/375

43% 42% 39%

Range

.36 .53 .57

98/303 265/531 200/375

32% 50% 53%

21

27

Location of the Possessive Subject Object

.40 .53 Range

99/259 400/948

38% [.52] 42% [.49]

13

101/259 463/948

13

39% 49%

.61 .47 14

Grammatical Gender of Possessee Feminine Masculine

.57 .45 Range

I=input

12

253/549 247/661

46% 37%

.46 .53

245/549 319/661

45% [.45] 48% [.54]

7

I=.40 500/1210 41% I=.47 564/1210 47% I=.07 146/1210

12%

As illustrated in (14) and (15), apparently, the subject position facilitates the occurrence of possessive periphrases, which constitute the variant that most accurately denotes possession. That is, syntactically, the subject constitutes the location where possessive periphrases appear to be constrained the least. Conversely, as occurs in Barranquilla (Orozco 2010a: 208), the favorable effect of possessives in the object on possessive adjectives appears to be caused, at least partially, by the presence of prepositions. This is illustrated in (19) where a noun marked for possession (limitaciones) occurs as the object of a preposition (de). (19) Entonces yo creo que en ese aspecto somos... hay que ser humilde, humilde en el sentido de ser uno consciente de sus limitaciones. [NM02028]

MURIEL/OROZCO

224

31/5/12

14:24

Página 224

RAFAEL OROZCO

In general, we obtain the same tendencies as in Barranquilla. However, some differences emerge in terms of the significance of location of the possessive on the variants explored. This constraint significantly conditions definite articles in Barranquilla but not in New York, and the opposite obtains with possessive periphrases.

4.1.6. Grammatical Gender of the Possessee My exploration of this constraint originally included grammatical number with the four factors listed in 2.3. Plural and singular feminine possessees on the one hand, and singular and plural possessees, on the other, registered similar statistical tendencies in preliminary analyses. These tendencies indicate that New York Colombians, like barranquilleros, make a greater distinction in terms of grammatical gender than in terms of grammatical number. This is somewhat similar to what occurs with the possessor, which is not conditioned by grammatical number. Thus, I merged the factors with similar tendencies, and I proceeded to test feminine versus masculine possesses in all subsequent analyses. The grammatical gender of the possessee, as in Barranquillla, conditions the occurrence of both possessive adjectives and articles (see examples 16 and 19) but not that of possessive periphrases. The individual tendencies registered by each factor are also similar to those in Barranquilla. In the absence of possessive periphrases, possessive adjectives and definite articles acting as possessors yield complementary results. While feminine possessees favor possessive adjectives (.57), masculine possessees disfavor this variant (.45). Conversely, definite articles are favored by nouns marked for possession that are grammatically masculine (example 20), with a value of .53. They are also disfavored by those grammatically marked as feminine (.46). (20) Mira, por lo menos nosotros ya estamos atados aquí quien sabe hasta cuándo porque los niños; ellos no quieren salir de aquí. [NF06004]

In general, these results corroborate the Barranquilla findings for possessive adjectives. They seem to indicate that determiners, the most neutral of the three variants, are favored by masculine plurals, the most neutral of the possessees, in term of grammatical gender since plural masculines regularly incorporate plural feminines. For instance, in (20) los niños (both male and female)’ is marked for masculine grammatical gender but also includes las niñas. Having discussed the internal factors which promote the occurrence of the three variants, I will now turn to the discussion of the external factors significant

MURIEL/OROZCO

31/5/12

14:24

Página 225

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

225

in the occurrence of possessive adjectives, determiners, and possessive periphrases. The effects of external, social constraints answer our third main research question.

4.2. EXTERNAL CONSTRAINTS In the particular case of the New York Colombian community, external constraints can tell us how a recent population shift impacts on processes already in progress prior to the onset of simultaneous, direct language and dialect contact. Learning about how changes in people’s social situations affect their linguistic behavior will lead to a better understanding of language variation and change in immigrant and multilingual settings. Six of the seven external constraints explored reached statistical significance: (1) educational attainment, (2) length of residence (LOR), (3) age, (4) socioeconomic status, (5) sex, and (6) age of arrival. Seen from another perspective, all six of them condition possessive adjectives, three (education, LOR, and age) condition definite articles, and two (education and LOR) possessive periphrases. Only the speaker’s linguistic repertoire did not reach statistical significance for any of the variants. It is interesting to observe that there are similarities in the effect that external factors have on the expression of futurity and on the expression of nominal possession. Specifically, the results for sex and age reveal the same general tendencies. A discussion of the social constraints that most significantly condition nominal possession follows.

4.2.1. Education This is one of two external constraints conditioning all three variants under study. The results (Table 5) clearly indicate that possessive adjectives are favored by speakers pursuing their higher education in the US (.64) and by those who attended college in Colombia (.57). These same speakers also disfavor the occurrence of possessive periphrases with statistical weights of .25 and .41, respectively. Moreover, those with higher education in the US also disfavor definite articles (.40) whereas those who attended college in Colombia have a neutral effect with a value of .50. Conversely, the high school dropouts are simultaneously the strongest promoters of possessive periphrases (.71) and those who most strongly disfavor possessive adjectives (.32). These results seem to indicate that speakers who have attained more education disfavor the expansion of the periphrases, exhibiting a conservative linguistic behavior.

MURIEL/OROZCO

31/5/12

14:24

Página 226

226

RAFAEL OROZCO

The fact that the high school dropouts promote possessive periphrases more strongly than they promote determiners while simultaneously disfavoring possessive adjectives suggests that contact with English does not influence how they express possession. These individuals also register the lowest degree of fluency and literacy in English. Thus, if we were to interpret education as an indicator of the effect of contact with English on the possessive, we could say that the more Spanish dominant speakers favor possessive periphrases and determiners while disfavoring possessive adjectives. Spanish dominant individuals comprise perhaps both the largest segment of the New York Colombian community and also the largest portion of the immigrant Latino community in the greater NYC area. Perhaps these speakers have preserved basilectal features in their speech as a result of their immigration. Further, it is also arguable that the results reflect a certain degree of dialect leveling caused by their close contact with Spanish speakers from all over the Hispanic world (Orozco 2009b: 51).

TABLE 5 Effects of Education and LOR Factor

PA

N

%

DA

N

%

PP

N

%

Education US Post Secondary Education Attended College in Colombia High School Only Incomplete High School Range

32

.64

67/138

49%

.40

64138

46%

.25

7/138

5%

.57 .49

169/375 95/272

45% 35%

.49 .48

169/375 140/272

45% 51%

.41 .59

37/375 37/272

10% 14%

.32

56/217

26%

.60

120/217

55%

.71

41/217

19%

40/178 72/721 24/154 10/157

22% 10% 16% 6%

20

46

LOR Less than 5 Years 5-10 Years 10-20 Years More than 20 Years

.39 .55 .30 .59

Range

29

I=input

42/178 343/721 56/154 59/157

24% 48% 36% 38%

.44 .42 .57 .67 25

96/178 306/721 74/154 88/157

54% 42% 48% 56%

.46 .53 .68 .26 42

I=.40 500/1210 41% I=.47 564/1210 47% I=.07 146/1210

12%

MURIEL/OROZCO

31/5/12

14:24

Página 227

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

227

4.2.2. Length of Residence (LOR) As indicated on Table 5, LOR is the second of two social constraints significantly conditioning all three variants. Speakers who have lived in the US the longest favor the occurrence of possessive adjectives (.59) and determiners (.67) to the detriment of possessive periphrases (.26). Colombians who have been in the US 10 to 20 years strongly favor possessive periphrases with a value of .68. These individuals also favor definite articles (.57) and disfavor possessive adjectives (.30). Speakers with a length of residence of five to ten years favor possessive adjectives and possessive periphrases with statistical weights of .55 and .53, respectively, at the expense of definite articles (.42). There is a clear pattern with the use of definite articles: the longer speakers have resided in the US, the more they favor this variant. However, we do not have clearly discernible patterns for the possessive adjectives or the periphrases. The results for the PP show an increase in the favoring effect on this variant that starts with newcomers and stops after 20 years in the US. At the same time, those with the longest length of residence disfavor this variant. These tendencies may be due to increased contact with other Hispanics originally from regions where this variant occurs more frequently on the one hand, and the influence of English for those who have resided in the US for over 20 years. While accounting for these results may present a formidable challenge, it also provides an incentive to explore how the expression of possession will continue to evolve in this community. Further research should allow us to unravel the impact of LOR on the expression of possession.

4.2.3. Age Age conditions possessive adjectives and definite articles but not possessive periphrases. The results (Table 6) show that the youngest speakers favor possessive adjectives with a statistical weight of .68 while disfavoring definite articles (.44). Speakers in their 30s strongly favor definite articles (.71) and disfavor possessive adjectives with .29. Those in their 40s also disfavor possessive adjectives (.40) while exerting a neutral effect on definite articles (.50). The tendencies of the oldest speakers are similar to those for the youngest, i.e., they favor possessive adjectives with a value of .65 while disfavoring definite articles (.39). These results, like those for the expression of futurity (Orozco 2007b), lack clear patterns for the effects of age. However, they tell us that age grading does not play a role in the existing variation in the expression of possession. Nor would the results support the premise that the increasing use of the possessive periphrases

MURIEL/OROZCO

31/5/12

14:24

Página 228

228

RAFAEL OROZCO

constitutes a change from below promoted by the youngest speakers. At this juncture we need to further explore the effects of age and how they correlate with competence in both languages in terms of the expression of possession both in Colombia and NYC. TABLE 6 Effects of Speaker’s Age, Socioeconomic Status and Age Factor

PA

N

%

DA

N

%

PP

N

%

147/351 104/186 228/495 85/178

42% 56% 46% 48%

[.49] [.38] [.55] [.50]

40/351 12/186 61/495 33/178

11% 6% 12% 18%

Speaker’s Age 15 to 30 Years Old 31 to 40 Years Old 41 to 50 Years Old Older than 50

.68 .29 .40 .65

Range

39

164/351 70/186 206/495 60/178

47% 38% 42% 34%

.44 .71 .50 .39 32

Socioeconomic Status Blue-collar in Colombia & US White-collar in Colombia & US White-collar in Col., blue in US Range

.43

291/732

40% [.53]

350/732

48% [.51]

91/732

12%

.52

81/174

47% [.42]

82/174

47% [.39]

11/174

6%

.66

128/304

42% [.45]

132/304

43% [.52]

44/304

15%

23 Speaker’s Sex

Women Men

.44 .55 Range

I=input

209/571 291/639

37% [.52] 46% [.49]

297/571 267/639

52% [.52] 42% [.48]

65/571 81/639

11% 12%

500

41% I=.47

564

47% I=.07

146

12%

11 I=.40

4.2.4. Socioeconomic Status My exploration of this constraint tests Labov’s assertion that lower-middle and upper-working class individuals promote linguistic innovations (1990: 226). Socioeconomic status only conditions the occurrence of possessive adjectives.

MURIEL/OROZCO

31/5/12

14:24

Página 229

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

229

These are disfavored by individuals who remained in blue-collar occupations with a value of .43. In contrast, those with white-collar status in Colombia favor possessive adjectives. Those who have maintained their white-collar status slightly favor possessive adjectives (.52) whereas individuals with white-collar status in Colombia but blue-collar status in the US favor possessive adjectives more strongly with a statistical value of .66. Although socioeconomic status does not condition possessive periphrases significantly, the results appear to suggest that Labov’s postulate showing members of the upper-working class as linguistic innovators is extensive to immigrants’ use of their L1. In general terms, these results may suggest that individuals who have endured a decline in their socioeconomic status due to their immigration now try to express, by means of their sociolinguistic behavior, that they may have been negatively affected by their own social mobility. In a way, these individuals’ borderland situation caused by a socioeconomic reality that places them socially on a different level from the one they enjoyed in Colombia appears to affect their sociolinguistic behavior as has been attested in other communities (Fought 2008).

4.2.5. Speaker’s Sex As with previous research on this community (Orozco 2004, 2007b), and this linguistic variable (Orozco, 2009a, 2009b, 2010a), I analyzed the effect of sex in terms of oppositional categories. The results (Table 6) show that possessive adjectives are favored by men (.55) and disfavored by women (.44). These statistical tendencies run contrary to those obtained in Barranquilla. The differences between NYC and Barranquilla are consistent with Eckert’s observation (1989: 253) that “gender does not have a uniform effect on linguistic behavior for the community as a whole.” These results are also congruent with James’ view (1996: 119) that, in terms of linguistic behavior, “there are a variety of different factors which can give rise to differences between women and men.” If we consider that, in diachronic terms, the preferential use of possessive adjectives indicates grammatical correctness and formality, then New York Colombians do fall within established sociolinguistic behavior patterns. That is, men favor the more “correct” and conservative variant, a fact congruent with one of Labov’s principles of sexual differentiation, since he postulates that women are most often the innovators in change from below (1990: 215). When we compare these results to the sex effect on the expression of futurity (Orozco 2007b: 108), the reversal in the effect of speaker’s sex on both linguistic variables is remarkably similar. In both cases, men maintain a conservative sociolinguistic

MURIEL/OROZCO

31/5/12

14:24

Página 230

230

RAFAEL OROZCO

attitude by favoring the form traditionally considered more correct. While this appears to suggest that New York Colombians reflect their adjustment to their new sociolinguistic landscape through their sociolinguistic behavior, it may also be a result of dialect leveling and direct contact with English. In general, the results identify the social constraints conditioning nominal possession in the Spanish of New York Colombians after a few years in NYC where it is simultaneously in contact with English and with Spanish from all corners of the Hispanic world but mainly from the Caribbean. It is interesting to point out the existing similarities between Barranquilla and New York. However, the absence of clearly defined age patterns indicates the need to further explore their effect on the expression of possession in both speech communities.

5. Summary and conclusions This empirical analysis of the expression of nominal possession, a follow-up to Orozco (2010a), has focused on a linguistically under investigated community. I have answered three main research questions and tested several hypotheses. The distribution of variants answers our first question, revealing that definite articles occur more frequently than possessive adjectives and possessive periphrases, respectively. The results refute the hypothesis that convergence with English would trigger an increased use of possessive adjectives over the other two variants. Contrary to our expectations, possessive adjectives have lost ground. Their frequency has dropped by 6.5% whereas possessive periphrases occur almost twice as frequently as they do in Colombia. In general, the distribution of variants appears to suggest that, at the turn of the 21st century, dialect leveling exerts a stronger force on the expression of possession than contact with English. As with the expression of futurity (Orozco 2004), the possessive is conditioned by a complex interaction of linguistic and social constraints. The internal conditioning answers our second research question revealing that nominal possession is conditioned by eight linguistic constraints. Semantic category, grammatical person, and type of subject exert the strongest pressures on the three variants, as indicated by the range values. Besides being the same constraints that condition the possessive in Barranquilla (Orozco 2010a), the effects of individual factors reflect the same general tendencies. As in Barranquilla, semantic category has the strongest effect and its results are quite meaningful in several ways. The differences in the ranges between NYC and Barranquilla appear to be the result of something other than chance or fluctuation between two populations or corpora. Perhaps this is a first piece of evidence of the effects of convergence on the possessive. We have a clear indication of the influence of contact on

MURIEL/OROZCO

31/5/12

14:24

Página 231

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

231

semantic category and, to a lesser extent, on grammatical person. However, type of subject and distance between referent and possessive still do not show effects of contact with English, suggesting that some linguistic constraints are more susceptible to the effects of language contact than others. In answering our third research question, we have identified the six external constraints conditioning nominal possession. Two of them, education and LOR condition all three variants. The social constraints that condition the possessive in this community provide us with important information about what happens in the initial stages of contact between Spanish and English. In general, besides having a greater number of social constraints conditioning the expression of possession in New York, these constraints exert stronger conditioning effects. Although we know that the external constraints conditioning the possessive in both communities are, for the most part, different, and those which are the same have different effects, it is virtually impossible to separate the effects of language contact from those of dialect leveling. Contrary to what occurs with linguistic constraints, social constraints exert different effects in NYC and Barranquilla. This is especially noticeable in the case of sex, which shows opposite effects on the occurrence of possessive adjectives in both communities. We could be witnessing differences in linguistic behavior triggered by the new sociolinguistic environment Colombians find in NYC. If these results are typical of recent situations of linguistic contact (Orozco 2007b), we can assume that this is another situation where changes in the effects of external constraints simply precede inevitable changes in the effects of linguistic factors. Our results indicate that nominal possession is under the same internal conditioning in NYC and Barranquilla. The effects of individual factors are also largely the same in both places. Since this is analogous to what occurs with the expression of futurity (Orozco 2007a: 324), at this point, we can consider these results as further support to the theory of dialectal parallelism (Guy 2000). That is, New York Colombians and barranquilleros still belong to the same speech community despite their different sociolinguistic environments. A closer comparison between the expressions of futurity and possession shows that their similarities in Barranquilla and NYC extend from the distribution of variants to the internal conditioning and to the effects of speaker’s sex. However, their similarities do not extend to the remaining external constraints perhaps, among multiple reasons, because of the different sociolinguistic configurations of the communities involved. Thus, we can also conclude that a community’s social makeup could account for differences in the effects of external constraints despite similar internal conditioning. Concurrently, although the theory of dialectal parallelism can account for what happens with linguistic factors, apparently, it does not help predict the effects of social constraints.

MURIEL/OROZCO

232

31/5/12

14:24

Página 232

RAFAEL OROZCO

The emergence of possessive periphrases, genitive phrases that constitute the most accurate way to express nominal possession in Spanish, provides another instance of grammaticalization of an analytic paradigm. This development represents another interesting parallel between the expressions of futurity and possession, analogous to the occurrence of verbal periphrases which eventually replace inflections, discussed by Fleischman (2009/1982: 31) and Schwegler (1990: 144). The increasing occurrence of possessive periphrases appears to constitute another manifestation of cyclicity, a crosslinguistic evolutionary process whose effects on Spanish have been discussed by Gutiérrez (1995: 214) and Orozco (2007a, 2009a, and references therein), inter alios. These results show that the effects of cyclicity, mainly a series of internal syntactic and morphological adjustments, extend beyond verbal morphology. The spread of this phenomenon to other morphosyntactic domains has engendered more periphrastic paradigms of different kinds (Orozco 2004: 273). The large scale consequences of cyclicity obtain in a cycle where the language changes from being primarily synthetic to predominantly analytic and eventually becomes synthetic again (Fleischman 2009/1982: 152, Givón 1971, Schwegler 1990: 146-147). The fact that possessive periphrases are the newest of the three variants supports the premise that we are in the presence of another case of change in progress where an analytical innovation is expanding at the expense of an older, synthetic form. Further, the genesis and grammaticalization of the possessive periphrases appear to represent a new evolutionary cycle where an analytic paradigm develops as part of the ongoing recasting of the Latin/Spanish pronominal and possessive systems (cf. Kany 1969: 63-70, Gili y Gaya 1964: 240). Thus, considering this situation in view of the evolution of Spanish, we could hypothesize that, the development of the possessive periphrases, being subsequent to that of the periphrastic future, would fit perfectly within the wider scope of cyclicity (Orozco 2010a: 216). This would emerge as another manifestation of that process because the Latin synthetic possessive inherited by Spanish had an analytic realization at one point in its evolution which gave way, during the middle ages, to the modern synthetic possessive adjectives (Penny 2002, Eberenz 2004: 617 and references therein). The existing variation in the expression of possession appears to be part of a series of internally motivated changes in progress. These started in Colombia and have accelerated in the New York Colombian community (Orozco 2007a). In the expression of possession, this change would trigger the occurrence of possessive periphrases as reflected in the distribution of variants. According to Silva-Corvalán (1994), the proliferation of periphrastic forms and the acceleration of internally motivated change would be consequences of contact with English in NYC. The case of the possessive is particularly interesting because it is simultaneously under the effect of pressures that push and pull it in different

MURIEL/OROZCO

31/5/12

14:24

Página 233

THE EXPRESSION OF NOMINAL POSSESSION IN NEW YORK CITY

233

directions, as evidenced by the effects of both internal and external constraints. On the one hand, dialect leveling appears to favor the proliferation of the possessive periphrases; an example of this would be the pressures exerted by the high school dropouts who favor the periphrases. On the other hand, the influence of English appears to favor possessive adjectives, a pressure that would slow down the apparently incipient change in progress. Cases of linguistic contact curbing the speed of an ongoing change have precedents in other Hispanic communities as occurs with the expression of futurity in Castellón, explored by Blas Arroyo (2008). It would also be necessary to follow this situation closely to determine, after a few generations, which of the opposing pressures ultimately prevails in determining the fate of the expression of nominal possession. This analysis of the expression of possession in the New York Colombian community sought to determine the impact of contact with English and with other varieties of Spanish on the genitive. However, at this early stage of language contact in this community, it is impossible to readily account for every difference between Barranquilla and NYC in terms of contact with English or dialect leveling. This paper has discussed what is happening with the expression of possession at this stage in its evolution. The findings also open the question of how much longer this will obtain among New York Colombians since the ultimate fate of the expression of possession in this community is inextricably tied to the future maintenance of Spanish in its linguistic repertoire. Our results are important in terms of what occurs in the early stages of a direct language contact situation. After having documented the existing variation in the expression of nominal possession, we can expect subsequent research in other speech communities to help more accurately assess what appears to be another instance of change in progress. Further research will also help answer the questions that remain open regarding the conditioning exerted by internal as well as external constraints on this and other linguistic variables.

MURIEL/OROZCO

31/5/12

14:24

Página 234

MURIEL/OROZCO

31/5/12

14:24

Página 235

REFERENCES

ALARCOS LLORACH, Emilio (1973): Estudios de gramática funcional del español. Madrid: Gredos. ALBA, Orlando (1982): “Función del acento en el proceso de elisión de la /s/ en la República Dominicana”, in: Alba, Orlando: El español del Caribe: ponencias del VI simposio de dialectología. Santiago: Universidad Católica Madre y Maestra, 17-26. — (2004): Cómo hablamos los dominicanos: Un enfoque sociolingüístico. Santo Domigo: Grupo León Jimenes. ALBA DE DIEGO, Vidal/SÁNCHEZ LOBATO, Jesús (1980): “Tratamiento y juventud en la lengua hablada: aspectos sociolingüísticos”, in: Boletín de la Real Academia Española 60, 95-127. ALBOR, Hugo R. (1986): “Uso e interpretación de ser en construcciones galicadas y en Él necesita es descansar”, in: Thesaurus XLI, 173-186. ALCINA FRANCH, Juan/BLECUA, José Manuel (2001/1975): Gramática española. Barcelona: Editorial Ariel. ALFARAZ, Gabriela (2002): “Miami Cuban Perceptions of Varieties of Spanish”, in: Long, Daniel/Preston, Dennis (eds.): Handbook of Perceptual Dialectology, vol. 2. Amsterdam/Philadelphia: John Benjamins, 1-11. AMASTAE, Jon/ELÍAS-OLIVARES, Lucía (eds.) (1982): Spanish in the United States: Sociolinguistic Aspects. New York: Cambridge University Press. ARANGO CANO, Jesús (1994): El idioma español en Colombia desde la conquista hasta hoy. Armenia: Editorial QuinGráficas. ARRÁZOLA, Roberto (1970): Palenque, primer pueblo libre de América. Cartagena: Ediciones Hernández. Atlas lingüístico etnográfico de Colombia (ALEC) (1981-1983). Bogotá: Instituto Caro y Cuervo. 5 vols. BAAYEN, R. Harald (2001): Word Frequency Distributions. Dordrecht/Boston: Kluwer Academic. BAILEY, Raleigh (2005): “New Immigrant Communities in the North Carolina Piedmont . Triad: Integration Issues and Challenges”, in: Goz´dziak, Elz bieta M./Martin, Susan (eds.): Beyond the Gateway: Immigrants in a Changing America. Lanham: Lexington Books, 57-86. BARRUTIA, Richard/SCHWEGLER, Armin (1994): Fonética y fonología españolas. New York: Wiley. BARTENS, Angela (2002): “Another Short Note on Creoles in Contact with Non-Lexifier Prestige Languages”, in: Journal of Pidgin and Creole Languages 17, 2, 273278. — (2003): “A Contrastive Grammar Islander-Caribbean Standard English-Spanish”, in: Annales Academiae Scientiarum Fennicae/Suomalaisen Tiedeakatemian Toimituksia, Humaniora 327, 1-175.

MURIEL/OROZCO

236

31/5/12

14:24

Página 236

REFERENCES

— (2004): “Notas sobre el uso de las formas de tratamiento en el español colombiano actual“ (Last consulted: 3 July 2010). — (2009): “A Comparison of the English-based Creoles of Nicaragua and San Andres and Old Providence,” in: Neuphilologische Mitteilungen 110, 3, 299-318. BATLLORI DILLET, Montse (1998): “La impersonalización en español medieval: recursos formales y Semánticos”, in: García, Claudio/González, Fabián/Mangado, Javier (eds.): Actas del IV CIHLE 1. Logroño: Universidad de la Rioja, 381-393. BAYLEY, Robert (2004): “The Quantitave Paradigm”, in: Chambers, J. K./Trudgill, Peter/Schilling-Estes, Natalie (eds.): The Handbook of Language Variation and Change. Malden: Blackwell Publishing, 117-141. BAYONA, Patricia (2006): “Sociolinguistic Competences in the Use of Colombian Pronouns of Address” (Last consulted: 1 mayo 2008). BECKMAN, Jill N. (1999): Positional Faithfulness: An Optionality Theoretic Treatment of Phonological Asymmetries. New York: Garland. BELLACK, Alan/HERSEN, Michel/LAMPARSKI, Danuta (1979): “Role-play Tests for Assessing Social Skills: Are They Valid? Are They Useful?”, in: Journal of Consulting and Clinical Psychology 47, 2, 335-342. BELLETTI, Adriana (2004): “Aspects of the Low IP Area”, in: Rizzi, Luigi (ed.): The Structure of IP and CP: The Cartography of Syntactic Structures. New York: Oxford University Press, 16-51. BELLO, Andrés (1978): Gramática de la lengua castellana. Madrid: Edaf. BELLO, Andrés/CUERVO, Rufino J. (1941): Gramática de la lengua castellana. Buenos Aires: Ediciones Anaconda. BENTIVOGLIO, Paola/SEDANO, Mercedes (1993): “Investigación sociolingüística: sus métodos aplicados a una experiencia venezolana”, in: Boletín de Lingüística 8, 3-35. BERGEN, John (ed.) (1990): Spanish in the United States: Sociolinguistic Issues. Washington: Georgetown University Press. BICKERTON, Derek/ESCALANTE, Aquiles (1970): “Palenquero: A Spanish Based Creole of Northern Colombia”, In: Lingua 32, 254-267. BLAS ARROYO, José Luis (2007): “El contacto de lenguas como factor de retención en procesos de variación y cambio lingüístico: Datos sobre el español en una comunidad bilingüe peninsular”, in: Spanish in Context 4, 2, 263-291. — (1995): “Tú y usted: dos pronombres de cortesía en el español actual. datos de una comunidad peninsular”, in: Estudios de Lingüística de la Universidad de Alicante (ELUA) 10, 21-44. — (2008): “The Variable Expression of Future Tense in Peninsular Spanish: The Present (and Future) of Inflectional Forms in the Spanish spoken in a Bilingual Region”, in: Language Variation and Change 20, 1, 85-126. BLUST, Robert (2004): ì* t to k : An Austronesian Sound Change Revisitedî, in: Oceanic Linguistics 43, 365-410. BOERSMA, Paul/WEENINK, David (2010): “Praat: Doing Phonetics by Computer (version 5.1.43)” (Last consulted: 4 August 2010).

MURIEL/OROZCO

31/5/12

14:24

Página 237

REFERENCES

237

BOOMERSHINE, Amanda/CURRIE HALL, Kathleen/HUME, Elizabeth/JOHNSON, Keith (2008): “The Impact of Allophony versus Contrast on Speech Perception”, in: Avery, Peter/ Dresher, Elan/ Rice, Keren (eds.): Contrast in Phonology: Perception and Acquisition. New York: Mouton de Gruyter, 143-172. BORDEN, Gloria J./GAY, Thomas (1979): “Temporal Aspects of Articulatory Movements for /s/-stop Clusters”, in: Phonetica 36, 21-31. BOSQUE, Ignacio (1999): “On Focus vs. Wh-movement: The Case of Caribbean Spanish”, in: Sophia Linguistica 44-45, 1-32. BOSQUE, Ignacio/DEMONTE, Violeta (eds.) (1999): Gramática descriptiva de la lengua española. Madrid: Espasa Calpe. BOUCHER, Victor J. (1994): “Alphabet-Related Biases in Psycholinguistic Inquiries: Considerations for Direct Theories of Speech Production and Perception”, in: Journal of Phonetics 22, 1, 1-18. BROWN, Dolores (1975): “The Use of ‘Tú’ and ‘Usted’ with Parents by Some Mexican American Students”, in: Hispania 58, 1, 126-127. — (1993): “El polimorfismo de la /s/ explosiva en el noroeste de México”, in: Nueva Revista de Filología Hispánica 1, 159-176. BROWN, Earl K. (2009a): “The Relative Importance of Lexical Frequency in Syllableand Word-Final /s/ Reduction in Cali, Colombia”, in: Collentine, Joseph/García, Maryellen/Lafford, Barbara/Marcos Marín, Francisco (eds.): Selected Proceedings of the 11th Hispanic Linguistics Symposium. Somerville: Cascadilla Proceedings Project, 165-178. — (2009b): A Usage-Based Account of Syllable- and Word-Final /s/ Reduction in Four Dialects of Spanish. LINCOM Studies in Romance Linguistics 62. München: Lincom Europa. BROWN, Esther L. (2000): “La reducción de la /s/ inicial en el habla de Chihuahua, México”. Conference paper. 29th Meeting of the Linguistic Association of the Southwest (LASSO). Puebla. — (2004): The Reduction of Syllable Initial /s/ in the Spanish of New Mexico and Southern Colorado: A Usage-Based Approach. Doctoral Dissertation. Albuquerque: University of New Mexico. — (2005): “New Mexican Spanish: Insight into the Variable Reduction of ‘la ehe inihial’ (/s-/)”, in: Hispania 88, 813-824. — (2006): “The Effects of Discourse Context on Phonological Representation”. Conference paper. Linguistic Society of America. Albuquerque. BROWN, Esther L./TORRES CACOULLOS, Rena (2002): “¿Qué le vamoh aher?: Taking the syllable out of Spanish /s/ reduction”, in: University of Pennsylvania Working Papers in Linguistics: Papers from NWAV 30. Philadelphia: University of Pennsylvania Press, 17-32. — (2003): “Spanish /s/: A different Story from Beginning (Initial) to End (Final)”, in: A Romance Perspective in Language Knowledge and Use. Selected Papers from the 31th Linguistic Symposium on Romance Languages (LSRL), Chicago. Amsterdam/ Philadelphia: John Benjamins, 22-38.

MURIEL/OROZCO

238

31/5/12

14:24

Página 238

REFERENCES

BROWN, Penelope/LEVINSON, Stephen (1987): Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. BROWN, Roger/GILMAN, Albert (1960): “The Pronouns of Power and Solidarity”, in: Sebeok, Thomas (ed.): Style in Language. Cambridge: The M.I.T. Press, 253-276. BRUYNE, Jacques de (2004/1995): A Comprehensive Spanish Grammar. Malden: Blackwell Publishing. BRZOZOWSKA-ZBURZYNSKA, Beata (2005): “Los marcadores de la deixis espacial en español y en frances”, in: Estudios de Lingüística Universidad de Alicante (ELUA) 19, 65-84. BUTT, John/BENJAMIN, Carmen (2004): A New Reference Grammar of Modern Spanish, Fourth Edition. New York: McGraw Hill. BYBEE, Joan (1985): Morphology: A Study of the Relation between Meaning and Form. Amsterdam/Philadelphia: John Benjamins. — (2000): “The Phonology of the Lexicon: Evidence from Lexical Diffusion”, in: Kemmer, Suzanne/Barlow, Michael: Usage-Based Models of Language. Stanford: CSLI Publications, 65-85. — (2001): Phonology and Language Use. Cambridge: Cambridge University Press. — (2002): “Word Frequency and Context of Use in the Lexical Diffusion of Phonetically Conditioned Sound Change”, in: Language Variation and Change 14, 261-290. — (2003): “Mechanisms of Change in Grammaticalization: The Role of Frequency”, in: Joseph, Brian D./Janda, Richard D. (eds.): Handbook of Historical Linguistics. Oxford: Blackwell Publishing, 602-623. — (2006): “From Usage to Grammar: The Mind’s Response to Repetition”, in: Language 82, 4, 711-733. BYBEE, Joan/PERKINS, Revere/PAGLIUCA, William (1994): The Evolution of Grammar: Tense, Aspect and Modality in the Languages of the World. Chicago: University of Chicago Press. BYBEE, Joan/SCHEIBMAN, Joanne (1999): “The Effect of Usage on Degree of Constituency: The Reduction of don’t in American English”, in: Linguistics 37, 575-596. BYBEE, Joan/THOMPSON, Sandra A. (1997): “Three Frequency Effects in Syntax”, in: Berkeley Linguistics Society 23, 378-388. CAMACHO, José (2006): “In Situ focus in Caribbean Spanish: Towards a Unified Account of Focus”, in: Sagarra, Nuria/Toribio, Almeida Jacqueline (eds.): Selected Proceedings of the 9th Hispanic Linguistics Symposium. Somerville: Cascadilla Proceedings Project, 13-23. CANFIELD, D. Lincoln. (1988): El español de América. Barcelona: Editorial Crítica. CANTOR, Erik (2000): Ni aniquilados, ni vencidos. Los Emberá y la gente negra del Atrato bajo el dominio español. Siglo XVII. Bogotá: Instituto Colombiano de Antropología e Historia. CARBONERO CANO, Pedro (1979): Deixis espacial y temporal en el sistema lingüístico. Sevilla: Universidad de Sevilla. CARRASCO, Félix (1986): “La indeterminación del sujeto y la práctica sociolingüística”, in: Moreno de Alba, José (ed.): Actas del II Congreso Internacional sobre el español de América. México: Universidad Autónoma de México, 390-397.

MURIEL/OROZCO

31/5/12

14:24

Página 239

REFERENCES

239

CARRICABURO, Norma (1997): Las fórmulas de tratamiento en el español actual. Madrid: Arco Libros. CASHMAN, Holly (2007): “An Overview of Linguistic Politeness in the Spanish of the United States”, in: Placencia, María Elena/García, Carmen (eds.): Research on Politeness in the Spanish-Speaking World. Mahwah: Lawrence Erlbaum, 121-140. CÁSSERES ESTRADA, Solmery (2005): Diccionario lengua afro palenquero-español. Cartagena de Indias: Ediciones Pluma de Mompox. CASTELLANOS, Isabel M. (1980): “Sociolinguistic Attitudes toward Caribbean Spanish”, in: Lenguaje 11, 73-91. CASTILLO MATHIEU, Nicolás del (1984): “Negro-African Vocabulary in San Basilio de Palenque”, in: Thesaurus 39, 80-169. CASTRO-MITCHELL, Amanda (2000): Pronominal Address in Honduran Spanish. München: Lincom Europa. CEDERGREN, Henrietta I. (1973): The Interplay of Social and Linguistic Factors in Panamá. Doctoral Dissertation. Ithaca: Cornell University. — (1978): “En torno a la -s final de sílaba en Panamá”, in: Corrientes actuales en la dialectología del Caribe hispánico. Río Piedras: Editorial Universitaria, 35-50. CHASTON, John (1987): Aspect Choice in Preterit and Imperfect Usage in the Speech of Mexican-American College Students in Texas: A Sociolinguistic Approach. Doctoral Dissertation. Austin: University of Texas at Austin. CHAUDENSON, Robert (2003): La créolisation: théorie, applications, implications. Paris: L’Harmattan. CHELA FLORES, Godsuno (1986): “Las teorías fonológicas y los dialectos del Caribe hispánico”, in: Núñez Cedeño, Rafael/Páez, Iraset/Guitar, Jorge (eds.): Estudios sobre la fonología del español del Caribe. Caracas: Ediciones La Casa Bello, 21-30. CHOMSKY, Noam (2000): “Minimalist Inquiries: The Framework”, in: Martin, Roger/ Michaels, David/Uriagereka, Juan (eds.): Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik. Cambridge: MIT Press, 89-155. CID-HAZARD, Susana Margarita (2003): “Variación de estilo en relación a la variable fonológica /s/ en el español de Santiago de Chile”, in: Southwest Journal of Linguistics 22, 13-43. CINQUE, Guglielmo (1999): Adverbs and Functional Heads: A Cross-Linguistic Perspective. New York: Oxford University Press. COLMENARES, Germán (1991): Los esclavos en la Gobernación de Popayán 1680-1780. Tunja: Universidad Pedagógica y Tecnológica de Colombia. COMPANY COMPANY, Concepción (2008): “Gramaticalización, género discursivo y otras variables en la difusión del cambio sintáctico”, in: Kabatek, Johannes (ed.): Sintaxis histórica del español y cambio lingüístico. Madrid/Frankfurt: Iberoamericana/Vervuert, 17-51. CONTRERAS, Heles (1991): “On the Position of Subjects”, in: Syntax and Semantics: Perspectives on Phrase Structure 25, 63-79. CORREA RAMIREZ, Amanda (1990): “On a Peculiarity of Rural Spanish”, in: Glotta 5, 2, 20-25.

MURIEL/OROZCO

240

31/5/12

14:24

Página 240

REFERENCES

COVARRUBIAS, Patricia (2002): Culture, Communication, and Cooperation. Interpersonal Relations and Pronominal Address in a Mexican Organization. Lanham: Rowman & Littlefield. CRIADO DE VAL, Manuel (1966): Gramática española y comentario de textos, tercera edición. Madrid: Editorial Saeta. CUERVO, Rufino José. (1939): Apuntaciones críticas sobre el lenguaje bogotano con frecuente referencia al de los países de Hispano-América. Bogotá: Instituto Caro y Cuervo. CURNOW, Timothy Jowan/TRAVIS, Catherine E. (2004): “The Emphatic es Construction of Colombian Spanish”, in: Moskovsky, Christo (ed.): Selected Papers from the 2003 Conference of the Australian Linguistic Society, http://www.als.asn.au. — (2008): “Locational Adverbs in Non-Spatial Settings: The Case of ahí in Colombian Spanish Conversation”, in: Curnow, Timothy Jowan (ed.): Selected Papers from the 2007 Conference of the Australian Linguistic Society; http://www.als.asn.au. DAVIS, Martha Swearingen (1993): “The Anterior and Object Clitic Placement in Palenquero”, in: Southwest Journal of Linguistics 12, 1-2. — (2000): “The Past Imperfect in Palenquero”, in: Studies in Language 24, 3, 565-581. DEMELLO, George (2000): "Tú impersonal en la habla culta", in: Nueva Revista de Filología Hispánica 48, 2, 359-372. DILLEY, Laura/SHATTUCK-HUFNAGEL, Stefanie/OSTENDORF, Mari (1995): “Glottalization of Word-Initial Vowels as a Function of Prosodic Structure”, in: Journal of Phonetics 24, 423-444. DONOHUE, Mark (2002): “Which Sounds Change: Descent and Borrowing in the Skou Family”, in: Oceanic Linguistics 41 (1), 171- 221. DUBOIS, John W./SCHUETZE-COBURN, Stephan/CUMMING, Susanna/PAOLINO, Danae (1993): “Outline of Discourse Transcription”, in: Edwards, Jane/Lampert, Martin (eds.): Talking Data: Transcription and Coding in Discourse. Hillsdale: Lawrence Erlbaum Associates, 45-89. DURÁN, Richard (ed.) (1981): Latino Language and Communicative Behavior. Norwood: Ablex. EBERENZ, Rolf (2004): “Cambios morfosintácticos en la Baja Edad Media”, in: Cano, Rafael (coord.): Historia de la lengua española. Barcelona: Ariel, 613-641. ECKERT, Penelope (1989): “The Whole Woman: Sex and Gender Differences in Variation”, in: Language Variation and Change 1, 3, 245-267. EDMONDS, Joseph (1978). “The Verbal Complex V’ - V in French”, in: Linguistic Inquiry 9, 151-175. ELÍAS-OLIVARES, Lucía (ed.) (1983): Spanish in the U.S. Setting: Beyond the Southwest. Rosslyn: National Clearinghouse for Bilingual Education. ELÍAS-OLIVARES, Lucía/LEONE, Elizabeth/CISNEROS, René/GUTIÉRREZ, John (eds.) (1985): Spanish Language Use and Public Life in the USA. Berlin: Mouton. ENGSTRAND, Olle/KRULL, Diana (2001): “Segment and Syllable Reduction: Preliminary Observations”, in: Working Papers, Lund University, Department of Linguistics 49, 26-29.

MURIEL/OROZCO

31/5/12

14:24

Página 241

REFERENCES

241

ERKER, Daniel (2010): “A Subsegmental Approach to Coda /s/ Weakening in Dominican Spanish”, in: The International Journal of the Sociology of Language 203, 9-26. ESCALANTE, Aquiles (1954): “Notas sobre el Palenque de San Basilio, una comunidad negra en Colombia”, in: Divulgaciones Etnológicas (Barranquilla) 3, 207-359. — (1979): El Palenque de San Basilio: una comunidad de negros cimarrones. Barranquilla: Editorial Mejoras. — (1988). “Influencia bantú en la cultura popular de la costa atlántica colombiana”, in Desarrollo 87, 37-47. ESLING, John. (2005): “There is not Back Vowels: The Laryngeal Articulator Model”, in : Canadian Journal of Linguistics 50, 13- 44. ESLING, John/FRASER, Katherine/HARRIS, Jimmy (2005): “Glottal stop, Glottalized Resonants, and Pharyngeals: A Reinterpretation with Evidence from a Laryngoscopic Study of Nuuchahnulth (Nootka)”, in: Journal of Phonetics 33, 383- 410. FÉLIX-BRASDEFER, J. César (2007): “Natural Speech vs. Elicited Data: A Comparison of Natural and Role Play Requests in Mexican Spanish”, in: Spanish in Context 4, 2, 159-185. FERGUSON, Charles A. (1990): “From Esses to Atiches: Identifying Pathways of Diachronic Change”, in: Croft, William/Denning, Keith/Kemmer, Suzanne (eds.): Studies in Typology and Diachrony for Joseph H. Greenberg. Amsterdam/Philadelphia: John Benjamins, 59-78. FERNÁNDEZ, Salvador (1951): Gramática española. Madrid: Revista de Occidente. FILE-MURIEL, Richard J. (2006): “Spanish Adjective Position: Differences between Written and Spoken Discourse”, in: Clements, J. Clancy/Yoon, Jiyoung: Functional Approaches to Spanish Syntax. New York: Palgrave MacMillan, 203-218. — (2007): The Role of Lexical Frequency and Phonetic Context in the Weakening of Syllable-Final Lexical /s/ in the Spanish of Barranquilla, Colombia. Doctoral Dissertation. Bloomington: Indiana University. — (2009): “The Role of Lexical Frequency in the Weakening of Syllable-Final Lexical /s/ in the Spanish of Barranquilla, Colombia”, in: Hispania 92, 348-360. — (2010): “Lexical Frequency as a Scalar Variable in Explaining Variation”, in: The Canadian Journal of Linguistics/La Revue canadienne de Linguistique 55, 1-25. FILE-MURIEL, Richard J./BROWN, Earl (2010): “The Gradient Nature of s-Lenition in Caleño Spanish”, in: Penn Working Papers in Linguistics: Selected Papers from NWAV 38 16, 2. Philadelphia: University of Pennsylvania Press, 45-55. — (2011): “The Gradient Nature of S-Lenition in Caleño Spanish” in Language Variation and Change 23, 223-243. FILE-MURIEL, Richard J./DÍAZ-CAMPOS, Manuel (2003): “Grammatical Judgments and Phonetic Reality: A Study of Internal Constraints in Phonological Variation”. Conference paper. 114th Meeting Acoustical Society of America. Austin. FISHMAN, Joshua/COOPER, Robert/MA, Roxana (eds.) (1971): Bilingualism in the Barrio. Bloomington: Indiana University. FISHMAN, Joshua/KELLER, Gary (eds.) (1982): Bilingual Education for Hispanic Students in the United States. New York: Teachers College, Columbia University.

MURIEL/OROZCO

242

31/5/12

14:24

Página 242

REFERENCES

FLEISCHMAN, Suzanne (2009/1982): The Future in Thought and Language: Diachronic Evidence from Romance. New York: Cambridge University Press. FLÓREZ, Luis (1950): “El habla del Chocó”, in: Thesaurus 6, 110-116 — (1951): La pronunciación del español en Bogotá. Bogotá: Instituto Caro y Cuervo — (1961): El atlas lingüístico-etnográfico de Colombia (ALEC). Nota informativa, in: Thesaurus: Boletín del Instituto Caro y Cuervo 16, 77-125. — (1973): Las “Apuntaciones Críticas” de Cuervo y el español eogotano cien años después: pronunciación y fonética. Bogotá: Instituto Caro y Cuervo. — (1978): “Sobre algunas formas de pronunciar muchos colombianos el espanol”, in: Thesaurus: Boletin del Instituto Caro y Cuervo 33, 197-246. FONTANELLA DE WEINBERG, Maria Beatriz (1973a): “Comportamiento ante -s de hablantes femeninos y masculinos del español bonaerense”, in: Thesaurus 25, 12-22. — (1973b): “Comportamiento ante -s de hablantes femeninos y masculinos del español bonarense,” in: Romance Philology, 27, 50-58. FORREST, Karen/WEISMER, Gary/MILENKOVIC, Paul/DOUGALL, Ronald N. (1988): “Statistical Analysis of Word-Initial Voiceless Obstruents: Preliminary Data”, in: Journal of the Acoustical Society of America 84, 1, 115-123. FOUGHT, Carmen (2008): “On the Borderlands of Communities: Taking Linguistic Research to la Frontera”, plenary addresss. NWAV 37, Houston: 8 November 2008. FRIEDEMANN, Nina S. de/PATIÑO ROSSELLI, Carlos (1983): Lengua y sociedad en el Palenque de San Basilio. Bogotá: Instituto Caro y Cuervo. GARCÍA GONZÁLEZ, Francisco (1996): “Se le(s)/lo(s)/la(s) impersonal”, in: Casado, Manuel (ed.): Scripta philologica in memoriam Manuel Taboada Cid. La Coruña: Universidade da Coruña, 387-405. GAHL, Susanne (2008): “Time and Thyme are not Homophones: The Effect of Lemma Frequency on Word Durations in Spontaneous Speech”, in: Language 84, 474-496. GEESLIN, Kimberly (2003): “A Comparison of Copula Choice: Native Spanish Speakers and Advanced Learners”, in: Language Learning 53, 4, 703-764. GEESLIN, Kimberly/GUIJARRO-FUENTES, Pedro (2007): “Linguistic and Social Predictors of Copula Use in Galician Spanish”, in: Potowski, Kim/Cameron, Richard (eds.): Spanish in Contact: Policy, Social and Linguistic Inquiries. Amsterdam/Philadelphia: John Benjamins, 253-274. GILI Y GAYA, Samuel (1964): Curso superior de sintaxis española. Barcelona: Bibliograf. GIVÓN, Talmy (1971): “Historical Syntax and Synchronic Morphology: An Archaeologist’s Field Trip”, in: Papers from the Seventh Regional Meeting of the Chicago Linguistic Society, April 16-18, 1971. Chicago: Chicago Linguistic Society, 394-415. GÓMEZ, Nora Lucia (1999). “Basic structures and phonological processes in Colombian Sign Language”, in: Lenguaje 27: 50-77. GÓMEZ TORREGO, Leonardo (1998): La impersonalidad gramatical: descripción y norma. Madrid: Arco Libros. GOOD, Jeff (2005): “Split Prosody and Creole Simplicity: The Case of Saramaccan”, in: Journal of Portuguese Linguistics 3, 11-30. GOROSCH, Max (1967): “Un sujeto indeterminado o general expresado por la segunda persona del singular: tú”, in: Revue Romane 1, 13-26.

MURIEL/OROZCO

31/5/12

14:24

Página 243

REFERENCES

243

GRANDA, Germán de (1968): “La tipología “criolla” de dos hablas del área lingüística hispánica”, in: Thesaurus, 23, 193-205. — (1971): “Onomástica y procedencia africana de los esclavos negros en las minas del sur de la Gobernación de Popayán (siglo XVII)”, in: Revista española de Antropología Americana 4, 381-422. — (1977): Estudios sobre un área dialectal hispanoamericana de población negra. Las tierras bajas occidentales de Colombia. Bogotá: Instituto Caro y Cuervo. — (1978): Estudios lingüísticos hispánicos, afrohispánicos y criollos. Madrid: Gredos. — (1988): “Los esclavos del Chocó: su procedencia africana (siglo XVIII) y su posible incidencia lingüística en el español del área”, in: Thesaurus 43, 65-80. GREGORY, Michelle/RAYMOND, William/BELL, Alan/FOSLER-LUSSIER, Eric/JURAFSKY, Daniel (1999): “The Effects of Collocational Strength and Contextual Predictability in Lexical Production”, in: Papers from the Regional Meetings, Chicago Linguistic Society 35, 2, 151-166. GRIES, Stefan Th. (2009): Statistics for Linguistics with R: A Practical Introduction. Berlin: Mouton de Gruyter. GUITART, Jorge (1989): “On Spanish Cleft Sentences”, in: Studies in Romance Linguistics: Selected Papers from the 17th Linguistic Symposium on Romance Languages (LSRL) XVII, 129-154. GUTIÉRREZ, Manuel J. (1995): “On the Future of the Future Tense in the Spanish of the Southwest”, in: Silva-Corvalán, Carmen (ed.): Spanish in Four Continents: Studies in Language Contact and Bilingualism. Washington: Georgetown University Press, 214-223. GUTIÉRREZ AZOPARDO, Ildefonso (1980): Historia del negro en Colombia: ¿sumisión o rebeldía? Bogotá: Editorial Nueva América. GUTIÉRREZ-REXACH, Javier (2001): “Adverbial Weak Pronouns: Derivation and Interpretation”, in: Gutiérrez-Rexach, Javier/Silva-Villar, Luis (eds.): Current Issues in Spanish Syntax and Semantics. Berlin: Mouton de Gruyter, 143-174. GUY, Gregory R. (2000): “A identidade lingüística da comunidade de fala: paralelismo interdialectal nos padrões de variação lingüística”, in: Organon 14, 28/29, 17-32. HAENSCH, Günther/WERNER, Reinhold (eds.) (1993): Nuevo diccionario de americanismos, vol. 1: Nuevo diccionario de colombianismos. Bogotá: Instituto Caro y Cuervo. HAMMOND, Robert (1982): “El fonema /s/ en el español jíbaro: cuestiones teóricas”, in: Alba, Orlando: El Español del Caribe: Ponencias del VI simposio de dialectología. Santiago: Universidad Católica Madre y Maestra, 157-169. HANSSEN, Federico (1913): Gramática histórica de la lengua castellana. Halle: Max Niemeyer. HARRIS, James W. (1983): Syllable Structure and Stress in Spanish: A Nonlinear Analysis. Cambridge: MIT Press. HASPELMATH, Martin (1997): From Space to Time: Temporal Adverbials in the World’s Languages. München: Lincom Europa. HAWKINS, John A. (1994): A Performance Theory of Order and Constituency. Cambridge/New York: Cambridge University Press.

MURIEL/OROZCO

244

31/5/12

14:24

Página 244

REFERENCES

HENRÍQUEZ UREÑA, Pedro (1921): “Observaciones sobre el español en América”, in: Revista de Filología Española 8.357-90. HERNÁNDEZ, José Esteban (2002): “Accommodation in a Dialect Contact Situation”, in: Revista de Filología y Lingüística de la Universidad de Costa Rica 28, 2, 93-110. HERNANZ, María Luisa (1990): “En torno a los sujetos arbitrarios: La segunda persona del singular”, in: Demonte, Violeta/Garza, Beatriz (eds.): Estudios de lingüística de España y México. México: Instituto de Investigaciones Filológicas, 151-178. HIDALGO NAVARRO, Antonio (1996-97): “Sobre los mecanismos de impersonalización en la conversación coloquial: El tú impersonal”, in: Estudios de Lingüística Universidad de Alicante (ELUA) 11, 163-176. HOOPER, Joan Bybee (1976): “Word Frequency in Lexical Diffusion and the Source of Morphophonological Change”, in: Christie, William M. (ed.): Current Progress in Historical Linguistics. Amsterdam: North Holland, 96-105. HOPPER, Paul J./TRAUGOTT, Elizabeth Closs (1993): Grammaticalization. Cambridge: Cambridge University Press. HORVATH, Julia (1986): Focus in the Theory of Grammar and the Syntax of Hungarian. Dordrecht: Foris. HOSMER, David/LEMESHOW, Stanley (2000): Applied Logistic Regression. New York: Wiley. HOTTENROTH, Priska-Monika (1982): “The System of Local Deixis in Spanish”, in: Weissenborn, Jürgen/Klein, Wolfgang (eds.): Here and There: Cross-Linguistic Studies on Deixis and Demonstration. Amsterdam/Philadelphia: John Benjamins, 133-153. HSU, Yu-Yin (2008): “The Sentence-Internal Topic and Focus in Chinese”, in: 20th North American Conference on Chinese linguistics (NACCL-20), Columbus: The Ohio State University. HUALDE, José Ignacio (1991): “Aspiration and Resyllabification in Chinato Spanish”, in: Probus 3, 55-76. — (2005): The Sounds of Spanish. New York: Cambridge University Press. HUALDE, José Ignacio/SCHWEGLER, Armin (2008): “Intonation in Palenquero”, in: Journal of Pidgin and Creole Languages 23, 1-31. HURTADO, Luz Marcela (2001): La variable expresión del sujeto en el español de los colombianos y colombo-americanos residentes en el condado de Miami-Dade. Unpublished Doctoral Dissertation. Gainesville: University of Florida. — (2002): “El mantenimiento del español en Miami, Florida”, in: Litterae 11, 148-170. — (2005a): “El uso de tú, usted y uno en el español de los colombianos y colombo-americanos”, in: Ortiz López, Luis/Lacorte, Manel (eds.): Contactos y contextos lingüísticos: El español en los Estados Unidos y en contacto con otras lenguas. Madrid/ Frankfurt: Iberoamericana/Vervuert, 187-200. — (2005b): “Condicionamientos sintáctico-semánticos de la expresión del sujeto en el español colombiano”, in: Hispania 88, 2, 335-348. HURTADO IDÁRRAGA, Lucía (1994): Synchronic Variation in Colombian Speech: A Sociolinguistic Analysis of the Usage of the Pronouns of Address tú, vos, and usted, in Medellín (Antioquia) and Manizales (Caldas). Masters Thesis. Arizona State University.

MURIEL/OROZCO

31/5/12

14:24

Página 245

REFERENCES

245

HYMAN, Larry (2003): “Segmental Phonology”, in: Nurse, Derek/Philippson, Gerárd (eds.): The Bantu Languages. London: Routledge/Curzon; (Last consulted: 27 january 2009). ITKONEN, Esa (1991): Universal History of Linguistics: India, China, Arabia, Europe. Amsterdam/Philadelphia: John Benjamins. JACQUOT, André (1962): “Notes sur la phonologie du beembe”, in: Journal of African Languages 5, 232-242. J AMES , Deborah (1996): “Women, Men and Prestige Speech Forms: A Critical Review”, in: Bergvall, Victoria L./Bing, Janet M./Freed, Alice F. (eds.): Rethinking Language and Gender Research: Theory and Practice. London/New York: Longman, 98-125. JARAMILLO, June (1990): “Domain Constraints on the Use of Tú and Usted”, in: Bergen, John (ed.): Spanish in the United States: Sociolinguistic Issues. Washington: Georgetown University Press, 14-22. — (1995): “Social Variation in Personal Address Etiquette”, in: Hispanic Linguistics 6/7, 191-224. — (1996): “Tú and Usted: Address Etiquette in the Mexican American Family”, in: Hispanic Journal of Behavioral Sciences 18, 4, 522-532. JOHNSON, Keith (2004): “Massive Reduction in Conversational American English”, in: Proceedings of the 1st Session of the 10th International Symposium. Tokyo: The National Institute for Japanese Language, 29-54. KANY, Charles E. (1969): Sintaxis hispanoamericana. Madrid: Gredos. — (1975): American-Spanish Syntax (2nd ed.). Chicago: Chicago University Press. KASPER, Gabriele/DAHL, Merete (1991): “Research Methods in Interlanguage Pragmatics”, in: Studies in Second Language Acquisition 13, 2, 215-247. KATO, Mary A. (2008): “Clivadas sem operador no portugues brasileiro”. Conference paper. Sintaxe Gerativa do Português Brasileiro na Entrada do Século XXI: Minimalismo e Interfaces. São Paulo. — (2009): “VP-Remnant Movement in Portuguese Focus Constructions”. Conference paper. 39th Linguistic Symposium on Romance Languages (LSRL). Tucson. KENISTON, Hayward (1937): The Syntax of Castilian Prose: The Sixteenth Century. Chicago: The University of Chicago Press. KLEE, Carol/LYNCH, Andrew (2009): El español en contacto con otras lenguas. Washington: Georgetown University Press. KLEIN, Philip (1968): Modal Auxiliaries in Spanish. Doctoral Dissertation, Seattle: University of Washington. KOHLER, Klaus (1998): “The Disappearance of Words in Connected Speech”, in: ZAS Working Papers in Linguistics 11, 21-34. — (1999): “Plosive-Related Glottalization Phenomena in Read and Spontaneous Speech - A stød in German?”; (Last consulted: 28 september 2008). KUTEVA, Tania and Bernd Heine (2008): “On the Explanatory Value of Grammaticalization. Linguistic Universals and Language Change”, in: Good, Jeff (ed.): Linguistic Universals and Language Change. New York: Oxford University Press, 215-230.

MURIEL/OROZCO

246

31/5/12

14:24

Página 246

REFERENCES

LABOV, William (1972): Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. — (1990): “The Intersection of Sex and Social Class in the Course of Linguistic Change”, in: Language Variation and Change 2, 205-254 — (2001): Principles of Linguistic Change: Social Factors. Malden: Blackwell Publishing. — (2003): “Words Floating on the Surface of Sound Change”. Conference paper. NWAV 32. Philadelphia. LACORTE, Manel/LEEMAN, Jennifer (eds.) (2009): Español en Estados Unidos y otros contextos de contacto: sociolingüística, ideología y pedagogía. Madrid/Frankfurt: Iberoamericana/Vervuert. LADEFOGED, Peter (1964): A Phonetic Study of West African Languages. Cambridge: Cambridge University Press. — (2001): A Course in Phonetics. Boston: Heinle & Heinle. — (2003): Phonetic Data Analysis: An Introduction to Fieldwork and Instrumental Techniques Malden: Blackwell Publishing. LAFFORD, Barbara (1986): “Valor diagnóstico-social del uso de ciertas variantes de /s/ en el español de Cartagena, Colombia”, in: Núñez Cedeño, Rafael/Páez, Iraset/Guitart, Jorge (eds.) Estudios sobre la fonología del español del Caribe. Caracas: La Casa de Bello, 53-75. LASTRA DE SUÁREZ, Yolanda (1972): “Los pronombres de tratamiento en la Ciudad de México”, in: Anuario de Letras 10, 213-217. LAUZEN, Martha/DOZIER, David (2002): “You Look Mahvelous: An Examination of Gender and Appearance Comments in the 1999-2000 Prime-Time Season”, in: Sex Roles 46, 11/12, 429-437. LENNES, Mietta/AHO, Eija/TOIVOLA, Minnaleena/WAHLBERG, Leena (2006): “On the Use of the Glottal Stop in Finnish Conversational Speech”, in: The Phonetics Symposium 2006, Publications of the Department of Speech Sciences, University of Helsinki, 53, 93-102. LEÓN REY, José Antonio (1983): “El breve diccionario de colombianismos”, in: LópezMorales, Humberto/Molina, José Andrés de/Quilis, Antonio (eds.): Philologica Hispaniensia in Honorem Manuel Alvar, I. Madrid: Gredos, 361-371. LEONETTI, Manuel (1999): “El artículo”, in: Bosque, Ignacio/Demonte, Violeta (eds.): Gramática descriptiva de la lengua española. Madrid: Espasa Calpe, 787-890. LEVINSON, Stephen (1983): Pragmatics. Cambridge: Cambridge University Press. LEWIS, Anthony M. (2001): Weakening of Intervocalic /p, t, k/ in Two Spanish Dialects: Toward the Quantification of Lenition Processes. Doctoral Dissertation. Chicago: University of Illinois. LINDAU, Mona (1980): “Phonetic Differences in Nigerian Languages”, in: UCLA Working Papers in Phonetics, 51, 105-112. LIPSKI, John (1985a): “/s/ in Central American Spanish,” in: Hispania, 68, 143-149. — (1985b): “Black Spanish: The Last Frontier of Afro-America”, in: Crítica 2(1), 53-75. — (1986a): “El lenguaje de los negros congos del Panamá, estudio lingüístico”, in: Lexis 10 (1), 53-76. — (1986b): “Instability and Reduction of /s/ in the Spanish of Honduras”, in: Revista Canadiense de Estudios Hispánicos 11, 27-47.

MURIEL/OROZCO

31/5/12

14:24

Página 247

REFERENCES

247

— (1987): “Language Contact Phenomena in Louisiana Isleño Spanish”, in: American Speech 62, 4, 320-331. — (1990): “Sabine River Spanish: A Neglected Chapter in Mexican-American Dialectology”, in: Bergen, J. (ed.): Spanish in the Unites States: Sociolinguistic issues. Washington: Georgetown University Press, 1-13. — (1994): Latin American Spanish. London/New York: Longman. — (1999a): “The Many Faces of Spanish /s/-Weakening: (Re)Alignment and Ambisyllabicity”, in: Gutiérrez-Rexach, Javier/ Martínez-Gil, Fernando: Advances in Hispanic Linguistics: Papers from the 2nd Hispanic Linguistics Symposium. Somerville: Cascadilla Press, 198-213. — (1999b): “Evolución de los verbos copulativos en el español bozal”, in: Zimmermann, Klaus (ed.): Lenguas criollas de base lexical española y portuguesa. Madrid/ Frankfurt: Iberoamericana/Vervuert, 145-176. — (2000): “Back to Zero or Ahead to 2001?: Issues and Challenges in U. S. Spanish Research”, in: Roca, Ana (ed.): Research on Spanish in the United States: Linguistic Issues and Challenges. Somerville: Cascadilla Press, 1-41. — (2002): “Génesis y evolución de la cópula en los criollos afro-ibéricos”, in: Moñino, Yves/Schwegler, Armin (eds.): Palenque, Cartagena y Afro-Caribe: historia y lengua. Tübingen: Niemeyer, 65-101. — (2005): A History of Afro-Hispanic Language: Five Centuries and Five Continents. Cambridge: Cambridge University Press. — (2007): “El cambio /R/ > [D] en el habla afrohispánica. ¿Un rasgo ‘congo’?”, in: Boletín de Lingüística 19 (27), 94- 114. — (2008a): Afro-Bolivian Spanish. Madrid/Frankfurt: Iberoamericana/Vervuert. — (2008b): Varieties of Spanish in the United States. Washington: Georgetown University Press. — (2010): “Pitch Polarity in Palenquero: A Possible Locus of H Tone”, in: Colina, Sonia (ed.): Linguistic Studies in Romance Languages. Amsterdam/Philadelphia: John Benjamins, 111-127. — (2011): “El ‘nuevo’ palenquero y el español afroboliviano: ¿es reversible la descriollización?”, in: Ortiz López, Luis (ed.): Proceedings of the13th Hispanic Linguistic Symposium. Somerville: Cascadilla Proceedings Project, 1-16. LÓPEZ CHÁVEZ, Juan (1977): “El fonema /s/ en el habla de La Cruz, Sinaloa”, in: Nueva Revista de Filologia Hispánica 26, 332-340. LÓPEZ MORALES, Humberto (1989): Sociolingüística. Madrid: Gredos. — (2003): Los cubanos de Miami. Miami: Ediciones Universal. — (ed.) (2009): Enciclopedia del español en los Estados Unidos. Madrid: Instituto Cervantes/Santillana. LÓPEZ SCOTT, Alma (1983): A Sociolinguistic Analysis of /S/ Variation in Honduran Spanish. Doctoral Dissertation. Minneapolis: University of Minnesota. LORENZO, Emilio (1984): “Novedades sobre la segunda persona y la expresión impersonal”, in: Bernabé, A./Gangutia, E./Cuenca, L.A. de/López Facal, J. (eds.): ATHLON Satura Grammatica in honorem Francisci R. Adrados 1, 279-286.

MURIEL/OROZCO

248

31/5/12

14:24

Página 248

REFERENCES

— (1989): “Relación interpersonal y expresión impersonal”, in: Rodríguez González, F. et. al. (eds.): Comunicación y lenguaje juvenil. Madrid: Fundamentos, 217-230. LYNCH, Andrew (1999): The Subjunctive in Miami Cuban Spanish: Bilingualism, Contact, and Language Variability. Doctoral Dissertation. Minneapolis: University of Minnesota. — (2009): “Otras procedencias”, in: López Morales, Humberto (coord.): Enciclopedia del español en los Estados Unidos. Madrid: Instituto Cervantes, 382-389. LYONS, John (1977): Semantics. Cambridge: Cambridge University Press. MANN, Virginia A/REPP, Bruno H. (1980): “Influence of Vocalic Context on Perception of the S-[s] Distinction”, in: Perception and Psychophysics 28, 3, 213-228. MALDONADO, Ricardo (1999): “Espacios mentales y la interpretación del se impersonal”, in: Colombo, Fulvia (coord.): El centro de lingüística hispánica y la lengua española. México: Instituto de Investigaciones Filológicas, 205-227. MARCOS-MARÍN, Francisco (1992): “Corpus oral de referencia del español contemporáneo. Universidad autónoma de Madrid”; http://www.lllf.uam.es/corpus/corpus_oral. html. MARÍN, Diego (1972): “El uso de “tú” y “usted” en el español actual”, in: Hispania 55, 4, 904-908. MÁRQUEZ REITER, Rosina (2000): Linguistic Politeness in Britain and Uruguay: A Contrastive Study of Requests and Apologies. Amsterdam/Philadelphia: John Benjamins. MARTÍNEZ-CELDRÁN, Eugenio (1996): El sonido de la comunicación humana. Barcelona: Octaedro MARTÍNEZ-CELDRÁN, Eugenio/FERNÁNDEZ-PLANAS, Ana M./CARRERA-SABATÉ, Josefina (2003): “Castilian Spanish”, in: Journal of the International Phonetic Association 33 (2), 255-260. MASON, Keith William (1994): Comerse las eses: A Selective Bibliographic Survey of /s/ Aspiration and Deletion in Dialects of Spanish. Doctoral Dissertation. Ann Arbor: University of Michigan. MAYA, Luz A. (1998): “Demografía histórica de la trata por Cartagena 1533-1810”, in: Geografía Humana de Colombia. Tomo IV. Los Afrocolombianos. Bogotá: Instituto Colombiano de Cultura Hispánica. MCWHORTER, John H. (1995): “The Scarcity of Spanish-Based Creoles Explained,” in: Language in Society 24, 2, 213-244. — (1997): Towards a New Model of Creole Genesis. New York: Peter Lang. — (2000): Missing Spanish Creoles. USA: University of California Press. — (2005): Defining Creole. Oxford: Oxford University Press. MEDINA MORALES, Francisca (2010): “La metodología en los estudios sobre formas y fórmulas de tratamiento en español”, in: Hummel, Martin/Kluge, Bettina/Vázquez Laslop, María (eds.): Formas y fórmulas de tratamiento en el mundo hispánico. México/Graz: El Colegio de México/Karl Franzens Universität, 21-56. MEGENNEY, William (1982): “Elementos subsaháricos en el español dominicano”, in: Orlando Alba (ed.): El español del Caribe. Santiago: Universidad Católica Madre y Maestra, 185-201.

MURIEL/OROZCO

31/5/12

14:24

Página 249

REFERENCES

249

— (1986): El palenquero: un lenguaje post-criollo colombiano. Bogotá: Instituto Caro y Cuervo. — (1992): “El español de las Cumbes de Barlovento (Venezuela) y sus posibles vestigios Afronegroides: la filtración por la creolodización”, in: América Negra. 3, 54-67. MEJÍAS-BIKANDI, Errapel (1992): “The VP-Internal Subject Hypothesis and Spanish Sentence Structure”, in: Contemporary Research in Romance Linguistics. Papers from the 22nd Linguistic Symposium on Romance Languages (LSRL), 123, 275-289. MELLO, George de (2000): “Tú impersonal en el habla culta”, in: Nueva Revista de Filología Hispánica 48, 2, 359-372. MÉNDEZ DOSUNA, Julián (1985): “La duración de S en los grupos SP, ST, SK: A propósito del orden regular de difusión en algunos cambios fonéticos”, in: Melena, José L. (ed.): Symbolae Ludovico Mitxelena Septuagenario Oblatae. Vitoria: Universidad del País Vasco, 647-655. — (1987): “La aspiracion de ‘s’ como proceso condicionado por el contexto de silabas”, in: Revista Espanola de Linguistica 17, 15-35. — (1996): “Can Weakening Processes Start in Initial Position?”, in: Hurch, Bernhard/ Rhodes, Richard (eds.): Natural Phonology: The State of the Art. New York: Mouton de Gruyter, 97-106. MÉNDEZ VALLEJO, Dunia Catalina (2009): Focalizing Ser (‘to be’) in Colombian Spanish. Unpublished Doctoral Dissertation. Bloomington: Indiana University. MENÉNDEZ PIDAL, Ramón (1968/1904): Manual de gramática histórica española, decimotercera edición. Madrid: Espasa Calpe. MIKOLAJCZAK, Sylwia (2003): “Os tipos das construções com clivagem em portugués”, in: Studia Romanica Posnaniensia, 187-196. MILROY, Lesley (1987): Language and Social Networks. Oxford: Blackwell Publishing. MILROY, Lesley/GORDON, Matthew J. (2003): Sociolinguistics: Method and Interpretation. Malden: Blackwell Publishing. MINNICK FOX, Michelle Annette (2006): Usage-Based Effects in Latin American Spanish Syllable-Final /s/ Lenition. Philadelphia: University of Pennsylvania. MIYAMOTO, Masami (1993): Lista contrastiva de frecuencias del vocabulario del español hablado y escrito de Colombia. Kobe: Universidad Municipal de Estudios Extranjeros de Kobe. MIYOSHI, Junnosuke (1999): “Algunas reflexiones sobre acá y allá”, in: Samper, José Antonio/Troya, Magnolia (eds.): Actas del XI Congreso Internacional de la ALFAL. Las Palmas de Gran Canaria: Universidad de Las Palmas de Gran Canaria, 359-365. MOLHO, Maurice (1992): “La deixis española: lectura del significante”, in: Luna Traill, Elizabeth (ed.): Scripta philologica in honorem Juan M. Lope Blanch: a los 40 años de docencia en la UNAM y a los 65 años de vida. México: Universidad Autónoma de México, 203-217. MOLINA, Isabel (1993): “Las fórmulas de tratamiento de los jóvenes madrileños: estudio sociolingüístico”, in: Lingüística Española Actual 15, 2, 249-263. MOLINER, María (1984): Diccionario de uso del español. Madrid: Gredos.

MURIEL/OROZCO

250

31/5/12

14:24

Página 250

REFERENCES

MONTES GIRALDO, José Joaquín (1962): “Sobre el habla de San Basilio de Palenque (Bolívar, Colombia)”, in: Thesaurus 17, 446-450. — (1967): “Sobre el voseo en Colombia”, in: Thesaurus 22, 21-44. — (1974): “El habla del Chocó: notas breves”, in: Thesaurus 29, 409-428. — (1982): “El espanol de Colombia: Propuesta de clasificacion dialectal”, in: Thesaurus 37, 23-92. — (1985): Estudios sobre el español de Colombia. Bogotá: Instituto Caro y Cuervo. — (1992): “Historia del español en Colombia”, in: Hernández Alonso, César (coord.): Historia y presente del español de América. Valladolid: Junta de Castilia y León/ Pabecal, 501-517. — (1995): Dialectología general e hispanoamericana. Bogotá: Instituto Caro y Cuervo. — (1996): “Colombia”, in: Alvar, Manuel (dir.): Manual de dialectología hispánica. Barcelona: Ariel, 134-145. MONTES GIRALDO, José Joaquín/FIGUEROA LORZA, Jennie/MORA MONROY, Siervo/LOZANO RAMÍREZ, Mariano/RAMÍREZ CARO, Ricardo Aparicio/ESPEJO OLAYA, María Bernarda/DUARTE HUERTAS, Gloria Esperanza (1997): El español hablado en Bogotá: relatos semilibres de informantes pertenecientes a tres estratos sociales. Bogotá: Instituto Caro y Cuervo. MONTES GIRALDO, José Joaquín/FIGUEROA LORZA, Jennie/MORA MONROY, Siervo/LOZANO RAMÍREZ, Mariano/RAMÍREZ CARO, Ricardo Aparicio/ESPEJO OLAYA, María Bernarda/DUARTE HUERTAS, Gloria Esperanza (1998): El español hablado en Bogotá: análisis previo de su estratificación social. Bogotá: Instituto Caro y Cuervo. MONTOYA, María C. (2010): Expression of Possession in Spanish in Contact with English: A Sociolinguistic Study across two Generations in New York. Doctoral Dissertation. Albany: The State University of New York. MOÑINO, Yves (2002): “Las construcciones de genitivo en palenquero: ¿una semantaxis africana?”, in: Moñino, Yves/Schwegler, Armin (eds.): Palenque, Cartagena y AfroCaribe: historia y lengua. Tübingen: Niemeyer, 227-248. MOÑINO, Yves/SCHWEGLER, Armin (eds.) (2002): Palenque, Cartagena y Afro-Caribe: Historia y lengua. Tübingen: Max Niemeyer Verlag. — (2003): “Lengua e identidad afro-americana: el caso del criollo de Palenque de San Basilio (Colombia)”, in: Alès, Catherine/Chiappino, Jean (eds.): Caminos cruzados: ensayos en antropología social, etnoecología y etnoeducación. Paris/Mérida: IRDÉditions/Universidad de los Andes, GRIAL, 517-531. — (2007): “Convergencias lingüísticas iberocongolesas en Palenquero: ¿integrarse a la sociedad mayoritaria o distinguirse de ella?”, in: Mihatsch, Wiltrud/Sokol, Monika (eds.): Language contact and language change in the Caribbean and beyond/Lenguas en contacto y cambio lingüístico en el Caribe y más allá. Frankfurt: Peter Lang, 37-57. MORALES, Amparo (1995): “The Loss of the Spanish Impersonal Particle ‘se’ Among Bilinguals: A Descriptive Profile”, in: Silva-Corvalán, Carmen (ed.): Spanish in the Four Continents. Washington: Georgetown Universtiy Press, 148-162. MORENO F., Patricio (1985): “Deixis espacial en francés y español”, in: RLA: Revista de Lingüística Teórica y Aplicada 23, 157-162.

MURIEL/OROZCO

31/5/12

14:24

Página 251

REFERENCES

251

MORTON, Thomas (2005): Sociolinguistic Variation and Language Change: In El Palenque de San Basilio (Colombia). Doctoral Dissertation. Philadelphia: University of Pennsylvania. MUFWENE, Salikoko (2001): The Ecology of Language Evolution. Cambridge: Cambridge University Press. MUÑIZ CACHÓN, Carmen (1998): Impersonalidad y despersonalización: Estudio contrastivo. Oviedo: Universidad de Oviedo. NAVARRETE, María Cristina (2008): San Basilio de Palenque: memoria y tradición. Cali: Programa Editorial, Universidad del Valle. NILSSON, Kåre (1983): “En torno a los adverbios de lugar aquí, acá, etc. en castellano y sus formas correspondientes en catalán y portugués”, in: Actes de VIIIème Congrès des Romanistes Scandinaves, 257-268. NUÑEZ-CEDEÑO, Rafael (1982): “El español de Villa Mella, en desafío a las teorías fonológicas modernas”, in: Alba, Orlando (ed.): El español del Caribe. Santiago: Universidad Católica Madre y Maestra, 223-236. OBAID, Antonio H. (1973): “The Vagaries of the Spanish ‘S’”, in: Hispania 56, 60-67. OCHOA FRANCO, José (1945): Consideraciones generales sobre costumbres y lenguaje palenqueros. Cartagena: Dirección de Educación Pública de Bolívar. OLIVEIRA, Marco Antônio de/BRAGA, Maria Luiza (1996): “On Focusing Sentences in Brazilian Portuguese”, in: Guy, Gregory R./Feagin, Crawford/Shiffrin, Deborah/ Baugh, John (eds.): Towards a Social Science of Language: Papers in Honor of William Labov. Amsterdam/Philadelphia: John Benjamins, 2, 207-221. ORDÓÑEZ, Francisco (1997): Word Order and Clause Structure in Spanish and other Romance Languages. Doctoral Dissertation. New York: The City University of New York. ORLOV, Ann/UEDA, Reed (1980): “Central and South Americans”, in: Thernstrom, Stephan (ed.): Harvard Encyclopedia of American Ethnic Groups. Cambridge: Harvard University Press, 210-217. ORNSTEIN-GALICIA, Jacob/GREEN, George/BIXLER-MÁRQUEZ, Dennis (eds.) (1988): Research Issues and Problems in United States Spanish: Latin American and Southwestern Varieties. Brownsville: Pan American University at Brownsville. OROZCO, Rafael (2004): A Sociolinguistic Study of Colombian Spanish in Colombia and New York City. Doctoral Dissertation. New York: New York University/Ann Arbor: ProQuest. — (2005): “Distribution of Future Time Forms in Northern Colombian Spanish”, in: Eddington, David (ed.): Selected Proceedings of the 7th Hispanic Linguistics Symposium. Somerville: Cascadilla Proceedings Project, 56-65. — (2007a): “The Impact of Linguistic Constrains on the Expression of Futurity in the Spanish of New York Colombians”, in: Potowski, Kim/ Cameron, Richard (eds.): Spanish in Contact: Policy, Social and Linguistic Inquiries. Amsterdam/Philadelphia: John Benjamins. 311-327. — (2007b): “Social Constraints on the Expression of Futurity in Spanish-Speaking Urban Communities”, in: Holmquist, Jonathan/Lorenzino, Augusto/Sayahi, Lotfi

MURIEL/OROZCO

252

31/5/12

14:24

Página 252

REFERENCES

(eds.): Selected Proceedings of the Third Workshop on Spanish Sociolinguistics. Somerville: Cascadilla, 103-112. — (2009a): “El castellano del Caribe colombiano a comienzos del siglo XXI”, in: Revista Internacional de Lingüística Iberoamericana (RILI) 7, 2, 95-113. — (2009b): “La influencia de factores sociales en la expresión del posesivo”, in: Lingüística 22, 25-60. — (2010a): “Variation in the Expression of Possession in Costeño Spanish”, in: Spanish in Context 7, 2, 194-220. — (2010b): “A Sociolinguistic Analysis of three Linguistic Variables in two Communities: Barranquilla & New York”. Conference paper. Fifth International Workshop on Spanish Sociolinguistics. April 8, 2010: Raleigh. OROZCO, Rafael/GUY, Gregory (2008): “El uso variable de los pronombres sujetos: ¿Qué pasa en la costa Caribe colombiana?”, in: Westmoreland, Maurice/Thomas, Juan Antonio (eds.): Selected Proceedings of the Fourth Workshop on Spanish Sociolinguistics. Somerville: Cascadilla, 70-80. ORTIZ LÓPEZ, Luis/LACORTE, Manel (eds.) (2005): Contactos y contextos lingüísticos: el español en los Estados Unidos y en contacto con otras lenguas. Madrid/Frankfurt: Iberoamericana/Vervuert. OTHEGUY, Ricardo/ZENTELLA, Ana Celia (2007): “Apuntes preliminares sobre el contacto lingüístico y dialectal en el uso pronominal del español en Nueva York”, in: Potowski, Kim/Cameron, Richard (eds.): Spanish in Contact: policy, social and linguistic inquiries. Amsterdam/Philadelphia: John Benjamins, 275-295. OTHEGUY, Ricardo/ZENTELLA, Ana Celia/LIVERT, David (2007): “Language and Dialect Contact in Spanish in New York: Toward the Formation of a Speech Community”, in: Language 83, 4, 770-802. PALACIOS, Azucena (2005): “Aspectos teóricos y metodológicos del contacto de lenguas: el sistema pronominal del español en áreas de contacto con lenguas amerindias”, in: Noll, Volker/Zimmermann, Klaus/Neumann-Holzschuh, Ingrid (eds.): El español en América: Aspectos teóricos, particularidades, contactos. Madrid/Frankfurt: Iberoamericana/Vervuert, 63-92. — (2007a): “¿Son compatibles los cambios inducidos por contacto y las tendencias internas al sistema?”, in: La Romania en interacción: entre historia, contacto y política. Ensayos en homenaje a Klaus Zimmermann. 263-283. — (2007b): “Cambios lingüísticos de ida y vuelta: los tiempos de pasado en la variedad emergente de los inmigrantes ecuatorianos en España”, in: Revista Internacional de Lingüística Iberoamericana (RILI) 5, 2, 109-125. PARKVALL, Mikael (2000): Out of Africa: African Influences in Atlantic Creoles. London: Battlebridge Publications. PASTIZZO, Matthew J./CARBONE Jr., Robert F. (2007): “Spoken Word Frequency Counts Based on 1.6 million Words in American English”, in: Behavior Research Methods 39, 4, 1025-1028. PATIÑO ROSSELLI, Carlos (1983): “El habla en El Palenque de San Basilio”, in: Friedemann, Nina/Patiño Rosselli, Carlos: Lengua y sociedad en El Palenque de San Basilio. Bogotá: Instituto Caro y Cuervo, 85-287.

MURIEL/OROZCO

31/5/12

14:24

Página 253

REFERENCES

253

— (1986): “On the Linguistic Question of San Andres and Providencia,” in: Glotta 1, 2, 8-12. — (1999): “Aspectos de la estructura del criollo Palenquero”, in: Zimmermann, Klaus: Lenguas criollas de base lexical española y portuguesa. Madrid/Frankfurt: Iberoamericana/Vervuert, 205-230. — (1991): Español, lenguas indígenas y lenguas criollas en Colombia. Presencia y destino: el español de América hacia el siglo XXI, vol. I, 145-207. Bogotá: Instituto Caro y Cuervo. — (2002): “History and Society in the Origin of Creole Languages”, in: Revista de Estudios Sociales 13, 109-115. PAVY, David (1967): “The Provenience of Colombian Negroes”, in: The Journal of Negro History 52 (1), 35-58. PENNY, Ralph (2000): Variation and Change in Spanish. New York: Cambridge University Press. — (2002): A History of the Spanish Language, Second Edition. New York: Cambridge University Press. — (2008): Gramática histórica del español. Barcelona: Ariel. PÉREZ, Lisandro (1992): “Cuban Miami. Miami Now”, in: Grenier, Guillermo J./Stepick III, Alex (eds.): Immigration, Ethnicity and Social Change. Gainesville: University Press of Florida, 83-108. PFLEIDERER, Bettina (1998): Sprachtod und Revitalisierung der spanisch basierten Kreolsprache Palenquero (Kolumbien). Berlin: Freie Universität Berlin. PICALLO, M. Carme/RIGAU, Gemma (1999): “El posesivo y las relaciones posesivas”, in: Bosque, Ignacio/Demonte, Violeta (eds.): Gramática descriptiva de la lengua española. Madrid: Espasa Calpe, 973-1024. PIERREHUMBERT, Janet/TALKIN, David (1991): “Lenition of /h/ and Glottal Stop”, in: Docherty, Gerard/ Ladd, Robert (eds.): Papers in Laboratory Phonology II. Cambridge: Cambridge University Press. 90-117 PIÑEROS, Carlos-Eduardo (2003): “Accounting for the Instability of Palenquero Voiced Stops”, in: Lingua 113, 12, 1185-1222. PLACENCIA, María Elena (1997): “Address Forms in Ecuadorian Spanish”, in: Hispanic Linguistics 9, 165-202. PLUYMAEKERS, Mark/ERNESTUS, Mirjam/BAAYEN, R. Harald (2005): “Lexical Frequency and Acoustic Reduction in Spoken Dutch”, in: Journal of the Acoustical Society of America 118, 2561-2569. POPLACK, Shana (1979): Function and Process in a Variable Phonology. Doctoral Dissertation. Philadelphia: University of Pennsylvania. PORCEL, Jorge (2003): Tense, Modality and Aspect in the Spanish Verb. (A Description of the Usage of Present Progressive, Simple and Periphrastic Future, and Preterite and Imperfect of the Indicative in the Miami Cuban Community). Doctoral Dissertation. Pittsburgh: University of Pittsburgh. POUPLIER, Marianne/GOLDSTEIN, Louis (2005): “Asymmetries in the Perception of Speech Production Errors”, in: Journal of Phonetics 33, 47-75.

MURIEL/OROZCO

254

31/5/12

14:24

Página 254

REFERENCES

QUESADA PACHECO, Miguel Ángel (2000): El Español de América. Cartago: Editorial Tecnológica de Costa Rica. — (ed.) (2010): El español hablado en América Central. Madrid/Frankfurt: Iberoamericana/Vervuert. R DEVELOPMENT CORE TEAM. R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing; http://www.R-project.org. RAE, Real Academia Española (1973): Esbozo de una nueva gramática de la lengua española. Madrid: Espasa Calpe. RAMÍREZ, Dora (2003): “L1 regression in Spanish-English Bilinguals”, in: Sayahi, Lotfi (ed.): Selected proceedings of the First Workshop on Spanish Sociolinguistics. Somerville: Cascadilla Proceedings Project, 89-95. — (2007a): Language Attrition and Language Maintenance: The Case of Colombian Immigrants in New York State. Unpublished Doctoral Dissertation. Albany: State University of New York. — (2007b): “Subject Personal Pronouns and Impersonal Sentences in Adult Colombian Immigrants' Spanish”, in: Holmquist, Jonathan/Lorenzino, Augusto/Sayahi, Lotfi (eds.): Selected Proceedings of the Third Workshop on Spanish Sociolinguistics. Somerville: Cascadilla Proceedings Project, 122-128. RAYMOND, William D./DAUTRICOURT, Robin/HUME, Elizabeth (2006): “Word-Internal /t,d/ Deletion in Spontaneous Speech: Modeling the Effects of Extra-Linguistic, Lexical, and Phonological Factors”, in: Language Variation and Change 18, 55-97. REDI, Laura/SHATTUCK-HUFNAGEL, Stefanie (2001): “Variation in the Realization of Glottalization in Normal Speakers”, in: Journal of Phonetics 29, 407- 429 REPP, Bruno H./WILLIAMS, David R. (1985): “Influence of Following Context on Perception of the Voiced-Voiceless Distinction in Syllable-Final Stop Consonants”, in: Journal of the Acoustical Society of America 78, 2, 445-457. RESNICK, Melvyn C. (1975): Phonological Variants and Dialect Identification in Latin American Spanish. The Hague: Mouton. REY, Alberto (1994a): “Social Correlates of the Usage of the Spanish You Subject Pronouns in Colombia”, in: Hispanic Journal 15, 2, 291-305. — (1994b): “The Usage of Usted in Three Societies: Colombia, Honduras, and Nicaragua”, in: Language Quarterly 32, 193-204. REY CASTILLO, Marisol (2005): El uso de tú, ud, y sumercé como fórmulas de tratamiento en Funza, Cundinamarca. Masters Dissertation. Bogotá: Instituto Caro y Cuervo. RICHARDSON, Bill (1996): “Spanish Spatial Deictic Features: Indices of Entities, Location and Movement”, in: International Review of Applied Linguistics 34, 3, 215-231. RINCÓN, Luz Mary (2006): “Usted, the Pronominal Address of Familiarity in an Urban Community”, in: Hispanic Journal 27, 1, 11-22. ROCA, Ana (ed.) (2000): Research on Spanish in the United States: Linguistic Issues and Challenges. Somerville: Cascadilla Press. ROCA, Ana/COLOMBI, María Cecilia (eds.) (2003): Mi lengua: Spanish as a Heritage Language in the United States. Washington: Georgetown University Press. ROCA, Ana/LIPSKI, John (eds.) (1993): Spanish in the United States: Linguistic Contact and Diversity. Berlin: Mouton de Gruyter.

MURIEL/OROZCO

31/5/12

14:24

Página 255

REFERENCES

255

RODRIGUEZ-CASTELLANO, Lorenzo/PALACIO, Adela (1948): “El habla de Cabra”, in: ROTP 4, 378-418. ROSSOMONDO, Amy (2002): “Pronominal Address Forms in Madrid”, in: Lee, James/Geeslin, Kimberly/Clements, J. Clancy (eds.): Structure, Meaning, and Acquisition in Spanish: Papers from the 4th Hispanic Linguistics Symposium. Somerville: Cascadilla Press, 115-129. RUIZ MORALES, Hildebrando (1987): “Desplazamiento semántico en las formas de tratamiento del español de Colombia”, in: López Morales, Humberto/Vaquero, María (eds.): Actas del I Congreso Internacional sobre el Español de América. San Juan: Academia Puertorriqueña de la Lengua Española, 765-775. SACKS, Norman P. (1954): “Aquí, acá, allí, and allá”, in: Hispania 37, 3, 263-266. — (1987): “An Analysis of Spanish Adverbs of Place Beyond the Rest / Motion Distinction: Sound Symbolism, Relativity of Focus, and Correspondences to English there”, in: Morgan, Terrel/Lee, James F./Van Patten, Bill (eds.): Languages and Language Use: Studies in Spanish. Lanham: University Press of America, 289-315. SALCEDO RAMOS, Alberto/RODRÍGUEZ, Cynthia (2005): El oro y la oscuridad: la vida gloriosa y trágica de Kid Pambelé. Bogotá: Debate. SALVÁ, Vicente (1830/1988): Gramática de la lengua castellana. Madrid: Arco Libros. SAMPER PADILLA, José Antonio (1990): Estudio sociolingüístico del español de Las Palmas de Gran Canaria. Santa Cruz de Tenerife: La Caja de Canarias. SCHMIDELY, Jack (1975): “Déictiques spatiaux de l’espagnol”, in: Sephiha, Haïm Vidal (ed.): Mélanges offerts a Charles Vincent Aubrun. Paris: Editions Hispaniques, 239-252. SCHREFFLER, Sandra (1994): “Second-Person Singular Pronoun Options in the Speech of Salvadorans in Houston, TX”, in: Southwest Journal of Linguistics 13, 1-2, 101-119. SCHUCHARDT, Hugo Ernst Mario (1885 [1972]): “On Sound Laws: Against the Neogrammarians”, in: Vennemann, Theo/Wilbur, Terence H. (eds.): Schuchardt, the Neogrammarians, and the Transformational Theory of Phonological Change. Frankfurt: Athenäum Verlag, 39-72. SCHWARTZ, Martin F. (1970): “Duration of /s/ in /s/-plosive blends”, in: Journal of the Acoustical Society of America 47, 1143-1144. SCHWEGLER, Armin (1989): “Notas etimológicas palenqueras: casariambe, túngananá, agü., monicongo, maricongo y otras voces africanas y pseudo-africanas”, in: Thesaurus 44, 1-28. — (1990): Analyticity and Syntheticity: A Diachronic Perspective with Special Reference to Romance Languages. Berlin/New York: Mouton de Gruyter. — (1991): “El habla cotidiana del Chocó (Colombia)”, in: América Negra 2:85-119. — (1996): “Chi ma nkongo”: lengua y rito ancestrales en El Palenque de San Basilio (Colombia). Madrid/Frankfurt: Iberoamericana/Vervuert. 2 vols. — (2002): “On the (African) Origins of Palenquero Subject Pronouns”, in: Diachronica 19(2), 273-332. — (2007a): “A Fresh Consensus in the Marking: Plural ma and Bare Nouns in Palenquero”, in: Mihatsch, Wiltrud/Sokol, Monika (eds.): Language Contact and Language Change in the Caribbean and Beyond/Lenguas en contacto y cambio lingüístico en el Caribe y más allá. Frankfurt: Peter Lang, 59-75.

MURIEL/OROZCO

256

31/5/12

14:24

Página 256

REFERENCES

— (2007b): “Bare Nouns in Palenquero: A Fresh Consensus in the Making”, in: Baptista, Marlyse/Guéron, Jacqueline (eds.): Noun Phrases in Creole Languages: A Multi-Faceted Approach. Amsterdam/Philadelphia: John Benjamins, 205-222. — (2010): “State of the Discipline. Pidgin and Creole Studies: Their Interface with Hispanic and Lusophone Linguistics”, in: Studies in Hispanic and Lusophone Linguistics 3, 431-481. — (2011a): “Palenque (Colombia): Multilingualism in an Extraordinary Social and Historical Context”, in: Díaz-Campos, Manuel (ed.): The Handbook of Hispanic Sociolinguistics. Malden: Wiley/Blackwell Publishing, 446-472. — (2011b): “Palenque(ro): The Search for its African Substrate”, in: Lefebvre, Claire (ed.): Creoles: Their Substrates, and Language Typology. Amsterdam/Philadelphia: John Benjamins, 225-249. — (forthcoming): “Sobre el origen africano de la lengua criolla de Palenque (Colombia)”, in: Maglia, Graciela/Schwegler, Armin (eds.): Palenque (Colombia): oralidad, identidad y resistencia. Un enfoque interdisciplinario. Bogotá: Instituto Caro y Cuervo/Universidad Javeriana SCHWEGLER, Armin/GREEN, Kate (2001): “Palenquero (Creole Spanish)”, in: Holm, John/Patrick, Peter (eds.): Comparative Creole Syntax. London: Battlebridge, 273-306. SCHWEGLER, Armin/KEMPFF, Juergen/AMEAL-GUERRA, Ana (2010). Fonética y fonología españolas. Hoboken: John Wiley & Sons, Inc. SCHWEGLER, Armin/MORTON, Thomas (2003): “Vernacular Spanish in a Microcosm: Kateyano in El Palenque de San Basilio (Colombia)”, in: Revista Internacional de Lingüística Iberoamericana (RILI) 1, 97-159. SCHWENTER, Scott (1993): “Diferenciación dialectal por medio de pronombres: una comparación del uso de tú y usted en España y México”, in: Nueva Revista de Filología Hispánica 41, 1, 127-149. SEDANO, Mercedes (1988): “Yo vivo es en Caracas: Un cambio sintáctico”, in: Studies in Caribbean Spanish Dialectology. Washington: Georgetown University Press, 115-123. — (1990): Hendidas y otras construcciones con ser en el habla de Caracas. Caracas: Universidad Central de Venezuela. — (1994a): “Presencia o ausencia de relativo: Explicaciones funcionales”, in: Thesaurus 49, 3, 491-518. — (1994b): “Evaluation of Two Hypotheses about the Alternation between aquí and acá in a Corpus of Present-Day Spanish”, in: Language Variation and Change 6, 2, 223-237. — (1995a): “Los adverbios demostrativos en tres ciudades hispano hablantes”. Conference paper. V Congreso Internacional del español de América, Burgos. — (1995b): “A la que yo escribí fue a María vs. Yo escribí fue a María: El uso de estas dos estructuras en el español de Caracas”, in: Boletín de Lingüística 9, 51-80. — (1996): “Las posibilidades de ahí como elemento central del sistema locativo”, in: Arjona, Marina (ed.): Actas del X Congreso Internacional de la ALFAL. México: Universidad Nacional Autónoma de México, 103-108. — (1999a): “¿Ahí o allí? Un estudio sociolingüístico”, in: Serrano, María José (ed.): Estudios de la variación sintáctica. Madrid/Frankfurt: Iberoamericana/Vervuert, 51-63.

MURIEL/OROZCO

31/5/12

14:24

Página 257

REFERENCES

257

— (1999b): “Los adverbios demostrativos y anafóricos en siete lenguas románicas”, in: Samper Padilla, José Antonio/Troya, Magnolia (eds.): Actas del XI Congreso Internacional de la ALFAL. Las Palmas de Gran Canaria: Universidad de Las Palmas de Gran Canaria, 433-446. — (2000): “Los adverbios demostrativos en diez ciudades hispanohablantes: resultados de una encuesta”, in: Carbonero Cano, Pedro/Casado Velarde, Manuel/Gómez Manzano, Pilar (eds.): Lengua y discurso. Estudios dedicados al Profesor Vidal Lamíquiz. Madrid: Arco Libros, 805-826. — (2003a): “Más sobre seudohendidas y construcciones con verbo ser focalizador en el habla de Caracas”, in: Moreno Fernández, F. et al. (eds.): Lengua, variación y contexto: Estudios dedicados a Humberto López Morales, Madrid: Arco Libros 2, 823-847. — (2003b): “Seudohendidas y oraciones con verbo ser focalizador en dos corpus del español hablado de Caracas”, in: Revista Internacional de Lingüística Iberoamericana (RILI), 1, 1, 175-204. SILBERT, Noah/DE JONG, Kenneth (2008): “Focus, Prosodic Context, and Phonological Feature Specification: Patterns of Variation in Fricative Production”, in: Journal of the Acoustical Society of America 123, 5, 2769-2779. SILVA-CORVALÁN, Carmen (1994): Language Contact and Change: Spanish in Los Angeles. New York: Oxford University Press. — (ed.) (1995): Spanish in Four Continents: Studies in Language Contact and Bilingualism. Washington: Georgetown University Press. — (2004): “Spanish in the Southwest”, in: Finegan, Edward/Rickford, John (eds.): Language in the USA: Themes for the Twenty-first Century. Cambridge: Cambridge University Press, 205-229. SIMARRA OBESO, Rutsely/MIRANDA REYES, Regina/PÉREZ TEJEDOR, Juana Pabla (2008): Lengua ri Palenge jende suto ta chitiá. Cartagena: Casa Editorial C. I. Organización Digital. SIMARRA REYES, Luís/TRIVIÑO DOVAL, Álvaro Enrique (2008): Gramática de la lengua palenquera: introducción para principiantes. Cartagena: Grafipapel. SOLÉ, Yolanda (1970): “Correlaciones socio-culturales del uso de tú/vos y usted en la Argentina, Perú y Puerto Rico”, in: Thesaurus 25, 2, 161-195. — (1978): “Sociocultural Determinants of Symmetrical and Asymmetrical Address Forms in Spanish”, in: Hispania 61, 4, 940-949. STEWART, Miranda (1999): The Spanish Language Today. London: Routledge. STURNER, Pamela (1995): “Colombian Americans”, in: Galens, Judy/Sheets, Anna/ Young, Robyn (eds.): Gale Encyclopedia of Multicultural America, V. 1. New York: Gale Research Inc., 326-338 TAGLIAMONTE, Sali (2006): Analysing Sociolinguistic Variation. New York: Cambridge University Press. TERRADO PABLO, Xavier (1990): “Sobre la forma de contenido de los adverbios de lugar. Cuestiones de diacronia”, in: Sintagma 2, 55-66. TERRELL, Tracy D. (1977a): “Universal Constraints on Variably Deleted Final Consonants: Evidence from Spanish”, in: The Canadian Journal of Linguistics/La Revue canadienne de Linguistique 22, 156-168.

MURIEL/OROZCO

258

31/5/12

14:24

Página 258

REFERENCES

— (1977b): “Constraints on the Aspiration and Deletion of Final /s/ in Cuban and Puerto Rican Spanish”, in: The Bilingual Review/La revista bilingüe 4, 35-51. — (1978a): “La aspiración y elisión de /s/ en el español porteño”, in: Anuario de Letras 16, 41-66.” — (1978b): “Sobre la aspiración y elisión de /s/ implosiva y final en el español de Puerto Rico,” in: Nueva Revista de Filología Hispánica 27, 24-38. — (1979): “Final /s/ in Cuban Spanish”, in: Hispania 62, 599-612. — (1981): “Diachronic Reconstruction by Dialect Comparison of Variable Contraints: S-Aspiration and Deletion in Spanish”, in: Sankoff, David: Variation Ombibus. Carbondale: Linguistic Research. — (1986): “La desaparición de /s/ posnuclear al nivel léxico en el habla dominicana”, in: Estudios sobre la fonología del español del Caribe. Caracas: Ediciones La Casa de Bello, 117-134. THOMASON, Sarah (2000): “Linguistic Areas and Language History”, in: Gilbers, Dicky/Nerbonne, John/Schaeken, Jos (eds.): Languages in Contact. Amsterdam: Rodopi, 311-327. — (2001): Language Contact: An Introduction. Washington: Georgetown University Press. THOMASON, Sarah/KAUFMAN, Terrence (1988): Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press. TORIBIO, Almeida Jaqueline (1992): “Proper Government in Spanish Subject Relativization”, in: Probus 4, 291-304. — (2002): “Focus on Clefts in Dominican Spanish”, in: Lee, James/Geeslin, Kimbery/ Clements, J. Clancy (eds.): Structure, Meaning, and Acquisition of Spanish, Somerville: Cascadilla Press, 130-146. TORRES, Lourdes (1991): “The Study of U.S. Spanish Varieties: Some Theoretical and Methodological Issues”, in: Klee, Carol/Ramos-García, Luis (eds.): Sociolinguistics of the Spanish-Speaking World: Iberia, Latin America, United States. Tempe: Bilingual Press/Editorial Bilingüe, 255-270. TOVAR, Lionel Antonio (2006): “Units of Linguistic Analysis in the Visual and Gestural Modality”, in: Lenguaje 34: 15-70. TRAUGOTT, Elizabeth Closs (2003): “Constructions in Grammaticalization”, in: Joseph, Brian D./Janda, Richard D. (eds.): Handbook of Historical Linguistics. Oxford: Blackwell Publishing, 624-647. TRAVIS, Catherine E. (2002): “La metalengua semántica natural: The Natural Semantic Metalanguage of Spanish”, in: Goddard, Cliff/Wierzbicka, Anna (eds.): Meaning and Universal Grammar: Theory and Empirical Findings. Volume I. Amsterdam/Philadelphia: John Benjamins, 173-242. — (2005): Discourse Markers in Colombian Spanish: A Study in Polysemy. Berlin/New York: Mouton de Gruyter. UBER, Diane (1984): “The Pronouns of Address in the Spanish of Bogotá, Colombia”, in: The SECOL Review 8, 59-74. — (1985): “The Dual Function of Usted: Forms of Address in Bogotá, Colombia”, in: Hispania 68, 2, 388-392.

MURIEL/OROZCO

31/5/12

14:24

Página 259

REFERENCES

259

— (1999): “Forms of Address in the Commercial Spanish of Five Latin American Cities”, in: Gutiérrez-Rexach, Javier/Martínez-Gil, Fernando (eds.): Advances in Hispanic Linguistics: Papers from the 2nd Hispanic Linguistics Symposium. Somerville: Cascadilla Press, 110-118. — (2000): “Addressing Business in Puerto Rico: Tú vs. Usted”, in: Roca, Ana (ed.): Research on Spanish in the United States: Linguistic Issues and Challenges. Somerville: Cascadilla Press, 310-318. — (2011): “Forms of Address: The Effect of the Context”, in: Díaz-Campos, Manuel (ed.): The Handbook of Hispanic Sociolinguistics. Oxford: Blackwell Publishing, 244-262. US CENSUS BUREAU (2010): American Fact Finder. (Last consulted: 19 july 2011) VALDÉS, Guadalupe/LOZANO, Anthony/GARCÍA-MOYA, Rodolfo (eds.) (1981): Teaching Spanish to the Hispanic Bilingual: Issues, Aims, and Methods. New York: Teachers College, Columbia University. VALENTÍN-MÁRQUEZ, Wilfredo. (2006): “La occlusion glotal y la construcción lingüística de identidades sociales en Puerto Rico”, in: Sagarra, Nuria/Almeida, Jacqueline (eds.): Selected Proceedings of the 9th Hispanic Linguistics Symposium, 326-341 VAN NAERSEN, Margaret (1983): “Ignoring the Reality of the Future in Spanish”, in: Bailey, Kathleen/Long, Michael/Peck, Sabrina (eds.): Second Language Acquisition Studies. Rowley: Newbury House, 56-67. VELÁSQUEZ, Rogerio (1960): Breve historia del Chocó 1500-1852. Unpublished manuscript. Bogotá: Instituto Colombiano de Antropología e Historia. VERA LUJÁN, Agustín (1990): Las construcciones pronominales pasivas e impersonales en español. Murcia: Universidad de Murcia. VIDA, Matilde (2004): Estudio sociofonológico del español hablado en la ciudad de Málaga. Alicante: Universidad de Alicante. VILA, María Rosa (1987): “La segunda persona gramatical en función no deíctica”, in: Revista Española de lingüística 17, 57-68. VILLA MEJÍA, Victor. (2001): Política idiomática en Colombia: visión sociolingüística. Medellín: Editorial Universidad de Antioquia. WEINREICH, Uriel (1967): Languages in Contact: Findings and Problems. London/The Hague/Paris: Mouton. WEST, Robert (1957): Las tierras bajas del Pacífico de Colombia. [Translated by Claudia Leal (2000)]. Bogotá: Instituto Colombiano de Antropología e Historia. WHERRITT, Irene/GARCÍA, Ofelia (eds.) (1989): US Spanish: The Language of Latinos. International Journal of the Sociology of Language 79. WIDDISON, Kirk A (1991): The Phonetic Basis for S-Aspiration in Spanish. Doctoral Dissertation. Berkeley: University of California. — (1994): “An investigation into the acoustic correlates of s-aspiration in Spanish,” in: Theoría 3, 113-118. — (1995): “Two models of Spanish s-aspiration”, in: Language Sciences 17, 4, 329-343. YANAGIDA, Yuko (1995): Focus projection and wh-head movement. Doctoral Dissertation. Ithaca: Cornell University.

MURIEL/OROZCO

260

31/5/12

14:24

Página 260

REFERENCES

YUAN, Yi (2001): “An Inquiry into Empirical Pragmatics Data-Gathering Methods: Written DCTs, Oral DCTs, Field Notes, and Natural Conversations”, in: Journal of Pragmatics 33, 2, 271-292. ZAGONA, Karen (2002): The Syntax of Spanish. Cambridge: Cambridge University Press. ZAMORA MUNNÉ, Juan C./GUITART, Jorge M. (1982): Dialectología Hispanoamericana: Teoría, Descripción, Historia. Salamanca: Ediciones Almar. ZENTELLA, Ana Celia (1990): “Lexical leveling in four New York City Spanish dialects: Linguistic and social factors”, in: Hispania 73, 4, 1094-1105. — (1997a): “Spanish in New York”, in: García Ofelia/Fishman, Joshua (eds.): The multilingual apple: Languages in New York City. New York: Mouton de Gruyter, 167-201. — (1997b): Growing up Bilingual: Puerto Rican Children in New York. Cambridge: Blackwell Publishers. — (2002): “Latin@ Languages and Identities”, in: Suárez-Orozco, Marcelo/Páez, Mariela (eds.): Latinos: Remaking America. Berkeley: University of California Press, 321-338. — (2004): “Spanish in the Northeast”, in: Finegan, Edward/Rickford, John (eds.): Language in the USA: Themes for the Twenty-first Century. Cambridge: Cambridge University Press, 182-204. ZIPF, George Kingsley (1935): Selected studies of the principle of relative frequency in Language. Cambridge: Cambridge University Press.

MURIEL/OROZCO

31/5/12

14:24

Página 261

EDITORS

Richard J. File-Muriel Assistant Professor of Spanish at the University of New Mexico (Ph.D. Indiana University Bloomington), his research focuses on how frequency of use impacts the production and perception of language, primarily segmental phenomena. He is also interested in questions of research methodology, such as how one defines lexical frequency and how the composition of a corpus can influence lexical frequency counts. His research has been published in journals such as Hispania, Canadian Journal of Linguistics, and Language Variation and Change. Rafael Orozco Assistant Professor of Spanish Linguistics at Louisiana State University (Ph.D. New York University, 2004) and a native of the Colombian Caribbean region. His teaching and research interests include Spanish sociolinguistics, Colombian Spanish, Caribbean Spanish, and Spanish in the United States. He is coeditor of the volume Lenguaje, arte y revoluciones ayer y hoy: New Approaches to Hispanic Linguistic, Literary, and Cultural Studies (Cambridge Scholars Publishing: 2011). His work has been published in journals such as Revista Internacional de Lingüística Iberoamericana (RILI), Spanish in Context, and Lingüística, as well as in several edited collections.

AUTHORS Earl K. Brown Assistant Professor of Spanish Linguistics in the Department of Modern Languages at Kansas State University (Ph.D. University of New Mexico, 2008). His research centers on language variation and change, especially as it pertains to the sounds of Spanish. The effect of frequency on reductive processes in language, as detected in the statistical analysis of large numbers of tokens, is also central to his research. Esther L. Brown Assistant Professor of Hispanic Linguistics in the Department of Spanish and Portuguese at the University of Colorado Boulder (Ph.D. University of New

MURIEL/OROZCO

262

31/5/12

14:24

Página 262

AUTHORS

Mexico, 2004). She studies linguistic variation from a usage-based perspective. Her work identifies and addresses the significant correlations between variation and use, principally between phonological reduction and word and discourse context frequencies. José Alejandro Correa Researcher at the Instituto Caro y Cuervo in Bogotá, Colombia. His linguistic interests include dialectology, phonetics, phonology and the intersection between these areas. He is currently investigating the intonation of Colombian varieties of Spanish and is also pursuing his postgraduate studies in Phonetics and Phonology at Consejo Superior de Investigaciones Científicas (Madrid/España). Timothy Jowan Curnow Program Director in the Research Centre for Languages and Cultures at the University of South Australia (Ph.D. Australian National University, 1998). His research focuses on the grammar of spoken Spanish, on the grammar and history of the Barbacoan language Awa Pit spoken in Colombia and Ecuador, and on language education and language policy, particularly in Australian schools. Luz Marcela Hurtado Cubillos Professor of Spanish in the Department of Foreign Languages, Literatures, and Cultures at Central Michigan University (M.A. Instituto Caro y Cuervo, Colombia, 1995; Ph.D. University of Florida, 2001). She has published articles and presented papers at national and international conferences in the areas of Spanish in the US, language variation and change, and second language acquisition. Her work focuses on sociolinguistic variation in Colombian Spanish, and has been published in journals such as the Modern Language Journal, Hispania, Litterae, as well as in edited collections. Scott Lamanna Assistant Professor of Spanish at Calvin College in Grand Rapids, MI. He received his M.A. and Ph.D. in Linguistics and Spanish from Indiana University. He works in the areas of contact linguistics, sociolinguistics, functional syntax, and second language acquisition. His current research focuses on the role of language and dialect contact in shaping varieties of Spanish in the United States. John M. Lipski Edwin Erle Sparks Professor of Spanish and Linguistics in the Department of Spanish, Italian and Portuguese at the Pennsylvania State University. He is the author of several books, including Latin American Spanish, Varieties of Spanish

MURIEL/OROZCO

31/5/12

14:24

Página 263

AUTHORS

263

in the United States, El español de América, and A History of Afro-Hispanic Language Contact. Dunia Catalina Méndez Vallejo Teaching Assistant Professor in the Department of World Languages, Literatures, and Linguistics at West Virginia. She received a dual Ph.D. in Linguistics and Hispanic Linguistics from Indiana University in 2009. While she specializes in syntax, she has also conducted research on the effects of prosody in requests and Spanish word order, sociolinguistic variation in Spanish future tenses, and sociopragmatic change in discourse markers and forms of address. Her current research projects focus on the socio-cultural meanings of the word marica in Colombian varieties, and the semantic and pragmatic features of the Focalizing Ser structure. Armin Schwegler Associate Editor of the Journal of Pidgin and Creole Languages and Co-Editor of Revista Internacional de Lingüística Iberoamericana (RILI), his research explores “marginal” speech communities that are of great historical importance to understanding the evolution of American Spanish. Included among these communities is the maroon village of El Palenque de San Basilio (Colombia). His publications include several books and about 50 scholarly articles. Catherine E. Travis Professor of Modern European Languages in the School of Language Studies at the Australian National University (Ph.D. La Trobe University, Australia, 2002). Her research focuses on the grammar of spoken discourse, in particular conversation, and through the study of variation seeks to improve our understanding of the ways in which grammar emerges from language use. She has worked with both contact and non-contact varieties of Spanish (namely Colombian and New Mexican Spanish), as well as English and Portuguese.

MURIEL/OROZCO

31/5/12

14:24

Página 264