Spanish Language Use and Public Life in the United States [Reprint 2015 ed.] 9783110852530, 9783110096286

Citation preview

Spanish Language Use and Public Life in the United States

Contributions to the Sociology of Language 35

Editor Joshua A. Fishman

MOUTON PUBLISHERS • BERLIN • NEW YORK • AMSTERDAM

Spanish Language Use and Public Life in the United States

Edited by

Lucía Elias-Olivares Elizabeth A. Leone René Cisneros John R. Gutiérrez

MOUTON PUBLISHERS • BERLIN • NEW YORK • AMSTERDAM

CIP-Kurztitelaufnahme

der Deutschen

Bibliothek

Spanish language use and public life in the United States / ed. by Lucia Elias-Olivares . . . — Berlin ; New York ; Amsterdam : Mouton, 1985. (Contributions to the sociology of language ; 35) ISBN 3 11 009629 3 Pb. ISBN 3-11-009628-5 geb. NE: Elias-Olivares, Lucia [Hrsg.]; GT

Library of Congress Cataloging in Publication Data

Main entry under title: Spanish language use and public life in the United States. (Contributions to the sociology of language ; 35) I. Spanish language - United States - Addresses, essays, lectures. 2. Spanish language — Social aspects — United States — Addresses, essays, lectures. I. Elias-Olivares, Lucía. II. Series. PC4826.S66 1985 460'.973 85-4836 ISBN 0-899-25054-8

Printed on acid free paper

© Copyright 1985 by Walter de Gruyter & Co., Berlin. All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced in any form - by photoprint, microfilm, or any other means - nor transmitted nor translated into a machine language without written permission from the publisher. Typesetting: Copo Typsetting, Bangkok/ Werksatz Marschall, Berlin. - Printing: Drukkerei Hildebrand, Berlin. - Binding: Lüderitz & Bauer Buchgewerbe GmbH, Berlin. Printed in Germany.

Contents

Introduction Section 1 : ON DESCRIBING SPANISH-SPEAKING COMMUNITIES

1

11

Mary Beth Floyd Spanish in the Southwest: Language Maintenance or Shift?

13

John J. Attinasi Hispanic Attitudes in Northwest Indiana and New York

27

Pedro Pedraza Language Maintenance Among New York Puerto Ricans

59

Lucinda Hart-González Pan-Hispanism and Subcommunity in Washington, D.C.

73

Elizabeth A. Leone and René Cisneros Children's Storytelling Strategies

89

Maryellen Garcia and Georgeanne Weiler In the Match Between Spanish Dialects, Who is the Referee?

113

Section 2: USING SPANISH IN PUBLIC LIFE Joan Rubin Spanish Language Planning in the United States Deborah Martinez, Elizabeth A. Leone and Jennifer Sternbach de Medina language as a Barrier to Health Care Florence Barkin-Riegelhaupt Testing Bilingual Language Proficiency: An Applied Approach

133

153

165

Contents

Etilvia Aijona The Court Interpreters Certification Test Design

181

Beverly B. McConnell Bilingual Education and Language Shift

201

John M. Lipski Spanish in United States Broadcasting

217

Epilogue

229

LIST OF CONTRIBUTORS Etilvia Aijona School of Education Stanford University

Elizabeth A. Leone Department of Foreign Languages San Antonio College

John Attinasi Department of Bilingual Education Indiana University-Northwest

John M. Lipski Department of Spanish University of Houston

Florence Barkin-Riegelhaupt Department of Foreign Languages Arizona State University

Deborah Martinez Graduate Student School of Public Health University of Minnesota

René Cisneros Edgewood Independent School District San Antonio, Texas Lucía Elias-Olivares Department of Spanish, Italian and Portuguese University of Illinois at Chicago Mary Beth Floyd Department of Foreign Languages Northern Illinois University Maryellen García National Center for Bilingual Research at Los Alamitos, California John Gutiérrez Department of Spanish University of Virgnia Lucinda Hart-González Department of Linguistics Oakland University

Beverly McConnell Individualized Bilingual Instruction Project Pasco School District Pullman, Washington Pedro Pedraza Center for Puerto Rican Studies Hunter College - C.U.N.Y. Joan Rubin Joan Rubin and Associates Pinole, California Jennifer Sternbach de Medina Graduate Student School of Public Health University of Minnesota Georgeanne Weiler Centro de Investigación para la Integración Social Mexico City, Mexico.

Introduction

The Hispanic population in the United States is like a sleeping giant that has only recently begun to awaken, and it is imminent that as this giant stirs, it will continue only to get bigger. Although we may have to wait until the next decade for the full political and social consequences, the statistics are indicative of the tremendous impact that this is likely to have. By the middle of the 1990s, Hispanics will certainly overtake blacks to become the largest minority. The U.S. Bureau of the Census estimated that there were more than 14 million Hispanics in the country in 1980. Add to this figure the estimated 6 million undocumented Hispanics and the number grows to roughly 20 million - almost 10 percent of the entire US population. If we were to look at this from another perspective, i.e., strictly in terms of numbers, after Mexico, Spain, and Colombia, the United States would rank as the fourth largest Spanish-speaking country in the world today. The statistics also tell us that the median age for this group of people is 32.2, that only 4.9 percent are over 65 and that 32 percent are under 15. If we add to this the fact that the birthrate for Hispanics in the United States is twice that of whites and 60 percent higher than that of blacks and that Hispanic immigration to this country is running at about one million per year, if one includes the legal with the undocumented, it can clearly be seen why demographers are projecting that the number of Hispanics will increase by • at least 58 percent to 23 million in the next two decades making the Hispanics the most numerous minority in the United States. Knowing that the unprecedented growth of the Hispanic population requires a number of local policy responses for meeting the sociolinguistic needs of their speakers, we have compiled this collection of papers focusing on the description of U.S. Spanish and its use in public domains. The authors examine Spanish varieties, language attitudes, language maintenance and shift, and language assessment, and discuss the use of Spanish in the health, legal, educational and mass media settings. In doing so, the authors have considered linguistic norms, translation, the development of test instruments and document design. The papers demonstrate the variety of theoretical approaches and methodological strategies currently being applied to the study of Spanish variation in Hispanic communities, on issues dealing with bilingualism, and on the identification of language planning needs. Some of these papers represent pioneering work in the field of Spanish language planning, and as such constitute

2

Introduction

research in progress, which points to the need for more basic descriptive studies dealing with the sociolinguistic characteristics of the bilingual speech community and leading to the identification and assessment of issues related to language and public life. Much of this pioneering work is necessarily qualitative as basic constructs have not been developed to the point where quantitative methods would be appropriate. We see language planning as the need for a broad identification of issues (Ferguson, 1979), its main goal as the identification of "the concrete areas of society that demand action regarding linguistic resources" (Jernudd and Das Gupta, 1971: 197), and as a need for a more broadly based language policy that respects the linguistic rights of linguistic minorities (Macias, 1979). Within this framework, the role of language has to be considered in the areas of governmental services, public education, media, and the delivery of social services to Hispanics. The need for the protection of the rights to access to all human services is particularly acute in contacts with the legal and medical processes. The inability to communicate effectively within a courtroom denies Hispanics full participation in the legal proceedings (Pousada, 1979). And in health, the healing process must be built on a comprehensive approach that includes the patient, his family, and his culture, which is not possible when we even lack the most basic tool of human communication — a common language and culture. Several papers in this volume address the question of Spanish being maintained in Hispanic communities, especially among the younger generations. Studies done in the Southwest offer strong evidence of intergenerational language shift from Spanish to English or from a monolingual community to a bilingual one. Greater use of English is associated with a greater degree of assimilation on the part of the bilingual. Studies conducted with other populations, however, especially in New York and the Midwest, seem to present a different picture of bilingualism, and brighter prospects for the maintenance and development of Spanish. The trend which is slowly emerging from these studies — which utilize a combination of quantitative and qualitative methods of research, this latter dimension lacking in the studies done in the Southwest, is one not of total language shift, or even diglossia, but of a more stable type of bilingualism. The choice is not English or Spanish, but both languages. These communities are neither abandoning Spanish nor resisting the learning of English. Furthermore, competence in both languages allows for a bilingual mode of communication - code-switching - which is as systematic and rule-governed as any other communication system (Poplack, 1982). This sociolinguistic information is extremely relevant to the educational arena, especially in the relatively new field of bilingual education, because as Rubin points out, language planning requires "the right kinds of

Introduction

3

information about the sociolinguistic habits of the target population and about the social basis for language policy in order to project productive directions for change" (Rubin, 1973 :v). An interesting finding of recent studies is the widespread support for bilingual education among those Hispanic researched, and this support is for a type of program which not only allows the child to become proficient in English but also helps him to maintain and develop the skills possessed in his original first language. This concept of bilingual education is clearly at odds with present language policies in the United States. Transitional bilingual programs, which have been legislated with little or no community input, provide for programs whose primary goal is not bilingual development but transition into English monolingualism, whereas the populations surveyed seem to consider the maintenance of cultural and linguistic resources as a fundamental right and obligation. At present there is no explicit national language policy in the United States for Hispanics, due to a historical approach under which federal policies have been rejected in favor of local ones. As Ferguson (1979) says, language rights for minorities have a better chance if they are pressed by the communities themselves. This is precisely what some communities have tried to do. The National Puerto Rican Task Force on Educational Policy, for example, has adopted the following 'agenda for discussion' to address the education of Puerto Rican children: (1) full bilinguality beyond the first generation; (2) spoken and written command of Spanish and English without downgrading nonstandard dialects and code-switching; (3) languages of instruction based on sensitivity to the community's needs and usage; and, (4) involvement in language events of the artistic, professional and organizational life of the community (NPRTFEP, 1977: 30-2). It should also be remembered that only community acceptance will ensure the success of policy implementation. Zentella (1980) points out that a language policy most likely to be accepted by a community is one based on previous research which has determined what is necessary, what is feasible and what is desirable, in community terms. Such a community-based language policy which is not dictated 'from the top', and which accomodates the desires and goals of the communities in question would involve profound social and philosophical changes, which may not take place until the speech communities concerned acquire a substantial amount of economic and political power to make known these needs and goals.

4

Introduction

On Describing Spanish-Speaking Communities The papers that comprise this section reveal the research efforts that are currently being carried out in the areas of Spanish language maintenance (Floyd, Attinasi, Pedraza) as well as in the description of the use of Spanish and English in the United States today (Hart-Gonzalez, Leone and Cisneros, Garcia and Weller). In the first three we are able to discern a difference in both the methodological approach as well as in the results that were obtained by the investigators. In the Southwest, for example, all studies point toward an overt language shift, while in the Midwest and East, Spanish seems to have stabilized itself alongside English. The last four papers deal with the actual description of the use of Spanish among Hispanics in California, along the Texas-Mexico border, in Minnesota and in Washington, D.C. The fact that all seven articles deal with the maintenance and description of the use of Spanish in the United States is essential in that this is the first step in the direction of language planning, for only after we are cognizant of the degree of maintenance and shift as well as the linguistic habits of the communities involved, can we begin to formulate a program of language planning that can indeed be implemented. In the first article, Mary Beth Floyd reviews the findings of several studies on language maintenance, one of which involved Puerto Ricans and Cubans while the rest examined Chicanos in Texas, New Mexico, California and Colorado. These studies, all quantitative in nature, were based on self-reporting techniques, that is, subjects indicated when, where and with whom they thought they used Spanish and English. The studies focused on language maintenance and shift and their correlation to such factors as generation, age, childhood language, language used at home, formal study of Spanish, level of education and sex. All of the studies provide evidence of an overt language shift in process and Floyd suggests that loss of Spanish may also be an index of the level of assimilation to the dominant English-based culture. In his study, John Attinasi compares the attitudes of three groups of Hispanics in the United States, these being a group of Puerto Ricans from New York, a group of Hispanic school teachers from New York and a group of Puerto Ricans and Mexican-Americans from Northwest Indiana. The subjects of this study were asked about their attitudes toward the English and Spanish languages and how these relate to being Hispanic, language maintenance and shift, and bilingual education. The results of his study suggest that the communities studied are not experiencing a total shift to English but a stabilized level of bilingualism with greater fluency in English. All the subjects are overwhelmingly in favor of Spanish being maintained and many feel that schools might be in an excellent position to enhance language maintenance.

Introduction

5

Pedro Pedraza, vis-à-vis Attinasi and Floyd, looks at various theoretical questions that are involved in the studies of Spanish language maintenance in the United States. Most, like those reported by Floyd, employ self-report methods, are quantitative and involve a micro-level analysis. The research on maintenance that is currently being carried out at the Centro de Estudios Puertorriqueños, CUNY, from which Pedraza's research emanates, involves the actual observance of members of the speech community in interaction with other bilinguals, as well as Spanish and English monolinguals. He signals the need for a qualitative micro-level analysis in future studies of language maintenance and shift, that is, we must look beyond what is merely quantitatively linguistic to a qualitative analysis at the political, economic, social and historic levels and how these interrelate to issues of language maintenance. He goes on to report that Puerto Ricans in New York are neither losing Spanish nor resisting English and like the subjects in Attinasi's study, have achieved a state of bilingualism with greater fluency in English. The stabilized use of Spanish is also enhanced, according to Pedraza's results, by the circulatory migration patterns of Puerto Ricans between New York and the Island. In a study that based itself in the heterogeneous Hispanic population of Washington, D.C., a speech community that has received virtually no attention from linguists, Hart-González sought to review behavior and attitudes of Hispanics and correlate this with group allegiance. She suggests that the Hispanic speech community of Washington is referentially bound by self and group identification, similar attitude structure and tolerance and minimization of difference. Experiential ties in the larger community include common norms of interaction and receptive competence in a common base of Spanish. This common base includes elements of English not found in the original group dialects, thus reflecting a mode of communication that is mutually comprehensible by D.C. Hispanics who share an English-based experience. This research is significant in that Hart-González may be describing a model for other groups of Hispanics in urban settings across the United States, i.e., where one finds groups of Hispanics, whose origins are different Spanishspeaking countries, living and interacting with one another, so one might also find a Spanish-based pan-Hispanic mode of communication. Leone and Cisneros assess the Spanish oral language skills of a group of Chicano children between the ages of 8 and 12 in Minnesota as each retold a stoiy to which they had previously listened on tape. The children's communicative competence in story retelling is described with the framework of the ethnography of communication and several of their strategies such as verb usage, use of conjunction and language switches are analyzed. The authors stress the importance of understanding the concept of communicative competence in language planning efforts, and suggest that planners could employ the results of studies of this nature as a base on which to

6

Introduction

further build the Spanish skills of Hispanic children. They also call for a need to develop methods for the ongoing development of these skills at the secondary and post-secondary levels and quite possibly effectuate the maintenance of Spanish in the adult lives of these children. In the last paper in this section, Garcia and Weller, in a study of linguistic norms, undertook to survey the awareness of Mexicans (from Mexico) and Mexican-Americans as to what they consider to be correct usage with regard to a number of Spanish constructions that were taken from actual interviews with Mexicans and Mexican-Americans. Their objective in doing this was to discover those areas where Mexican Spanish and Mexican-American Spanish overlap. Their study also uncovered Mexican attitudinal data on MexicanAmerican Spanish, thus accounting for some of the stigmatization of Mexican-American Spanish by Mexicans. The authors' implications for education are twofold: in the match between Spanish dialects there is no single referee because there appear to be two arenas, while in teaching writing to Hispanic bilinguals a more formal writing style is needed. In sum, the papers in this section are indicative of the investigation that still awaits researchers of Spanish/English bilingualism. Only with a basis in a sound and thorough body of research that uses a combination of quantitative and qualitative methods to collect attitudinal data, to describe the use of both languages and to accurately assess language maintenance and shift, can language planners proceed with the monumental task that awaits them.

Using Spanish in Public Life The Spanish language planning and language policy studies presented in this section demonstrate the variety of steps involved in conducting and formulating language planning. Rubin begins by offering a comprehensive discussion of Spanish language planning in the U.S. from a present-day perspective. She considers the identified language/communication inadequacies, the planners with the authority to make or influence language decisions, and the plans or goals which have been set to meet these inadequacies in the domains of health, law, employment, communication, citizenship, social welfare and education (Rubin). By establishing a paradigm for looking at the inadequacies, the planners, the plans, and the actual attempts at planning, Rubin effectively reviews the public domains, showing some significant language planning in Spanish that is taking place across these domains. In the many examples of language planning cited, there have always been a variety of groups interacting to influence the decision-making, and a variety of processes by which language

Introduction

1

planning becomes official policy. Rubin points out that often language policy becomes necessary in order to protect substantial rights which would otherwise be violated, and not necessarily to provide for language rights in and of themselves. All her examples are ones demonstrating instrumental or pragmatic functions of language — where communication is necessary for services to be provided. She discusses different policy processes, implementation being difficult when decisions are made in the judicial arena, and emphasizes that in the majority of language planning situations, the actors are not so clearly defined, and for this reason, the planning may be less effective. If persons who will eventually become involved in implementation participate early in the planning process, then the correcting of language inadequacies in the public domains can be more effective. This is followed by a statement of needs (Martinez, Leone, and Sternbach de Medina); a discussion of assessment, design and implementation at the state level (Barkin) and at the federal level (Aijona); a study documenting the success of a language policy, plan and program, and clarifying the needs of two distinct Chicano communities in the United States (McConnell); and finally, a discussion of the discovery and setting of standards in U.S. Spanish (Lipski). All six papers share a concern with meeting the language and communication needs of U.S. Hispanics in public domains. The papers in this section, while having certain elements in common, also vary in many aspects, such as approach or methodology. Aijona's paper begins with the impetus for the court certification project: a 1978 public law mandating 'more effective use of court interpreters', then proceeds to discuss the nine stages of the development of the court interpreter's certification examination, including the involvement of a diverse group of professionals, government officials, and native speakers who participated in editing, piloting, and actual certification. Both the oral and written sections of the exam consisted of a variety of specific tasks, and so rating scales and trained raters were also needed. By 1982, the impact of the court interpreter's certification legislation has been great: the number of full-time court interpreter positions has doubled (from 15 to 32), the salary and status classification of court interpreters has also doubled, and the way has been paved for the application of this same process to other languages such as sign language, Native American languages, and others. At the regional or state level, a process similar to the one documented by Aijona has been reported by Barkin. In order to deal with a pressing need in the state of Arizona, the Spanish Language Task Team designed a Spanish language proficiency exam for present and future teachers in bilingual programs in the state, calling on a number of institutions of higher education, persons involved in the bilingual education programs, and members of the communities of bilinguals in the state. Because of the specific communi-

8

Introduction

cative purpose involved in teaching bilingually, the Spanish Language Task Team had to determine some specific language functions needed by the bilingual teacher, and the determination of these functions depended on observations made from an ethnographic perspective. As with Aijona's study, Barkin accounts the process of applying language and communication information to a concrete situation, giving necessary details on the various steps involved in the task. Martinez, Leone, and Sternbach de Medina's study illustrates how one community has dealt with a documented need for health care that meets the language and cultural needs of Hispanic persons in the 'twin cities' of Minneapolis and St. Paul. A brief review of documentation by social scientists and health care workers in Hispanic communities in the U.S. places the study by Martinez et al. within the context of health care needs in general for the Spanish-speaking. The paper then gives basic community statistics, revealing a population of about 35,000 persons who are Hispanic, the majority of whom do use Spanish. Finally, the community organization that has begun a series of negotiations with institutions in the area, the Hispanic Health Care Coalition, is discussed. The HHCC's goals and current project to provide translation services for patients in area clinics and health care facilities are explained. Perhaps one of the significant features of this study, in addition to the fact that it illustrates a community-level language planning effort, is that it also exemplifies an effort begun by individual professionals and not by a state or federal institution. A study of the success of a language program in bilingual education, McConnell's paper discusses the findings of her seven-year study of two bilingual communities: one in Washington State that encourages language shift and one on the Texas-Mexican border that encourages language maintenance. One of the few longitudinal studies available that compares parallel bilingual programs in two diverse communities, this paper shows the effect of language maintenance/shift on academic progress in both English reading and mathematics, and discusses the effect of bilingual education in supporting both primary and second language development in both communities. One significant finding that school educators might not be aware of is the detrimental effect of language shift on the learning of math, though both groups made large gains in the program overall. Clearly the strongest aspects of the McConnell paper are its rigorous statistical documentation and its ability to analyze the bilingual education program within the context of the two very different language communities. The results of McConnell's study — which support bilingual education overwhelmingly - are important in Spanish language description and its nexus to language planning and language policy, since too frequently the complexity of the educational system and process tend to diminish the impact of such successful programs. In McDonnell's

Introduction

9

study bilingual education is presented as not only effective educational policy, but also as compatible within the existing language use patterns of Spanish-speaking persons in the United States. In the process of discovering and setting standards for Spanish language radio broadcasting in the U.S., Lipski moves from the national to the international use of Spanish in radio broadcasting in order to return to his focus on U.S. radio in Spanish. He deals with issues such as the training and ethnic background of U.S. Spanish-speaking radio announcers, early U.S. Spanish language programming, a general description of Hispanic broadcast style, and the hierarchy of radio broadcasting styles in Spanish. Focusing on two specific phonological features, Lipski proceeds to analyze four styles: formal (news), less formal (music), popular (sports), and actual speech. He suggests that this same hierarchy may be developing in the U.S. Spanish language broadcasting, and points to the need for further research into the various Latin American countries (13) already sampled in his study, in the two-step process of describing Spanish language use in broadcasting, and of planning for this use of Spanish. All in all, the papers in this section provide a variety of perspectives from which to view Spanish language planning in the United States. At the same time, all authors would concur that language planning has barely begun, and that persons informed about language and perceiving linguistic inadequacies have a great deal to contribute in applying language studies, assisting in language planning, and working with policy makers to ensure opportunity for Spanish-speaking persons in the health, legal, employment, citizenship, communication, social welfare, and educational domains. The complexities which underlie Spanish language planning and policy in the United States are obvious. Not only must we deal with different varieties of Spanish - both standard and nonstandard - but also with the emergence of different linguistic norms, with changing patterns of bilingualism, and with community-based goals for education which differ from those advocated by policy-makers and bureaucrats. As the issues are identified, and solutions proposed, all those interested in the development of the linguistic resources of Hispanic communities and a first class delivery of social services must become actively involved in the process. Those language researchers who are interested in greater social justice for minority language groups in the United States have a special role to play in reconceptualizing the nature of language and human rights in the United States (Macias, 1979), particularly in the establishment of a language data base such as the one proposed by Ferguson (1979) and in promoting changes in attitudes toward language planning. This information will provide language planners and policy-makers with the necessary research background on key issues such as census data, language choice in bilingual communities, the existing resources in terms of

10

Introduction

skills and usage, the identification of differences in linguistic norms, attitudes toward English and Spanish, and toward varieties of those languages, and knowledge of the relationship between language, culture and national identity. We hope that this collection will stimulate others to investigate the relevant issues addressed, so that we can obtain a better understanding of what United States Spanish is, and what the needs of its speakers are. Lucía Elias-Olivares Elizabeth A. Leone René Cisneros John Gutiérrez

References Ferguson, C. (1979). National Attitudes Toward Language Planning. In Language and Public Life. J.E. Alatis and G.R. Tucker (eds.). Washington, D.C.: Georgetown University Press. Jernudd, B. and J. Das Gupta. (1971). Towards a Theory of Language Planning. In Can Language Be Planned?, J. Rubin and B. Jernudd (eds.). Honolulu: East West Center Press. Macías, R. (1979). Language Choice and Human Rights in the United States. In Language and Public Life, J.E. Alatis and G.R. Tucker (eds.). Washington, D.C.: Georgetown University Press. National Puerto Rican Task Force on Educational Policy.(1977). Toward a Language Policy for Puerto Ricans in the United States: An Agenda for Community Involvement. New York: NPRTFEP. Poplack, S. (1982). Sometimes I'll Start a Sentence in Spanish y Termino en Español: Toward a Typology of Code-Switching. In Spanish in the United States, J. Amastae and L. Elias-Olivares (eds.). Cambridge: Cambridge University Press. Pousada, A. (1979). Interpreting for Language Minorities in the Courts. In Language and Public Life, J.E. Alatis and G.R. Thicker (eds.). Washington, D.C.: Georgetown University Press. Rubin, J. (1973). Introduction. In Language Manning: Current Issues and Research, J. Rubin and R. Shuy (eds.). Washington, D.C.: Georgetown University Press. Zentella, A.C. (1983). Language Planning: Acceptability and Adequacy Criteria. In Spanish in the United States: Beyond the Southwest, L. Elias Olivares (ed.). Rosslyn, VA.: National Clearinghouse for Bilingual Education.

Section 1 On Describing SpanishSpeaking Communities

Spanish in the Southwest: Language Maintenance or Shift? by Mary Beth Floyd

Introduction Spanish continues to be maintained as the mother tongue among significant proportions of the Hispanic population in the American Southwest. While this has long been assumed, it is only recently, that the specific topic of Spanish-language maintenance in the Southwest has begun to receive attention in the literature. Systematic investigation of patterns of language use among bilingual speakers in the Southwest and of factors which might relate to the maintenance or the loss of Spanish in not only this particular geographic region but in the U.S. in general has begun; the number of studies, however, remains limited. Moreover, while findings have been reported for some areas of the Southwest, primarily south Texas (Skrabanek, 1970; Thompson, 1974), New Mexico (Gutiérrez, 1980; Hudson-Edwards and Bills, 1980; Ortiz, 1975), southern California (López, 1978), and Colorado (Floyd, 1982), many other areas remain to be investigated. Fewer still are investigations which compare and contrast language-maintenance patterns across various Hispanic groups in the U.S.; a notable exception is a study (Laosa, 1975) which treats not only bilingual Chicano speakers in central Texas but also Cuban speakers in Miami and Puerto Rican bilingual speakers in New York. With respect to those studies which do constitute the literature to date, the findings are not always comparable, and although patterns of language use reported within the studies tend to be convergent, conclusions regarding language maintenance or shift within the same geographic area have at times been divergent. Given the state of the research at this point, then, the formulation of definitive conclusions regarding maintenance of the mother tongue, Spanish, or language shift in the direction of the dominant societal language, English, within the Southwest would not be justified. Nevertheless, the findings reported to date do reflect patterns of language use among bilingual speakers in the Southwest which merit consideration here. The intent of the present paper is to review the findings of those investigations which

14

Mary Beth Floyd

have treated Spanish-language maintenance in the Southwest, with particular attention to not only patterns of language maintenance or language shift which have been noted for the specific regions considered, but also with attention to various social and other non-linguistic factors, such as generation, age, and education, of the speakers involved which may relate to such patterns of language maintenance or shift among bilingual speakers in the United States.

Generation Differences in the use of Spanish among bilingual speakers according to generation have been noted in previous studies of Spanish-language maintenance in the Southwest. Language-maintenance studies for Texas (Laosa, 1975; Skrabanek, 1970; Thompson, 1974), for Los Angeles, California (Lopez, 1978), and for New Mexico (Hudson-Edwards and Bills, 1980; Ortiz, 1975) have reported patterns of language use among bilinguals as they relate to the generation of the speaker. The earliest study of those reviewed, that by Skrabanek (1970), examined Spanish language retention among 544 Mexican-American households in two areas of south Texas, i.e., rural Atascosa County and the urban San Antonio area. While Skrabanek's overall conclusion was that a high degree of retention of Spanish was evident among the Texas speakers, he did note that children within the households tended to use less Spanish than did their parents. Skrabanek noted, however, that it was only in the school situation that children used a high degree of English, and then only because it was required that they do so (1970: 276). Thompson (1974), in his study of 136 male heads of households in Austin, Texas, examined generational differences in language use. The patterns of usage revealed that first-generation speakers used all Spanish with children in the home. Second-generation fathers variably reported use of half Spanish or no Spanish, others reported use of all Spanish. Third-generation heads of households tended to report use of half Spanish or less with children at home (1974: 73-74). Sixty percent of the children of second-generation families were reported by their parents to use only English in the home even when their parents addressed them in Spanish. Thompson concluded that even in this predominantly Mexican-American area of Austin, English was becoming the language of the home (1974: 75). In a study of 295 children and their families, Laosa (1975) considered the use of 'language pattern' within specific social contexts among three Hispanic communities, namely Mexican Americans in Austin, Texas, Cubans in Miami, and Puerto Ricans in New York. Informants' reported use of

Spanish in the Southwest: Language Maintenance or Shift?

15

language patterns, i.e., either English, Spanish, English and Spanish, or a 'mixture' of the two languages as the single most frequent language, was considered in relation to the social contexts of family, for all informants, and school, for the children. The language pattern reported most frequently as used at home and at school by the Mexican American children in Austin tended t o be English; the most frequent pattern reported for their parents was a mixture of Spanish and English. Found to use English most frequently were not only the Mexican American children whose parents used English as their language pattern, but also those whose parents used a mixture of both English and Spanish. And even in those cases in which parents used Spanish most frequently in the home, Mexican-American children tended to use both English and Spanish (1975: 621-622). Differences between the Texas children's language use patterns and those of their parents were found to be significant (1975: 624). Laosa's study, a rarity in the literature in that it investigates language use patterns among several Spanish-speaking groups, indicates that significant differences prevailed among the various Hispanic sub-groups in this respect. For example, in contrast to the pattern of use of English among the MexicanAmerican children in Texas, the large majority of both parents and children in the Cuban and Puerto Rican groups reported using Spanish as the language pattern of the home. In contrast to the Texas children, the only Cuban and Puerto Rican children to use only English tended to be those whose parents also used English most frequently in the home. Statistical analysis showed differences among the three Hispanic sub-groups to be highly significant (1975: 6 2 0 - 6 2 1 ) . In the analysis of children's language use in relation to that of their parents, significant differences were shown for both the MexicanAmerican and the Cuban groups, the Texas group showing the greatest difference between language use of the two generations. Laosa's ultimate conclusion: the Hispanic group showing the greatest retention of mother tongue was the New York Puerto Rican group; in contrast, the group manifesting the greatest degree of language shift in the direction of English was the Mexican-American group in Texas (1975: 625). Lopez (1978), reporting on data deriving from a 1973 survey of language use in the home of 890 Chicano married women in Los Angeles, California, noted a striking reversal of language between first-generation and thirdgeneration speakers. While 84 percent of the first-generation informants continued to use primarily Spanish, 84 percent of the third generation reported using mostly English (1978: 271). Noting that intergenerational shift from Spanish to English was progressing rapidly, Lopez concluded that while the loss of Spanish among Chicanos in the future is not inevitable, " . . . in the absence of countervailing institutions it is exceedingly likely" (1978: 276).

16

Mary Beth Floyd

Ortiz (1975), examining patterns of language use among 48 children and their parents in the northern New Mexico community of Arroyo Seco, found that while parents and children tended to use Spanish predominantly with each other in the home, children used approximately as much English as they did Spanish at home with siblings (1975: 129). It was just this interaction between siblings which Ortiz observed as the vehicle by which English had been introduced into the family setting. In this traditionally Spanishspeaking community the majority of the children were judged as able to function well both in English and in Spanish. Increases in the number of people in the community claiming both Spanish and English as mother tongue as well as those claiming equal proficiency in both languages were taken by Ortiz as indicators of intergenerational language shift from an essentially monolingual Spanish community in 1940 to a bilingual community in 1975. Hudson-Edwards and Bills (1980), in al975 survey of 55 Chicano households in Martineztown, a barrio of Albuquerque, New Mexico, considered mother-tongue claiming, self-reported language proficiency and language use in relation to generational data. While 87 percent of second-generation heads of households claimed Spanish as their mother tongue, only 44 percent of third generation claimed Spanish (1980: 143—44). The greatest Spanish proficiency was reported by informants who were members of single-generation households; next were informants of multiple-generation homes without minor children; those reporting the lowest frequency of Spanish fluency were members of multiple-generation homes with younger children. With respect to the generation of the informant, all of the parents of heads of households reported fluency in Spanish; 85 percent of the heads of households did so; only 33 percent of the younger generation reported fluency in Spanish (1980: 148). Use of Spanish in the home was seen to be greatly influenced by the presence or absence of minor children in the home: the greatest use of Spanish was reported by those living in homes without young children (1980: 150). The authors concluded that intergenerational language shift was very much in evidence in Martineztown, New Mexico.

Age The factor of speakers' age has also been considered by several investigators (Floyd, 1982; Gutierrez, 1980; Hudson-Edwards and Bills, 1980; Ortiz, 1975; Skrabanek, 1970; Thompson, 1974) in relation to the maintenance of Spanish by Chicanos in the Southwest. Skrabanek (1970), in his study of south Texas, noted a "direct relationship between age and use of Spanish" among heads of households, with speakers aged 55 and older making the most use of Spanish and those under

Spanish in the Southwest:

Language Maintenance or Shift?

17

35 using Spanish the least (1970: 277). Similarly, younger children tended to use Spanish less than did their older siblings, and older children used less Spanish, in turn, than did their parents. Despite these patterns, it should be recalled, as noted earlier, that Skrabanek's overall conclusion was that a high degree of Spanish was being maintained even among the younger MexicanAmerican speakers (1970: 276). Thompson (1974), in his Austin study, found that the use of all Spanish with children in the home predominanted among the oldest heads of households, i.e., those 60 years of age and older, while the youngest group of fathers, those aged 18 to 29, reported use of half Spanish or less with their children (1974: 74). In his investigation of language maintenance in Arroyo Seco, New Mexico, Ortiz (1975) found in comparing across age groups that those who reported favoring Spanish as the language they spoke 'most comfortably' were the oldest adults and the youngest children. The overall pattern was one of Spanish monolingualism among the two age extremes, i.e., the pre-school children and the adults over 60. On the other hand, school-age children and younger adults who had attended school were characterized by a high degree of bilingualism (1975: 153-154). Although Ortiz observed 'considerable evidence' of a diglossic relationship between Spanish and English within the community, Spanish being the language used within the group and English being that used for intercultural communication, the relationship was, nevertheless, Ortiz noted, a 'relatively tenuous' one. Given this fluid rather than stable language situation, Ortiz concluded with respect to the future of Spanish in Arroyo Seco that it "remains to be seen whether the language situation will stabilize further with the passage of time" (1975: 198). With regard to the factor of age, Hudson-Edwards and Bills (1980) found a strong 'negative correlation' with the claiming of Spanish as the mother tongue. While nearly all informants over 25 claimed Spanish as their mother tongue, only 38 percent of those under 25 did so (1975: 142). Age was also found to be highly correlated with self-reported proficiency in Spanish: while 87 percent of those 26 years and over reported relatively high ability in Spanish, 86 percent of the younger members claimed fluency in English (1975: 147, 149). The primary purpose of a study by Gutie'rrez (1980) was to determine the functions of Spanish and English within 50 bilingual families in Martineztown, a predominantly Chicano area of Albuquerque, New Mexico. Language use within the domains of home, work, school, and recreation was considered in relation to the factors of informants' age, sex, religion, and education. Within the domain of school, English was overwhelmingly reported as the predominant language ; within that of work, the preferred language was also reported to be English. With respect to the language pattern in the home, the New Mexico speakers reported a slight preference for Spanish.

18

Mary Beth Floyd

However, when Gutierrez examined language used at home in relation to the age of the informants, the results were striking. While informants over 30 reported a preference for use of Spanish, the greatest use being reported by those informants aged 51 and over, a preference for English as the language of the home was reported by the informants under 30, the highest preference being shown by the 6 to 16-year-old group. The same pattern of language use was reported for the Martineztown speakers within the domain of recreation. English was the dominant language for the under-30 group, and for those informants 16 and under English was the only language used; Spanish was maintained only among the informants over the age of 30 (1980: 456—457). The greatest factor to be correlated with informants' language use was found to be that of age. And while Gutierrez noted that Spanish was still functioning as the language in the home domain, the 'maintenance stronghold' of Spanish, the author ultimately concluded that English "appears to be functioning as the language of all domains for the members of the speech community under thirty, implying that an overt language shift is in process in Martineztown" (1980: 459). In a study of 61 university-age informants in Colorado (Floyd, 1982), several indices of Spanish-language maintenance, i.e., scores on a standardized listening-comprehension test, self-reported use of Spanish at home, and self-reported skills in Spanish, were examined in relation to several factors including age of informants. The age of the Colorado speakers was found to be one of the most significant factors relating to the informants' maintenance of Spanish, with older speakers reporting more use of Spanish at home than younger informants. Age was also significant in relation to informants' listening comprehension scores: the older informants showed the greatest degree of comprehension. With regard to their self-reported language skills, again age was a significant factor: older informants reported more skills in Spanish than did younger informants (1982: 300).

Other Factors While most of the studies reviewed have considered language maintenance in relation to the age and/or generation of the informants, some studies have mentioned other factors which might relate to patterns of language use among bilingual speakers, variables such as childhood residence, language use at home, formal study of Spanish, education, and sex of the speakers involved. Differences in language use related to childhood residence of the speakers involved have been shown in some studies of language maintenance. Skrabanek's (1970) data reflect a pattern of relatively greater maintenance of Spanish among adults and their children from rural Atascosa county house-

Spanish in the Southwest: Language Maintenance or Shift?

19

holds than among speakers from the urban area of San Antonio, Texas. Thompson (1974) examined generational differences among heads of households in Austin in language retention in relation to the factor of childhood residence of the informants. Thompson found that first-generation speakers, i.e., those who used all Spanish with children in the home, had had a rural childhood residence; second-generation fathers, those reporting use of half Spanish or no Spanish, on the one hand, or all Spanish on the other, were divided between rural and urban childhood residences; third-generation heads of households, those reporting use of half Spanish or less with children in the home, reported largely urban childhood residence (1974: 74). It might be noted that it was Thompson's analysis of precisely this variable of childhood residence, or urban versus rural background, in relation to other data which revealed what Thompson considered to be the pattern of language shift in process in Austin. Lopez (1978: 273) indicated that although the effect of rural origin was not as great as that of generation, his findings nevertheless did give 'limited support' to the idea of greater maintenance of Spanish in rural areas than in the urban Los Angeles area. One factor which merits mention in a consideration of language maintenance among bilinguals is that of the language used in the home. Usually the use or non-use of Spanish in the home has been treated implicitly as an index of language maintenance. Rarely, however, has language use at home been isolated and considered in relation to other indices of language maintenance or in relation to non-linguistic factors which may be related to patterns of language use by informants. In his study of language maintenance in New Mexico, Ortiz (1975), however, considered the importance of this variable and noted emphatically that the greatest single factor in fostering the maintenance of Spanish in the Arroyo Seco community was its use in the home and within the family. Older parents of northern New Mexico children reported a significantly greater degree of use of Spanish in the home than did younger parents. An increase in the use of English within the home between siblings was attributed to the greater use of English among younger parents and the influence of older siblings who had learned English at school (1975: 161). Indices of language maintenance among Colorado speakers (Floyd, 1982) were analyzed in relation to parents' use of Spanish with each other as well as parents' use of Spanish with children in the family. With respect to the language used at home by parents, 74 percent of the Colorado informants indicated that their mothers used only English with children in the family; 75 percent of them reported that their fathers used only English at home with children. This pattern of preference for English with children in the home prevailed despite the fact that the informants reported that 62 percent of their parents used some or all Spanish with each other at home (1982:

20

Mary Beth Floyd

298). The Colorado informants' self-reported use of Spanish at home was seen to relate significantly both to the parents' use of Spanish with children at home and to parents' use of Spanish with each other. Not surprisingly, greater use of Spanish by informants was associated with greater use of Spanish by parents at home. Informants' listening comprehension scores were also seen to relate significantly to parents' use of Spanish with children at home. Similarly the Colorado informants' language skills were seen to relate significantly to parents' use of Spanish with their children and with each other: more skills were associated with greater parental use of Spanish (1982: 300). Another factor, that of years of previous formal study of Spanish, was considered in relation to measures of language maintenance among young Colorado speakers (Floyd, 1982). The only index of Spanish-language maintenance to show a relationship here was that of self-report of language skills. While reported skills in listening and speaking were not significantly related to years of study, informants' report of ability to read and/or write Spanish was significantly related to years of formal study of the language. The factor of informants' general level of education has rarely been considered among the studies reviewed in relation to language maintenance. Gutie'rrez (1980), however, examined this variable in relation to New Mexico informants' language use and found a slightly higher percentage of informants having no education reported using Spanish in the home than did those having completed twelfth grade. Similarly, within the domain of recreation, a higher percentage of informants with no education reported using Spanish than did those with a high-school education (1980: 457—458). One of the few studies to consider language maintenance in relation to the sex of informants was that by Gutierrez (1980). The investigator, however, found no differences between males and females with regard to their maintenance of Spanish in the domain of home or recreation.

Discussion Although the systematic study of Spanish-language maintenance in the Southwest among Hispanic bilinguals is relatively recent, with the majority of the studies spanning only the past decade, the literature reflects patterns of language use which are quite evident. Investigators of bilingual speakers of Austin, Texas (Laosa, 1975; Thompson, 1974), Los Angeles (Lopez, 1975) and New Mexico (Gutierrez, 1980; Hudson-Edwards and Bills, 1980) conclude that findings offer strong evidence of intergenerational language shift from Spanish to English, or from a monolingual Spanish community to a bilingual one (Ortiz, 1975). Of the studies considering language maintenance among

Spanish in the Southwest: Language Maintenance or Shift?

21

various generations, the only one which does not ultimately report such a conclusion (Skrabanek, 1970) is the oldest study to be considered here and was based on data which predated 1966. It has also seen suggested (Lopez, 1978) that in that case conclusions were drawn regarding the future viability of Spanish within the Texas community on the basis of adult usage at the time rather than on considerations of adult to children usage and usage between children themselves. Language maintenance in the Southwest has been studied in relation to several factors having to do with speaker circumstance. The most frequently considered factors have been generation and age of speaker, with less attention paid to such factors as speakers' language use at home, childhood residence, education and sex. The relationship between language maintenance and the factor of generation has consistently been reported in the literature. Generally, greater use of Spanish is reported for first-generation speakers, with successively less use of Spanish manifested by successive generations of speakers. It has been noted that even in the case of second-generation families who used Spanish with their children, the majority of the children used English with their parents (Thompson, 1974). Differences between Texas parents' and their children's patterns of language use were found to be significant; in fact, these bilingual Chicano speakers in the Southwest showed the greatest differences between language use of the two generations of all three sub groups of Hispanic speakers (Laosa, 1975). Among Los Angeles women a striking reversal from use of Spanish to English from first to third generations was observed (L6pez, 1978). Among speakers in New Mexico greater maintenance of Spanish was found among older generations of informants and successively less among younger generations of speakers whether the index of language maintenance was mother-tongue claiming, self-reported language proficiency, or report of language use at home (Hudson-Edwards and Bills, 1980). The factor of age has been considered very often in the language-maintenance literature. Almost without exception, greater use of Spanish as the mother tongue has been seen to be associated with age: the older the informants, the greater the degree of Spanish language use (Skrabanek, 1970; Thompson, 1974). An interesting pattern was noted (Ortiz, 1975) in which the speakers at the two age extremes, i.e., the oldest adults and the youngest pre-school children, favored Spanish and were monolingual while the schoolage children and younger adults were bilingual and favored less use of Spanish. Whether the index of language maintenance has been mother-tongue claiming (Hudson-Edwards and Bills, 1980), self-reported proficiency in Spanish (Floyd, 1982; Hudson-Edwards and Bills, 1980), listening comprehension (Floyd, 1982), or self-report of language use in the home (Floyd, 1982; Gutierrez, 1980; Ortiz, 1975; Thompson, 1974), greater maintenance of

22

Mary Beth Floyd

Spanish has been seen to be associated with older informants; the greater use of English has been associated with younger informants. The above findings which suggest differences between patterns of language use according to age and generation of the speakers involved indicate that among the speakers in the areas studied, language shift away from the language traditionally considered to be the mother tongue, Spanish, and in the direction of the dominant societal language, English, is in process. While the shift may be progressing at a faster pace in the large urban areas such as Austin and Los Angeles, the literature suggests that even in more remote areas like Arroyo Seco, New Mexico, language shift is in evidence. The demonstrated relationship between language maintenance and generation and/or age of the speaker is the expected one: the manifestation of greater mother-tongue retention on the part of the first generation and the older speakers of the language alongside the manifestation of greater use of the dominant language on the part of the younger and successive generations of speakers. Interestingly enough, from an historical perspective, the same language behavior which can be taken as a consequence of the fact that language shift has been in process in the past can be considered as a contributing causal factor in continued language shift in the direction of the dominant language in the future. Other factors such as informants' use of language at home, formal study of Spanish, general level of education, and childhood residence have been given less attention in the literature. The factor of language use at home has been treated variably in the literature. In most of the studies reviewed, language used at home was considered implicitly as an index of informants' language maintenance or language loyalty. In a few studies, however, the factor of language use within the family at home has been given explicit consideration in relation to indices of language maintenance. For speakers in New Mexico the factor of language use at home was considered to be the greatest single factor in the maintenance of Spanish (Ortiz, 1975). Among Colorado informants, significant relationships were found between parents' use of Spanish with each other and informants' use of Spanish at home, informants' listening comprehension, and informants' reported language skills. A pattern emerged in which the majority of the informants' parents, although they were reported to use Spanish with each other, were reported to have used English in interaction with their children in the home (Floyd, 1982). The importance of the factor of language use in the home is deserving of further investigation. Findings in these studies which have considered language use by the family in the home in relation to informants' language maintenance suggest that the factor of language use, and more specifically the language used by parents in interaction with children, is strongly related to

Spanish in the Southwest: Language Maintenance or Shift?

23

the degree of language maintenance on the part of these same children. The language used by parents with each other was seen to be significantly related to all indices of language maintenance for Colorado university-age informants (Floyd, 1982). It has been noted (Lopez, 1978) that the language behavior between adults and children and that between children themselves in the family should be good predictors of the language to be used by future generations of speakers. However, in some investigations of language maintenance children's use of language is reported to vary considerably from that of their parents, with a preference for English reported even when parents used all or some Spanish with their children at home (Laosa, 1975; Thompson, 1974). This factor of language use in the home in relation to language maintenance among bilinguals is deserving of more thorough investigation before definitive conclusions can be formulated. The factor of formal study of Spanish was considered in relation to indices of Spanish-language maintenance among Colorado speakers (Floyd, 1982). While no relationship was seen between formal study and self-reported listening and speaking skills, a significant relationship was seen between formal study of Spanish and self-reported literate skills of reading and writing. These findings suggest that while the aural-oral skills of listening and speaking may be acquired and maintained independent of formal study of the mother tongue, the literate skills of reading and writing are more likely to be acquired through the intervention of formalized study of the language. Findings in the literature suggest greater maintenance of Spanish is associated with speakers who report a rural childhood residence as opposed to an urban one. Greater maintenance of Spanish was observed among rural speakers in Texas than among speakers in the urban San Antonio area (Skrabanek, 1970). The relationship between Austin speakers' use of Spanish and their childhood residence was discovered by Thompson (1974). It might be noted that it was precisely this variable which Thompson suggested should be considered in any investigation of the effects of urbanization on language loyalty. A study of Los Angeles speakers (Lopez, 1978) lent limited support to the idea that in the past speakers in rural areas had manifested greater maintenance of Spanish than had speakers in the urban Los Angeles area. The factor of speakers' level of education has been considered in relation to language maintenance among New Mexico speakers. A somewhat greater use of Spanish was reported by informants with less education than among those with a high-school education (Gutierrez, 1980). These findings which have been reported in the literature suggest a possible relationship between language maintenance and factors such as education level or childhood residence of the speaker. If in fact these factors are related to language maintenance, such factors might be seen as indirect measures or indices of degree of assimilation to the dominant culture on the part of the

24

Mary Beth Floyd

speakers in question. Greater use of English would be expected to be associated with a greater degree of assimilation on the part of the bilingual speaker to the dominant monolingual culture and language. Nevertheless, again, further investigation of these factors and similar socioeconomic factors in relation to Spanish-language maintenance in the Southwest would clarify this relationship.

Concluding Remarks The present paper has been a review of the literature dealing with Spanishlanguage maintenance or language shift in the direction of English among Hispanic bilinguals in the American Southwest. Although an investigation of patterns of language maintenance among other Hispanic groups outside the Southwest, especially Cuban and Puerto Rican bilingual speakers, was beyond the scope here, such a review would be welcome. Comparisons between and among various Hispanic groups of speakers within the United States with respect to language maintenance would then be facilitated. A review of the current literature reflects certain limitations. The studies are few, the majority of them having been undertaken only in the past decade. Methodological differences between studies, e.g., the use of different measures such as self-reported language proficiency, language use, language loyalty, etc., to reflect language maintenance, make difficult cross comparisons between the findings of the various studies. Given these limitations, the formulation of conclusions regarding the maintenance or loss of Spanish among bilingual speakers in the Southwest must be couched in somewhat tentative terms at this point. Despite such limitations, however, certain conclusions seem justified. The existing investigations of Spanish-language maintenance among Chicano speakers in the Southwest offer evidence of a process of language shift from Spanish to the dominant societal language, English. The existing literature shows a relationship between such a shift and the factors of the speakers' age and generation, with younger informants and successive generations of speakers within the same families showing decreasing maintenance of Spanish. The language used at home is perhaps the most important factor related to Spanish language maintenance in the Southwest; Other factors related to speaker circumstances, such as childhood residence and degree of formal education, may be found to relate to patterns of language maintenance as indirect measures of degree of assimilation to the dominant monolingual culture. These factors and others not yet considered deserve thorough investigation before definitive conclusions can be drawn regarding their significance with respect to language maintenance among Chicanos in the Southwest.

Spanish in the Southwest: Language Maintenance or Shift?

25

The existing literature reflects that among Hispanic bilinguals in the Southwest language shift is in process away from Spanish and in the direction of English. These findings might be seen to have consequences for language planners and the notion of language planning in the United States. If the maintenance of ethnic language proficiency is seen as a value in a pluralistic society, such planning may be necessary to moderate attrition or loss of language skills among bilinguals and to foster first or ethnic language development. While the field of language planning is a relatively new one, particularly as it might be seen to relate to Spanish in the United States, the preeminence of Spanish as a minority language in our country may well argue for more attention being focused on such planning.

References Floyd, Mary Beth. (1982). 'Spanish Language Maintenance in Colorado.' In Florence Barkin, Elizabeth A. Brandt and Jacob Ornstein-Galicia (eds.), Bilingualism and Language Contact: Spanish, English, and Native American Languages, 290—303. New York: Teachers College Press. Gutierrez, John R. (1980). 'Language Use in Martineztown.' In Florence Barkin and Elizabeth Brandt (eds.), Speaking, Singing and Teaching: A Multidisciplinar)) Approach to Language Variation. Proceedings of the Eighth Annual Southwestern Areal Language and Linguistics Workshop. Anthropological Research Papers, No. 20, 454-459. Tempe: Arizona State University. Hudson-Edwards, Alan and Bills, Garland D. (1980). 'Intergenerational Language Shift in an Albuquerque Barrio.' In Edward L. Blansitt, Jr. and Richard V. Teschner (eds.), A Festschrift for Jacob Omstein, 139158. Rowley, Mass.: Newbury House. Laosa, Luis M. (1975). 'Bilingualism in Three United States Hispanic Groups: Contextual Use of Language by Children and Adults in Their Families.' Journal of Educational Psychology 67,617—627. López, David E. (1978). 'dùcano Language Loyalty in an Urban Setting.' Sociology and Social Research 62,267-278. Ortiz, Leroy I. (1975). A Sociolinguistic Study of Language Maintenance in the Northern New Mexico Community of Arroyo Seco. (Doctoral dissertation, University of New Mexico, 1975). Dissertation Abstracts International, 1976,37, 2159A. Skrabanek, R.I. (1970). 'Language Maintenance among Mexican-Americans.' International Journal of Comparative Sociology 11,272—282. Thompson, Roger M. (1974). "The 1970 Census and Mexican-American Language Loyalty: A Case Study.' In Garland Bills (ed.), Southwest Ared Linguistics, 65-78. San Diego: Institute for Cultural Pluralism.

Hispanic Attitudes in Northwest Indiana and New York by

John J. Attinasi

Spanish is linguistically the equal of the English language, with millions of literate speakers, a distinguished pantheon of masters and masterpieces, and active communities enlivening speech in variation. Sociolinguistically in the United States, however, English is the more powerful language, and the prestige and value of Spanish is not widely recognized (Sole, 1982: 255—6). The scholarly view of linguistic equality thus seems irreconcilable with the political and popular view of unequal linguistic advantage and power. Does Spanish imply an evolutionary, linguistic or sociohistorical 'disadvantage' justifying the supposed attempt to erase linguistic background and homogenize culture as an assimflationist educational system would do? Are there resistance and resources enough for ethnolinguistic viability? To answer these questions, the linguistic views of those affected are crucial. Attitudes in three key areas, language, identity and education, reflect concretely the sociolinguistic interaction of Hispanics with the United States, and are therefore the foci of this study. The present sociolinguistic situation demands that correspondences between language and culture be investigated, not in relation to perception or woridview, but in terms of the social and material resources of a culture and continuity across its generations. Community attitudes on language and culture influence cultural outcomes, but not in the strong, deterministic version of the Whorfian hypothesis that language influences woridview. Rather, attitudes, like images in speech, intensify perceptions and societal orientations. Attitudes are a special kind of language that "reticulated with culture, interact with the imagination", as Friedrich says, writing of poetry (1979: 492). Just as language most affects the individual in imagination and poetic use, language most affects societal images and orientations in expressed attitudes. But Hispanic attitudes are not simple, or homogeneous just because of Spanish language background. The thought of an ethnolinguistic group

28

John J. Attinasi

remains variable, and the culture mutable, just as languages continue despite individual differences and orientations of speakers. There may be, for example, two persons who consider themselves Mexican and are of the same generation, but who differ diametrically in language use and views of society, due to personal history and orientation. Because they are the views of the largest language minority in the U.S., the linguistic practices and attitudes of the population of Spanish background are crucial for linguistic projections of multinationalism in coming decades. The situation of Spanish may only be a partial indicator of the situation of other languages, due to differences in the size and social characteristics of the ethnolinguistic community (Giles, Bourhis and Taylor, 1977). Understanding the complete situation of Spanish requires as well the comparison of various speech communities, and analysis along dimensions of demography, history and economy. A goal of continued attitude study has been to determine whether Hispanic attitudes in New York are different from those in cities of smaller size, and in regions where Mexican and Puerto Rican communities live together, such as the Midwest. Even within each nationality there are many layers. Puerto Ricans enter the U.S. as citizens speaking Spanish, with a history of Afro-Indio-Hispanic admixture and colonialism under Spanish and North American rule. There are now U.S.-born-or-based Puerto Ricans, as well as those in various patterns of migration. Mexicans also have a history of two colonialisms, as both native-born and migrant persons in the U.S. But a large part of the territory of Mexico has been annexed by the U.S., so that the 'homeland' of many Mexican-Americans is now the 'sunbelt' of the U.S. While Cubans have had time t o create a U.S.-bred population, a new cohort of Cubans has arrived recently. Many professional and middle-class Cubans were welcomed as an 'embarrassment to Marxism' a generation ago. But Cuban refugees of the 1980s receive a much cooler welcome for reasons too complex to detail here. Although it may be true that the majority response to all ethnolinguistic groups has usually been an insensitive demand that they 'speak English and assimilate', attitudes toward the national minority language and culture have not been uniform. The understanding of attitude diversity needs to be informed not only by the historical and social layers in each ethnic nationality, but also by the assumptions and goals of attitudinal research itself. Movements which have been unmentioned in traditional linguistic treatments, civil rights, bilingual education and ethnic 'consciousness-raising' constitute the current dynamics of language related cultural change at present. Only when social movements and variation across and within groups inform conclusions regarding attitudes can meaningful predictions of ethnolinguistic maintenance be made. Static views, neglect of social forces or superficial

Hispanic Attitudes

in Northwest

Indiana and New York

29

treatment of issues unfortunately abound in much of the attitudinal research from psychological and educational disciplines. The critique of those inadequacies exists, regarding both theory and method (see Lee 1971; Williams 1970; Attinasi MS); but an alternative attitudinal literature is still to be written. Theoretically, the main disappointment lies in the one-sided orientation of studies. Attitudes toward English, or of Anglos toward ethnically-accented speakers, or toward minority languages in the abstract as home-based (nonstandard) systems are sought. The bias that these studies bear against bilingualism as a working solution to contemporary ethnic contact and international migration is conspicuous, and the living evidence of that bilingual solution in working-class communities is overlooked. Methodologically, there is a lamentable divorce in scholarly attitudinal studies from the rich and vital substance of people's attitudes. So often statistical refinement and 'doable studies' dissolve ultimately complex issues into simplistic positive and negative scales. The impersonal methodologies of mail-in and fill-in questionnaires further hamper the emergence of the quality of attitudes. Ethnographic familiarity and personal interviews partially overcome this sterility. Many films and works of literature crystallize attitudes in the voice and conflict of characters. Even though anecdotal and literary testimony are often the most vivid, they lack the broad-based validity that the survey technique affords. And so a trade-off between compelling personalism and generalizable statistics is needed, as is the balance of statistical significance and perspicacious relevance. Lastly, the use of school-age samples, or parents who are concerned with school success, often misrepresents the cultural orientation of the entire linguistic group (Ryan and Carranza 1980). This difficulty becomes serious when the out-of-school population holds values that diverge greatly from younger age-sets. Interference in interpreting community-wide patterns also arises from the natural desires of parents to conjecture what might be the object of the inquiry and answer accordingly. Where schooling is concerned, the 'right answers' are usually based in assimilation and standard language. Several research studies have confirmed the majority view toward speech variation, unsurprisingly finding that students with 'good voices', unaccented and standard speech are evaluated more favorably especially by teachers and non-minority-group members, but also by minority students (Seligman, et al. 1972; Williams 1976; Ryan and Carranza 1975). These judgments often involve the students as the object of a self-fulfilling prophecy: low evaluation will yield low achievement. Hence the high drop-out and push-out rates for Hispanics may be linked to language by complicity, not by causality. Ramirez (1981: 223) suggests that students who share their teachers' atti-

30

John J. Attinasi

tudes may ultimately achieve higher grades and Mora (1975) found that Mexican immigrants held more negative attitudes toward hispanicized English words (wachar, lonche, etc.) than Mexican-Americans, especially those who used such strategies. All of these studies confirm the usual hierarchy that links (academic) success with standard English, the avoidance of code-switching and unaccented speech, usually condeming Spanish as a disadvantage. Were it not for research on minority group attitudes and non-school populations, it might seem that the dismal self-fulfilling prophecy is the only prognostic for Spanish in the US. Still, most of the conclusions drawn from the research on school teachers and student populations recommend teacher awareness of the communicative competence of their students and change in attitudes to end the exclusion of language-minority (and dialect-minority) students from opportunities for academic success. But the subtext in these recommendations is that "Spanish is of little pragmatic value, since it limits and in many cases even hinders, social mobility in an alien world;" and that language shift for Hispanics, as for other minorities and immigrants, will be the ultimate sociolinguistic outcome for future generations (see Sole 1982: 255,267). Studies of attitudes held by ethnolinguistic groups themselves yield other interpretations more favorable to the future of Spanish in the US. But since such studies are fewer than those of accented and standard English, and since the minority populations are socially and linguistically marginated, their attitudes are considered less weighty, and the challenges they present to the educational system are often brushed aside. The overwhelming pro-maintenance sentiment among Hispanics for Spanish, whether from a cultural standpoint or as an economic or integrative advantage seems frequently reduced to insignificance by disclaiming qualifications (e.g., it is only the elite, the young radical, the barrio-ghetto residents who hold attitudes toward maintenance; or, even though a desire is voiced, few have actively sought to improve their Spanish; see Sawyer 1964; Sole 1977.) The conflict of attitudes implicit here is more at the societal level than at the theoretical, for whereas scholars respect differences in findings and use them to refine methods, school administrations and the general public usually do not. They would rather have one conclusion to operate upon, and the one which makes the least demand on the status quo is usually the one they would most readily adopt. Hence, the conflict model called for by Penalosa (1980, 1981) opposes the view of attitudinal consensus. This model has value in that language attitudes themselves are seen as reflections of societal stresses and political competition, and frequently involve economic resources. Comparison of samples and of characteristics within samples allows the definition of some of the lives of conflict in attitudinal differences. These

Hispanic Attitudes

in Northwest

Indiana and New York

31

may entail the desirability of Spanish maintenance, the salience of accentedness, the condemnation of code-switching, or bilingual education. In previous studies, among Cubans in Miami and Puerto Ricans in New York, ethnic density has been emphasized as a key factor in explaining continued use of Spanish and positive attitudes regarding both the language and culture of the Hispanic group.1 The question here is whether the values of Hispanics will 'melt' into the larger 'pot' of general North American Anglo attitudes when members are dispersed among other populations in suburbs and smaller cities. This question has both a positive and negative answer in the Midwest as the present study in Northwest Indiana will demonstrate. The Midwest is unique in the Hispanic situation in the United States because of the nationality mixture of Hispanics, a mixture that is not without its conflicts and challenges to unity. In the north central states, Hispanics of Mexican ancestry predominate with persons derived from either the annexed Southwest, with about a century of Anglo influence, or from Mexico by recent migration. It is anticipated that census figures will confirm estimates that two-thirds of the Hispanics in the greater Chicago area (including Northwest Indiana) are of Mexican background. About one-quarter are Puerto Rican in the Midwest, with greater political leverage due to unequivocal U.S. citizenship seeming to equalize the sheer numerical weight of Mexicans. Cubans, mostly from the first, more middle-class and anti-Castro migration, as well as some Central Americans and fewer South Americans and Spaniards make up the remainder. Table 1 Hispanics in Lake County

Lake County Calumet Township (incL Gary) Center Twp. Hobart Twp. North Twp (incL East Chicago) East Chicago city Tracts 308-309 (East Chicago)

Number

% of Total

43,932 11,789 342 2,670 26,239 16,818 6,473

11.9% 6.7 % 1.4% 6.2% 14.1 % 42.3 % 57.8%

Source: 1980 Census of Population and Housing, Advance Reports, Indiana, PHC80-V16, Table 1, pp. 13-14. For Tracts 308-309: Doc. P.L. 9 4 - 1 7 1 Counts, Lake County, IN 2/28/81 pp. 397-399.

Ethnic density, such as that found in New York Puerto Rican neighborhoods like El Barrio or the 'little Cubas' of Tampa or Miami does exist in

32

John J. Attinasi

some neighborhoods of Chicago and East Chicago (Indiana), but Hispanics are more dispersed in Gary, and in the rest of Northwest Indiana. (In all there are eighty-seven thousand Hispanics in Indiana, with over half that number (43,932) in Lake County. 2 See Table 1. Such patterns complement the single-nationality focus and high-density populations which have been studied in the Southwest, Florida and New York. It becomes instructive to determine if, under conditions of greater admixture and dispersal — conditions of the future for Hispanics in the United States — adherance to Spanish remains strong. The support for Spanish maintenance among U.S. Hispanics, even among respondents who do not speak it, has been established by many surveys and studies. Nearly one hundred percent of three samples, two in New York and one in Northwest Indiana, agree on the desirability of maintenance; some in New York were indignant that the issue of total language shift be even asked! Questions of actual usage and means to formulate probilingual policy were other issues, but need contextualization by sample.

The New York Block Over three years of intensive data collection, bracketed by periods of equal duration to pilot-test and analyze data, lend credibility and life-like resonance to statements by the Language Policy Task Force of Centro de Estudios Puertorriqueños, CUNY, regarding ethnographic attitudinal and sociolinguistic aspects of speaking (Language Policy Task Force, forthcoming). These statements are further refined through extensive discussion with Puerto Ricans, both research subjects and other researchers.3 The main sample for the research in the community, sponsored in part by the Ford Foundation, was a selected group of 91 persons from a single block in El Barrio, or East Harlem (a purposive sample). The group consisted of men, women and children over twelve, chosen to reflect a wide variety of experiences in, and migration to both New York and Puerto Rico. The New York block near Lexington and Third Avenues where members of this sample resided was nearly ninety percent Puerto Rican, a situation rare in Brooklyn, the Bronx and even in other parts of El Barrio. The ethnic density and sheer size of El Barrio is further reinforced by the now-prevalent Hispanic ambience in New York City, where between one-eighth and one-quarter of the population is Hispanic. The Puerto Rican situation, social and linguistic, is significantly strong yet filled with contradictions. A confidence of Puerto Rican permanence, without loss of cultural pride and language is evident in many homes and in public places, but so are the staggering statistics on employment, housing, educational completion and median income. Regarding language, however, wise

Hispanic Attitudes

in Northwest

Indiana and New York

33

and seasoned conversationalists among the block sample, persons who lived in Puerto Rico through imposed English-language instruction and 'bootstrap' economic development, are living testimony to the contradictions in US policy and in the myths of cultural assimilation. Their discussions of events and personal experiences are rich in context and formed in two languages, mixed in such a way that new insights on code-switching as a linguistic phenomenon are emerging from the research.4 Many of these persons are quite pro-American but do not speak English, nor do they intend to. So too, many habitual English speakers are staunchly pro-Puerto Rican. One of the most interesting insights to emerge from the three-pronged research was the cultural warp in apparent time as a means to predict linguistic change. If, judging by the greater use of English among adolescents, an observer would conclude that in one generation most adults and all children would be speaking English, then the error would lie in not considering the reintegrative effects of childbearing and migration that return the speaker to increased Spanish usage. The young adults in the New York Block exhibited ethnic identification in a way that not even the old timers did, and children seemed to continue to be raised in the Spanish language. Thus the young adults re-find the Spanish language and Puerto Rican values, and differ most from the adolescent in 12-17 year old peer groups.5 Nearly the entire Block sample felt that one could speak Spanish and be American, just as one could speak English and be Puerto Rican. The same attitudes were shared by a second sample, although for them the politics of language was a keener issue. Most Block members said Spanish is not necessary to Puerto Rican cultural identity, but most of the teachers interviewed in a second sample said that is was. More teachers also felt that English was a menace, a political more than linguistic threat, to Puerto Rican cultural integrity.

New York Bilingual Teachers The second sample studied among New York Puerto Ricans was a random selection of forty (of 256) teachers and administrators in the public school bilingual education program. Their schools were also in El Barrio (District 4) and 90 percent (36 teachers) were Puerto Rican; all were Hispanic. The teachers identified their nationality as Puerto Rican in much greater proportion than the Block (Figure 1). They have intense feelings regarding the necessary connection between Spanish and Puerto Ricanness, as has been mentioned; but the teachers report speaking English more so than do the members of the Block. All the teachers, of course, were bilingual and had achieved at least a bachelor's degree. On bilingual mixing, and other points

34

John J. Attinasi

reflective of cultural synthesis, the teachers were not as adamantly opposed as had been anticipated. Rather, they recognized the sociolinguistic situation of young Puerto Ricans and of the community in migration and were dedicated to work with it, a point to which we shall return in a discussion of codeswitching. These teachers were strongly in favor of bilingual education, and quite articulate about its prospects and weaknesses, for reasons of both professional commitment and also personal experience as young Latinos growing up needing but not receiving bilingual education. 6 Women comprised only 35 percent of the Block sample, but constituted 68 percent of the teacher sample, and 75 percent of the third sample, to which we now turn.

Northwest Indiana Latinos A sample of sixty-five Hispanic housewives, teachers and industrial workers were interviewed with the rationale that these persons, who were actively contributing to Midwestern American society, would give opinions regarding viable options for Hispanic ethnic uniqueness and linguistic maintenance. It was assumed that the varied acquaintances and multi-ethnic work situations and schools in the Midwest would place the forces for cultural identity in confrontation with those of assimilation. Attitudes in such an environment, if similar to those in more ethnically-dense barrios, would confirm the different sociolinguistic situation of Spanish in the United States. If, on the other hand, Hispanics under such conditions of dispersal were shifting in language and attitude, this would confirm the prophecy of some commentators that assimilation is proceeding for Hispanics in a pattern similar to that of European nationalities earlier in the twentieth century. 7 Attitudes, history and social conditions argue against such similarities. The Northwest Indiana sample was neither a survey of the first persons who would be interviewed (i.e., a convenience sample), nor a random or purposive sample as were the two New York samples. Rather, they are the Hispanics who were interviewed as part of a course at Indiana University Northwest on language attitudes. Groups of students chose purposive samples of twenty (housewives or teachers, for instance). Of the one hundred personal interviews, sixty-five were Hispanics; they comprise the Northwest Indiana sample. The instrument was a selection of the most salient forty-five questions from the instruments used in New York, which contained nearly 200 questions. Fortynine of the Northwest Indiana sample were women (75.4%) and twentyeight were housewives (43%). In fact, one of the purposive sub-groups sampled was specifically Latina housewives, with the rationale that these wives and mothers who have the most contact with children would have attitudes

Hispanic

Attitudes

in Northwest

Indiana and New

York

35

Table 2 Sex & Occupation o f Northwest Indiana Sample in numbers and percentage. number

(%)

Housewife Housewife/Student Housewife/Whitcollar Housewife Subtotal White Collar Teacher Professional Artesan/Foreman Service Administrator/Manager Machine Operators Student Manual Laborer Sales Unemployed

22 4 2 28 5 9 4 5 3 4 2 2 2 1

(33.8 %) (6.2 % ) (3.0% ) (43% ) (7.6 % ) (13.8 %) (6.1% ) (7.6 % ) (4.6 % ) (6.1% ) (3.0 % ) (3.0% ) (3.0% ) (1.5 % )

TOTALS

65

(100 % )

Occupation

sex

M

F 22 4 2

-

4 7

1 2 4

-

5 2

-

1 4 1 2

-

1

_

1

2 1

16 (24.6 %)

49 (75.4 %)

-

Table 3 Birthplace b y Age o n Arrival Northwest Indiana sample Birthplace Total

Age on Arrival * 7-12 U.S.Born 0-6

U.S. Puerto Rico Mexico

36 (55.4 %) 15 (23.1 %) 14 (21.5 %)

36

Totals

65 100%

36

13-19

20+

_

—

6 4

2 1

4 2

3 6

1

10

3

6

9

1

* 56 or 86.1 % have spent more than 15 years in the U.S. Table 4 Place o f Education o f Northwest Indiana Sample Place

N/R

number

(%)

U.S. Mexico Puerto Rico P.R. & U.S. Mexico & U.S.

49 6 4 4 2

(75.4 %) (9.2 %) (6.1 %) (6.1%) (3.1%)

TOTALS

65

(99.9 %)

36

John J.

Aitinosi

Table 5: Friendships by sample (in percentage) Samples All types Puerto Ricans Hispanics Mexicans A fro-Americans Anglo-American Other Totals

Northwest Indiana

New York Teachers

78.6 9.2 4.6 6.1

32.5 42.5 12.5

-

1.5 -

100.0

-

2.5 2.5 7.5 100.0

New Yoik Block

_ 90.0 1.0 -

7.0 2.0 -

100.0

formative of the future of the Hispanics as their child-raising practice followed their views. Their language usage was also considered important, since it will appear again in the language of their children.8 Latino teachers in the Northwest Indiana sample numbered nine (13.8%); seven of them were women. Their attitudes were similar to those of the teachers in New York on many points, but their number was not large enough to separate them out as a significant sub-group. As was found in New York among the Block sample, the attitudes of the women did not differ greatly from those of the men on the cultural and linguistic items studied. Differences by sex are better visible through ethnographic analysis as Pedraza (forthcoming) has shown, and are questions of style and content more than attitude. The greatest differences occurred in comparing the three samples. (Tables 2 - 5 display social information on the sample.) The Northwest Indiana sample has been connected to the U.S. experience for some time, another characteristic of the future of Hispanics as migrants settle out and disperse in concentration from ports of entry to where they will reside permanently. Over eighty-six percent had lived in the U.S. for at least fifteen years; fifty-five percent were born in the U.S.; and of those born elsewhere (Puerto Rico and Mexico) one-third came before age six, one-third as children ages 7 - 1 9 , and one-third came as adults over 20 years of age (Table 3). Seventy-five percent were educated in the U.S. and another 10 percent were at least partially educated in the U.S. In friendships, too, the Midwest sample seems more integrated to the diversity of US society. Whereas 90 percent of the Block claimed mostly Puerto Rican friends, over 78 percent of the Northwest Indiana sample said its friendships were of 'all types'. The New York teachers' friendships lie between these extremes (42.5 percent Puerto Rican, 32.5 percent all types). Although the numbers claiming AfroAmerican friendships are small for the New York samples (7 percent and 2.5 percent), the absence of any such claim in Northwest Indiana hints at a difference in racial orientation that has been observed in comparing New

Hispanic Attitudes

in Northwest

Indiana and New York

37

York with the Midwest over the last two years. Although the rivalry and competition between Puerto Ricans and Blacks in New York cannot be denied, residence in Harlem and the racial spectrum of Caribbean peoples, including Hispanics and English speakers, invalidate the black/white dichotomy that still seems to persist in the Midwest. Depending on their orientation, Hispanics in the Midwest seem to consider themselves as either Vhite', observing the two race dichotomy; or 'other', attempting to promote Hispanics as an alternative group, a population with growing political importance. The politics of the Hispanic presence are ultimately intertwined with the maintenance of culture, language and educational opportunity. The research to be reported here, however, is less sweeping in scope, attempting to find out the personal attitudes of a population, and leaving for another time their implications for the wider political and economic arena.

Cultural Indentity National definition, like any other, has two main facets: (a) being distinct from similar entities, and (b) having a positive substance or set of characteristics to constitute the uniqueness of the group. Attitudes toward distinctness will be treated after exploring the variation within one cultural constituent, national name, or self-identification. Other aspects of cultural definition: values, work, food, custom, family roles, and of course, social interaction through speech, are best treated through social analysis, not survey instruments. Such research is needed, but for the present, aspects of reported nationality and reported language use that correlated well with observation in the New York research will be treated for Northwest Indiana. A Latina in the research team, a student and mother, remarked that the instrument might encounter resistance because it was asking personal questions like "What is your nationality?'. After several years in New York, where nearly everyone has an international affiliation of one type or other, it struck me as particularly Midwestern (and not at all beyond the 'melting pot' theory) that such a question would be taken as personal and sensitive. This merited further exploration, and so the question was refined to 'What do you consider your nationality to be?'. Other items of cultural identity were eliminated from the instrument except those that might correlate with the self-identification question. This was done first to further elaborate the meaning of the terms of nationality, and second because in the New York research it was discovered that self-report is most valuable on general items rather than on issues of detail. (This is especially true for microlinguistic affairs, such as dialect variation and code-switching, but is also true for daily cultural practice.)

38

John J.

Attinasi

The question of self-report of nationality resulted in fourteen separate terms (even after merging mexicano/a with 'Mexican' and puertorriqueño/a with 'Puerto Rican'). Four major designations, however, account for 80 percent of the response and these are, as expected, 'Mexican, Puerto Rican, Mexican-American and American' (Table 6 A-D). Conspicuous by its absence was 'Puerto Rican-American', which accounted for the plurality in the New York Block sample (Figure 1), and the term 'Latino' which may be a politically strategic term of coalition, not an item of nationality. Table 6: Self-report of nationality, Northwest Indiana sample, 'What do you consider your nationality to be?' Term of Nationality

Number

Percent

A. Puerto Rican B. Mexican C. Mexican-American D. American E. Miscellaneous 1® F. Miscellaneous 2

16 17 13 6 7 6

24.6 26.2 20 9.2 10.8 9.2

65

100.0

Totals

a. Hispanic, Spanish-American, Latin-American, Chicano, Puerto-Rican Spanish-American, Spanish-native-bom-American, U.S. citizen, (1 each, 1.5 percent each). b. American Puerto Rican, American first-Mexican second, Spanish, (2 each, 3.1 percent each).

Two miscellaneous categories are used to collect several separate answers. Miscellaneous 2 with two persons each saying things like 'American first Puerto Rican second', 'American-Mexican' or 'Spanish'. By the way, two who said their nationality was 'Spanish' were born in Mexico and both came to the U.S. at a young age. The category 'Miscellaneous 1' contained seven distinct terms, each offered once. It surprised me that 'Chicano' was offered by only one person. Also included were the ungainly 'Spanish native bom American', and the clinical 'U.S. citizen'. With such small numbers it is fruitless to speculate on any of these items here. Insight regarding the varied character of Northwest Indiana Hispanics can be gained through further analysis and comparison of the main nationality designations by cultural affiliation and place of birth. The New York block, New York bilingual teachers and the Northwest Indiana samples vary in proportions of'pure' VS. 'hybrid' VS. 'USA' designations of nationality. Just over half the Northwest Indiana sample gave pure

Hispanic Attitudes in Northwest Indiana and New York

39

Figure 1. Self-report of nationality among three samples: New York Block (inner), New York Bilingual Teachers (middle); Northwest Indiana (outer circle). (All figures represent percentages.) (Mexican or Puerto Rican) terms of nationality, as compared to 56.8 percent of the teachers and under 40 percent of the block. The block, using Spanish most and considered the core group after analysis of report and usage, was actually the highest in hybrid terms (Puerto Rican American 41.8 percentandNuyorican 17.6 percent). 'Nuyorican' is quite different from the more traditional ethnic-American terms used for many nationalities (Greek-American, GermanAmerican, Franco-American, etc.), and signifies an alternative bilingual and bicultural identification. It seems, however, that political content, language use and such designations are not related by chains of implication. Depending on how one interprets the 'miscellaneous' categories, hybrid terms account for 20 to 40 percent of the Northwest Indiana sample, but only about 10 percent of the teachers. About the same size group of teachers (10.8 percent)

40

John J.

Attinasi

considered themselves 'American' in nationality; the proportion was slightly lower among the Northwest Indiana sample. Only two percent of the block termed themselves 'American'. The paradox is that avoidance of the 'USA' designation seems to correlate with strength of ethnic identity, but the avoidance of hybrid terminology among the teachers does not correlate with low involvement in the American system. The teachers are interesting in that they designate themselves by their pure national term in greatest proportion, they have the most political views of the interaction between culture and language, yet their English language use and reported lifestyle is most mainstream. The proportions of these various responses are graphed in concentric fashion for the three samples in Figure 1 (see also Attinasi (forthcoming)). How these proportions relate to birth, affiliation and use of Spanish among the Northwest Indiana is the next subject to be explored. Table 7: National identity by cultural affilation, Northwest Indiana sample. Nationality

Part of both

Of U.S. Of Homeland

Totals

A. Puerto Rican

14 (21.5) (87.5)

B. Mexican

13 (20 ) 76.5

2 ( 3.1) (12.5) 3 4.6 17.6

C. Mexican-American

1 (1.5) 7.7

12 (18.5) 92.3

D. American

3 (4.6) 50

3 (4.6) 50

E. Miscellaneous 2

3

3

-

6 Row number

F. Miscellaneous 1

4

3

-

7 Row number

38 (58.5 %) 26 (40 %)

-

1 (1.5) 5.9

16 Row number (24.6) Total % (100) Row % 17 Row number (26.2) Total % 100 Row % 13 Row number (20) Total % 100 Row %

—

1 (1.5 %)

6 Row number (9.2) Total % 100 Row %

65 (100%)

The issue of cultural affiliation was approached through a question that simply asked: "Do you consider yourself part of the U.S., part of your homeland, or part of both?'. All three samples had majorities of 60 percent that answered 'part of both', clear evidence of the bicultural character of Hispanics in the U.S. Briefly comparing the remainders, both New York samples had nearly equal (15-20 percent) portions answering 'part of the U.S.' and 'part of Puerto Rico'. Only one person from the Northwest Indiana sample answered

Hispanic Attitudes in Northwest Indiana and New York

41

'part of Mexico' and the group answering 'part of the U.S.' constituted 40 percent of the sample. Inspecting the national identification of these respondents reveals a striking split in the sample. 'Mexican' and 'Puerto Rican' designations are overwhelmingly correlated with feeling part of both places, whereas the Mexican-American designation correlates with feeling part of the U.S. Even the persons who describe themselves as 'American' are evenly divided on cultural affiliation (Table 7). All of those born in Puerto Rico felt part of both places and over three of every five born in Mexico felt part of both as well; but three of five born in the U.S. feel part of the U.S. (Table 8). Table 8: Birthplace by cultural affilation of Northwest Indiana sample. Birthplace A. Puerto Rico

B. Mexico

C. U.S.

Totals

Paît of both 15 (23.1) 100 9 (13.8) 64.3 14 (21.5) 38.9 38 (58.5 %)

Of U.S.

_

Of Homeland

Totals

_

15 Row number (23.1) Total % 100 Row %

4 (6.2) 28.6 22 (33.8) 61.1

1 (1.5) 7.1

14 Row number (21.5) Total % 100 Row %

_

36 Row number (55.4) Total % 100 Row %

26(40%)

1 (1.5 %)

-

-

-

65 (100 %)

The point here should be briefly concluded by saying that one of the main differences between the cultural affiliation of the New York samples and the Northwest Indiana sample is that the feeling of homeland affiliation is nearly not present in the Midwest sample. Beyond that, Mexican-Americans so designated did not even feel 'part of both', whereas a strong majority of self-termed Mexicans and Puerto Ricans did. A further difference between the two research sites can be seen in the scale of distinctiveness, a point mentioned earlier. The attitude toward remaining distinct as a group contained two peaks in the Northwest Indiana sample, one at midpoint and one at the upper end of a seven-point scale. For the New York samples, however, the curves are almost entirely at the upper end of the scale. The highest point of the scale, namely, It is very important for Hispanics to remain distinct', was chosen by 85 percent of the teachers, 70.3 percent of the block but only 20 percent of the Northwest Indiana sample (points 6 and 7 together constitute 40 percent for the Northwest Indiana sample; see Figure 2). If ethnicity scales could be created for the factors treated in this section, then, the locale of Northwest Indiana would

42

John J. Attinasi

Figure 2. 'How important is it for Hispanics to remain distinct?' (1) Should Not (2) Not Important (3) Makes No Différence (4) It is an Option (S) It is Good (6) Important (7) Very Important.

Hispanic Attitudes in Northwest Indiana and New York

43

be below New York; peisons born in the US would be below those born in Spanish speaking countries, and in national designations (independent of birthplace) Mexican-Americans would be below 'Mexicans' and these below 'Puerto Ricans'. Before a Mexican Midwest revolution is started over that last statement, several points should be emphasized: that a majority feels connected to both places; that the 'Mexican-American' group is only 20 percent; and that language questions which are the present focus of this report, are connected to cultural attitudes in complex ways. Thus there is no surprise in the scale proposed if the extremes are viewed: a New York Puerto Rican born in Puerto Rico is probably more firm in ethnic identity — and also Spanish language attitudes and use — than a U.S.-born Mexican-American. But between those extremes personal history and variation of a complex nature must be taken into account, in a manner beyond the scope of this essay, before one person or group is placed above another regarding attitudes, language maintenance or identity.

Languages Attitudes and Use Language provides one of the most visible and comprehensive ways to remain culturally distinct. Accordingly, the desire for ethnic distinction is substantial for all three samples. The New York block was unequivocal; 100 percent desired Spanish to be maintained in the Hispanic community; the teachers (94.6 percent) and Northwest Indiana (92.3 percent) samples were also nearly unanimous. National identity and cultural distinction through language maintenance, however, is more than a matter of claim;it also involves practice. Language performance and participation in the education of the young are part of securing the language and culture of Hispanics. Bilingual education is relevant here as a means to preserve cultural viability through Spanish language development in a society that demands skills and academic proficiency in English. The next several pages review findings on both language use and bilingual education. The attitudes and self-report of the block sample benefitted from verification by the ethnographic observation of Pedraza and the sociolinguistic studies of Pousada and Poplack (1982; Language Policy Task Force, forthcoming). From those investigations, the three items of self-report that were found to be most indicative and reliable were those of: (a) best (or 'dominant') language; (b) the language one felt most comfortable speaking; and, (c) the language used in socializing with friends. Ten to fifteen percent more block members reported their best language as Spanish in comparison to the other sample; about twenty percent fewer

44

John J. Attinasi

block members reported English as their best language (Table 9). For most comfortable language, the block is still higher in Spanish and lower in English than the other samples, but the spread narrows to about ten percentage points regarding English as the most comfortable language. Because of the English speaking ambience of New York (and the U.S. in general), some block members report being actually more comfortable speaking the language they do not consider their best! The Northwest Indiana sample scored somewhat higher than the teachers in three of the four Spanish and bilingual categories in Table 9. Their greater use of Spanish is also seen in Figure 3.

50. 48. 46. 44. 42. 40. 38. 36. 34. 32. 30. 28. 26. 24. 22. 20. 18. 16. 14. 12. 10. 8. 6. 4. 2.

Figure 3. Language used in socializing. S = Spanish B = Bilingual E = English

Northwest Indiana . _ New Yoik Teachers ....New York Block

Hispanic Attitudes

in Northwest

Indiana and New York

45

Table 9: Most comfortable and best language by sample (in percentage).

Best language Spanish Bilingual English Other Most comfortable Spanish Bilingual English N/R

Northwest Indiana

New York Teachers

New York Block

18.4 35.4 46.2

15.0 30.0 52.5 2.5

29.7 39.6 29.7 1.0

16.7 25.0 58.3

33.0 19.7 47.3

-

23.1 20.0 55.4 1.5

-

-

The report of language used in socializing indicates several key points of difference among the samples. First, the teachers and Northwest Indiana samples show quite low percentages of monolingual Spanish use. For the teachers, however, bilingual usage peaked and purely English usage was high (37.5 percent), but lower than that of the Northwest Indiana sample. The block shows a more even slope, all three points in the thirtieth percentile (Figure 3). A composite of several self-report measures (best language, most comfortable language, language used in socializing and in several other contexts) yields an even clearer picture of the language use of the three samples (Figure 4). The block's rather even distribution of English and Spanish (30 percent each) is complemented by a peak of bilingualism (40 percent). The bilingualism of the New York block, then, seems to be in a situation of stable bilingualism at present, with the young people probably moving toward a situation more like that of the teachers. The teachers, while lowest in Spanish monolingual usage of the three samples, are highest in bilingual usage (50 percent). Their usage of English is substantial (35 percent), and their situation appears to be one of a bilingualism that favors English, but retains Spanish adamantly with strong attitudinal support. Northwest Indiana shows an ascending slope toward English usage, a situation that seems to indicate language shift for the future. A conflict arises then in the Northwest Indiana situation that may be indicative of the future of Spanish in the less dense communities of the U.S. outside centers like New York, Miami, Chicago, San Antonio and Los Angeles. Although most want to see Spanish continue, the pattern of usage seem to indicate less Spanish, and even less of the bilingual usage that would retain Spanish as a component in the speech repertoire of Hispanics. Monolingual Spanish usage, which correlates with age and with age at time of migration,

46

John J. A ttinasi

is decreasing, but need not imply wholesale shift if bilingualism is active and stable. Bilingual education and a societal increase in the use of Spanish become key factors in the struggle for the survival of Spanish. Maintenance, therefore, will require the social conditions of the block — numbers, density, migration and the reactivation of Spanish among the young adults — or the consciousness and efforts as displayed by the bilingual teachers. The challenge for Spanish in the Midwest, most likely, will be to increase Spanish usage and to support bilingual education. The result would not, then, be language shift, but maintenance of an interpenetrating bilingualism with greater fluency in English, as seems to be the trend in New York among both teachers and younger adults (Flores, Attinasi and Pedraza, 1981: 195—202).

50. 45 40 35 30. 25 .

20. 15

Northwest Indiana

10.

New Yoik Teachers

5 .

New Yoik Block

Figure 4. Main language resources. S = Spanish B = Bilingual E - English

Hispanic Attitudes

in Northwest

Indiana and New York

47

Change in Use of the Languages Except for the New York block, which as we have seen, is already the most Spanish-speaking of the three samples, a large numbers report an increase in Spanish usage over the last few years. (From Table 10 this can be seen to be most pronounced among the New York teachers). At the same time the increase in English usage is also an undeniable pattern, with about 70 percent of both the New York block and the Northwest Indiana samples reporting speaking more English. Note also that more than a quarter of the Northwest Indiana sample reports increases in both Spanish and English, a sign that bilingualism is part of increased usage of language in general for many (Table 10, note a). It is also striking how few in the Northwest Indiana sample report a decrease in their use of Spanish. These results again argue against total shift and for bilingualism (Arce and Carranza 1981). Table 10: Reported change in language use by sample (in percentage). Spanish Increase Same Decrease N/R

Northwest Indiana 49.2 a 33.9. 13.8 b 3.1

New York Teachers 60.0 15.0 25.0

New York Block 35.0 27.5 36.2 1.3

English Increase Same Decrease N/R

70.8 21.5 7.7 C

50.0 42.5 7.5

68.1 14.2 13.8 3.9

-

-

a. 27.7 percent report an increase in both. b. 9.2 percent report more English and less Spanish. c. 3.1 percent report more Spanish and less English; 3.1 percent report decreases in both.

Bilingualism through Education Despite the persistence of the term 'dominant language' and the view that education should take place in only one language, whether Spanish or English, the use of both languages in the Hispanic community seems to be both a desired option and an actual practice. The perception in New York (by over 75 percent of both samples), is that there are 'many' in the community who speak 'both mixed'. In Northwest Indiana, about 40 percent report extensive bilingual usage, and the estimate of speakers who usually communicate bilingually is about the same size. It

48

John J. Aitinosi

is not appropriate here to explore the different attitudes toward mixing the languages VS. keeping them separate except to say that the reported use of 'both mixed' among block members in high proportions, compiled further with their minimal report of 'both separately', indicate that code-switching has the quality of a norm in New York that it does not in Northwest Indiana. Some focused comparisons regarding bilingualism further emphasize the kind of alternative to monolingualism in either language that is seen in the way many Hispanics speak. Comparing the teachers to the Northwest Indiana sample, in no case does a majority feel that mixing languages was harmful, although nearly half of the Northwest Indiana sample said mixing was linguistically detrimental. Many of these used the word 'confusing' in their explanation, others dogmatically stated, 'it's wrong', 'sounds bad' or 'should not be done', (linguistically, and in practice, such dogmatism has little force; Attinasi, et al. 1981; Poplack, 1980). It might be noted that in both samples, considerably fewer said it was culturally harmful to mix; and many Midwesterners mentioned the bicultural reminder that code-switching can provide.

Table 11: 'Do Hispanics who only speak one language divide the Community?' (Answers are in percentage)'. Northwest Indiana Those who speak only English Yes Somewhat No N/R Those who speak only Spanish Yes Somewhat No N/R

New York Block

43.1a 6.2

-

50. l b

68

-

32

-

3S.4C 6.2 56.9 d 1.5

13 -

86 1

a. 23.1 percent said 'yes' to both. b. 18.5 percent said speaking only English does divide, but speaking only Spanish does not c. 6.1 percent said speaking only Spanish divides, but speaking only English does not. d. 38.5 percent said 'no' on both.

Hispanic Attitudes in Northwest Indiana and New York

49

Another comparison of perceptions of the effects of the sociolinguistic situation may be seen in attitudes toward monolinguals of each language in the community. Does their presence divide the community? The Northwestern Indiana sample said no, as did the New York block; and in both samples monolingualism in English was seen as divisive by more persons than Spanish monolingualism. The overwhelming negative response among the New York block regarding those who only speak Spanish, however, indicates again the baseline quality of Spanish in that Hispanic community (Table 11). In Northwest Indiana, the norm of bilingualism seems to place a nearly equal obligation on Spanish speakers as well as English speakers to learn the language they do not know.

Spanish Maintenance through Bilingual Education It has been established that the Hispanic community does not favor leaving Spanish behind while it desires English; and that its speech pattern reflects this desire for bilingualism despite a low incidence of Spanish-only speech, and substantial English-only or mostly-English usage. Moreover, nearly all consider it an advantage to be bilingual and nine out of ten would want bilingual education for their own children (Table 12). The younger generation, however, is perceived to speak the two languages by only about twenty percent of the Northwest Indiana sample. As we shall see, it is hoped that bilingual education will be a means for that maintenance.

Table 12: Spanish Maintenance in Northwest Indiana (in percent).

Yes No Optional

Spanish maintained in Hispanic community

Advantage to be bilingual

Want children in bilingual education

92.3 4.6 3.1

98.5

89.2 7.7 3.1

-

1.5

Whether bilingual education was perceived as a privilege or a right is another issue that deserves mention. As a result of the Lau decision, numerous other litigations, and a publication entitled Igualdad en ¡a educación: un derecho (Illinois advisory committee to the U.S. Commission on Civil Rights, 1976), the right to bilingual education has been established and publicized. In the view of the Northwest Indiana sample, a close majority thought bilingual education to be a right. Nearly 48 percent said it was a privilege; but their

50

John J. Attinasi

responses were not all clear opinions that bilingual education is only an option or a frill. About a quarter of those who said bilingual education is a privilege said so for reasons that do not indicate a legal option, e.g., because it is 'not present in all schools', it is an 'advantage', or it should 'make one proud to have it'. Others in this sub-group gave ironic explanations for their views such as 'you're lucky if you can get it' and 'it's not a right in this country, Reagan is trying to cut it'. Another took a more forward-looking stance, calling for local control, 'Bilingual education should not be a government decision; they don't care, and will say there is no money for it'. Since three-quarters of the sample perceived Latino children as speaking English and ninety percent consider bilingualism desirable, the purposes and eligibility standards of bilingual education are logical next questions. The Northwest Indiana sample was emphatic regarding maintenance characteristics and biculturalism as the primary goals of bilingual education. Table 13: Purpose of bilingual education Northwest Indiana (in percent). a. To teach two languages and two cultures b. Teach English to non-English speakers c. Teach Spanish and foreign languages d. Improve image and self-concept e. Help children keep up studies f. N/R

52.2 19.4 11.9 8.9 1.5 6.2

Over fifty percent saw the bilingual aspect of bilingual education as primary, whereas only about twenty percent named English for non-English speakers (transitional schooling) as the primary purpose of bilingual education (Table 13, rows a, b). About a quarter said bilingual mastery should be the exit criterion (Table 14, row b), and nearly thirty-seven percent (Table 14, row a) implied their support for bilingual education as an alternative form of education in their opinion that children should remain in bilingual classes throughout school. Forty-five percent of the New York teachers, but only twenty percent of the block, registered the opinion that bilingual education should develop the languages and knowledge of two cultures. Another twentythree percent of the block, however, thought bilingual education should be teaching Spanish to Hispanics and twenty percent saw in it a means for general education. This indicates community support for language maintenance in bilingual education, but non-response on the part of 30 to 60 percent on bilingual education questions betrayed a lack of communication between the schools and the community.

Hispanic Attitudes

in Northwest

Indiana and New York

51

Table 14: Amount of time a child should be in bilingual education — Northwest Indiana sample (in percent). a. Throughout schooling b. Until two languages are mastered c. Until English is mastered d. Three years e. Depends f. N/R

36.9 24.6 23.1 9.2 1.5 6.2

As might be expected, the teachers directly involved in bilingual education perceive its success to be the greatest. Figure 5 graphs the split opinion of the Northwest Indiana sample regarding the success of the programs, and the overriding non-response of the New York block on the issue. It seems, then, that the opinions of the community might be potentially turned into demands for maintenance, biculturalism, and a responsive, successful program as a democratic right of taxpayers, if the subjective concerns registered here can be transformed from symbolic grounds to instrumental grounds for planning and intervention in the linguistic and educational state of affairs (Kelman, 1971). It is here that parent involvement, consciousnessraising and active participation in educational policy may result in the kinds of change that respond to the attitudes registered in these surveys and contribute to language maintenance.

The Prospect of Spanish Maintenance in the US Well over half the Northwest Indiana and New York teachers samples consider language maintenance a family responsibility, although it is clear that many looked to the schools for the elaboration of literate and academic skills (see Table 15). The family environment clearly sets the tone for school performance in language and other areas of academic training. Studies have shown that proficiency in the home language, whether Spanish, English or both, is the best indication of academic success (Troike, 1981; Cummins, 1980). The schools, especially bilingual maintenance models of education, can aid in the elaboration of the phonological, interactional and grammatical intuitions of the Spanish-speaking child. But the home environment, in both practical and attitudinal aspects, remains crucial to the maintenance of Spanish. The attitudinal issue regarding the maintenance of the Spanish language in the U.S. boils down to another question, which we were blunt enough to ask the three samples outright: 'Is it necessary to speak Spanish to be Hispanic?'. In the earlier studies 83 percent of the block said 'no', 62 percent of the

52

John J.

Attinasi

teachers said *yes\ But both groups considered Spanish the baseline of language of the culture and could easily characterize or envision themselves as a bilingual people, but never an English-speaking people. In Northwest Indiana 78.5 percent say speaking Spanish is not necessary to being Hispanic, 18.5 percent say that it is necessary and 3 percent did not respond. Thus, those who said Spanish is an essential part of cultural identity and who said, 'a person should at least communicate in their own language', were outvoiced by those who said things like: 'being Hispanic is in the blood', 'the heritage is more than just language', 'there are many non-Spanish speaking Hispanics in the U.S.', and, as one said, 'it (speaking Spanish) is nice, but not necessary'. About fifteen percent interpreted the question in another sense, reminiscent of some answers heard in New York. Several said things like 'there are people of other nationalities who know Spanish', and *we do not have a monopoly on the language'. The latter interpretation is the logical reverse of the question, for it states: One need not be Hispanic to speak Spanish; in other words, that Spanish language need not imply Spanish heritage (S ** H). We were actually asking whether Spanish heritage should imply knowing the Spanish language (H -> S). Table 15: 'Who is responsible for maintaining Spanish?' (answers represent percentage of indicated sample group).

Family All Family and schools Schools Community groups N/R

Northwest Indiana

New York Teachers

New York Block

61.6 27.7 3.1 4.6 1.5 1.5

56.8 24.3 13.5 5.4

37.4 49.5

_ 11.0

—

2.1

-

-

Beneath these intricacies there remains a strong attachment to the Spanish language that is difficult to tabulate, but emerges in qualitative inquiry. Few in the Northwest Indiana sample would agree with the assimilationist position of one Mexican-American who said that English is the 'main language' and Spanish 'should not be maintained'. (He also indicated that he speaks only English in every context surveyed, and considered himself part of the U.S. He did, however, consider bilingual education to be a right.) Another Mexican-American (self-termed) showed us further complexities in the sociolinguistic attitudes of Hispanics in the Midwest: he considered

Hispanic Attitudes in Northwest Indiana and New York

53

N/R of NY Block 62.5 %

Northwest Indiana N.Y. Teachers N.Y. Block

o a

1 o e

o if E

J .c« oaa o m dijo la giganta "yo creo que ya me siento más bien." #27 El esta comiendo tomates y después ella estaba enferma, después, uno da pan, otro da el trombón, y el otro le da plantas.

después

Perspectives on Strategies In this study we discover the strategies used by a small group of children whose task is to retell a story as part of a language proficiency test in Spanish. All of the children whose narratives are included in this study are growing up with Spanish as a first language or with both Spanish and English as first

94

Elizabeth A. Leone and René Cisneros

languages; for the majority Spanish has not been the primary language of instruction. The strategies identified for analysis and description fall into three general types or categories: narrative strategies, cohesive strategies, and communicative strategies. These categories are thus identified because such distinctions are supported by different language behaviors in the data; however, they may overlap and a particular example may include more than one type of strategy. This study does not consider the non-adult variants of the children to be 'errors'. Rather, it makes the assumption that these strategies can and will be maintained by the children and others in the community if they are recognized and valued by the speakers themselves. Storytelling strategies are assumed to be deliberate and can be defined as devices used to skillfully convey a message. Included in the evaluation of the effectiveness of strategies are channel, code, receiver of the message, setting (Hymes, 1972). Persons researching second language acquisition have identified various types of strategies: communication strategies, learning strategies, and production strategies (Tarone, 1980). Distinctions between communication strategies on the one hand and learning strategies on the other, focus on the purpose of the strategy - communication strategies emphasize the negotiation of meaning by speaker and hearer, whereas learning strategies emphasize the learning of the language and its strategies. Production strategies, similar to both communication and learning strategies, emphasize primarily the use of the linguistic system in the most efficient and clearest way (Tarone, 1980). For purposes of this study, the types of strategies can be considered to include all three of these types (communication, learning, and production), though the analysis and description will focus on the strategies in a slightly different fashion. Perhaps one of the most helpful notions related to strategies is that of planned and unplanned discourse (Keenan, 1977). Due to both situational and conceptual demands, speakers often do not plan their talk as much, and certain features tend to prevail. The four features identified by Keenan are the reliance on immediate context to express propositions and on morphsyntactic structures acquired earlier, as well as the repetition/replacement of lexical items and the similarity of form/content of sequential social acts. In the analysis of storytelling strategies in this study, the division of the strategies into three categories of strategies - narrative, cohesion, and communicative strategies confirm many of the assumptions made by Keenan concerning discourse that is relatively unplanned. There are some differences, nevertheless, to be kept in mind. There are two constraints on the situation: (1) the children have to retell a story that they have just heard, and (2) therefore this story 'frame' creates certain morph-syntactic, lexical, and speech event guidelines. Because of these the children do not have to plan

Children's Storytelling Strategies

95

as much, or at least may have to use different kinds of strategies in order to do the task required of them. Another notion related to discourse that is helpful for the analysis of the children's storytelling is the notion of repair. (Schegloff, 1973; Schegloff, Jefferson and Sacks, 1976.) Repair has been defined as an attempt to fix a problem in discourse, by the speaker or by the hearer, and can take a variety of forms, depending upon the item to be repaired, the repairable. In unplanned discourse, these repairs may appear more frequently, and in certain situations where the speakers are attempting to communicate according to certain 'rules', they may also need to use repairs more often. Shimanoff and Brunak (1977) discuss repairs in planned and unplanned discourse, examining the differences and similarities. They present a useful paradigm for the analysis of repairs, including the notion of hierarchies of repairs, markers of repairs, and some of the principal types of repairs. With respect to the analysis of children's storytelling in this study, the repairs of different kinds — morpho-syntactic, lexical, and pragmatic — will be discussed in the different sections describing the three kinds of strategies.

Narrative Strategies Strategies used to narrate the story by the children, reflecting the specific characteristics of the narrative as a genre, will be referred to as narrative strategies for the purposes of this study. Narrative characteristics noted by Labov and Waletzky (1967) and Labov (1972), include features such as introducers, orientation, complicating action, and resolution. Others have dealt with specific aspects of narratives: Kernan (1977) discusses elaboration in children's narratives, Walcutt (1977) discusses narrative boundedness and the back and forth movement between general and specific, for purposes of focusing. Although perhaps all strategies used could be seen as related to narration, this study examines four narrative strategies: introducers, orientation, complicating action, and resolution. The structures used to accomplish these narratives strategies will be discussed as part of the strategies themselves.

Introducers Introducers are those phrases which mark the following discourse as that belonging to the narrative of a story. Some are stylized as in the case of habia urn vez ('once there was') or esto es la historia de ('this is the story o f ) , or era una vez ('once there was'). In some cases, however, the introducer is simply a temporal adverb such as un dm ('one day') and in some cases the

96

Elizabeth A. Leone and René Cisneros

story begins with the verb ser or haber marked with the imperfect aspect as in era and había. With the exception of esto es la historia and un dm all the introducers included some use of the imperfect aspect to introduce background information into the narrative. This aspectual marker is also used in the orientation strategy to give background information. In the twenty-one narratives included in Table 1, the following introducers were used: (1) había una vez or una vez había - in 7 (2) era una vez — in 1 (3) era — in 3 (4) había - in 2 (5) una vez estaba — in 1 (6) esto es ¡a historia de — in 1 (7) un día — in 2 In other words, of the complete narratives produced, seventeen used some kind of introducer — about eighty percent. (The story input that the children heard included the introducer había una vez, which is perhaps the most common for this type of story and was the most common type found in the children's narratives as well.)

Orientation

and Complicating

Action

In addition to the introducers, Labov (1972) notes two other traits in narration, that of orientation and of complicating action. The children's narratives analyzed in this study also have these two characteristics. In order to express orientation and complicating action, the children's narratives make use of the syntactical device available in the Spanish language for marking or making the distinction between these two narrative traits. That is, the imperfect and preterite tenses in Spanish are used to indicate aspect: the preterite records actions that are completed already and the imperfect records actions descriptively in the past. Both tense-aspect categories together provide for the various conceptual 'angles' for viewing any event: at the beginning, at the end, and in the course of its occurrence. (Stockwell, Bowen and Martin, 1965: 134) The ability to use both these tense-aspect categories appropriately in the narratives was a strategy most of the children possessed, as summarized below. (Also see Table 1.) (1) all of the children used both the imperfect and preterite tense-aspect categories except two (of the twenty-one complete narratives analyzed): imperfect tense-aspect was used to describe:

Children's Storytelling Strategies

97

(a) (b) (c) (d)

in the introducers {había, era, estaba, hacía calor)', what the giant liked to eat {quería, gustaba, comía); what the giant thought the paint was {creía, parecía, era); how the giant felt after eating the paint {sentía, sentaba, estaba, enfermaba)', (e) what the giant was never going to do again {iba comer, volvía)', preterite tense-aspect was used to report: (a) what the giant ate {comió, fue a comer, tomó, quiso comer)', (b) what happened to the giant after eating the paint {se enfermó, se puso enferma, felió, sintió, hizo)', (c) that the friends came to visit {vinieron, vino, llegaron)', (d) that the friend gave gifts {dio, dieron, trajo, trajeron, trayó, trajió, trajimos)', (e) that the giant and friends talked {dijo, dijeron)', (f) how the paint tasted {supo, gustó); (g) that the giant thought about the paint/ice-cream {pensó); (2) of the two narratives that did not use both imperfect and preterite, one used a combination of present and preterite and the other used a combination of present and imperfect. (See narratives #17 and #18 in Table 1.) Thus, even the two children who did not use the combination of imperfect and preterite tenses did alternate between at least two tenses, to indicate a difference in time and aspect. The present tense was also used by other children, chiefly in the context of quotations of the friends and the giant speaking to each other. Compound tenses and sequencing of tenses in addition to direct and indirect quotations are noteworthy in the children's narratives and are summarized below: (3) of the 21 narratives 12 included direct quotations, though some of these also included indirect quotations. Most of the verb tenses within the quotations were either present or preterite, depending upon the message being conveyed: "estoy mal;" "no vuelvo a t o m a r . . ;" "no es helado, es pintura;" "ahora me siento un poco mejor;" "te trajimos unos regalitos;" (4) of the direct quotations used in the narratives all were framed but one. That is, all direct quotations were preceded by some marker such as dijo or dice to indicate that a direct quotation was to follow.

98

Elizabeth A. Leone and René Cisneros after receiving the gifts. (d) mention that the giant said 'thank you' to the friends for the gifts they brought.

All 21 included (a);the 12 complex ones included (b); 6 of the complex ones included (c); 4 of the complex ones included (d). All components were included in the original story used as input for the retellings. (5) of the 21 narratives only 6 contained indirect quotations. All contained tense sequences which maintained the sense of the indirect quotations, even though some used dijeron where preguntaron was intended: (a) (b) (c) (d) (e) (f) (g) (h) (i)

"dijo el gigante que ya no iba comer pintura roja." (#1) "y dijeron si se siente bien y dice la giganta que no." (#6) "comió pintura roja y luego dijo que no era." (#14) "el garganta grande dijo que le trajio un pan fresca." (#14) "entonces dijo que no se sentía bien." (#19) "la dijeron que como su . . . se sentía." (#20) "la giganta dijo que muy mal." (#20) "la giganta dijo que ya no volvía a comer pintura roja." (#20) "y dijo que no le supo bien." (#23)

Although some of the tense sequences are not in the most expected forms, all demonstrate an ability to handle two tense-aspect categories juxtaposed to each other. (6) there were few instances of compound or complex tenses in the children's narratives: one child used the imperfect progressive tense; one the subjunctive present perfect: estaba comendo (#18) haiga puesto (haga/haya) (#25) The child who used the imperfect progressive did not use either the simple imperfect nor the preterite tenses. The child who used the very sophisticated subjunctive tense had recently settled in the area directly from Mexico, already schooled in Mexico at least four years.

Resolution The final characteristic of narrative is giving an ending to the story. This ending consists of a solution of the conflict, and consists of several stages in the story of the giant analyzed in this study. After the apex of the story — the giant's mistaking of red paint for strawberry ice-cream and becoming

Children's Story telling Stra tegies

99

sick - all children ended with a resolution: 12 included a more complex version, 9 a simpler reduced one. The components of the resolution were: complete version : (a) mention that after the giant became ill, she was visited by a group of friends who brought her gifts. (b) mention that the giant said that she would never eat paint again. (c) mention that the giant said that she felt better

Cohesion Strategies In addition to narrative strategies — verb usage, introducers, indirect and direct quotations and framing — the children used certain strategies of coherence and cohesion that held the narratives together. These strategies can also be seen to overlap with the narrative strategies, especially conjunctive and temporal connectives, but they are also general strategies of cohesion characteristic of all texts (Halliday and Hasan, 1976).

Conjunction One of the most basic types of cohesion is the use of conjunctions such as and (y), then (luego) and adverbial phrases of time, place, etc. These words are cohesive because of the meanings attached to them and not necessarily because of any type of anaphoric relation with other words in a text (Halliday and Hasan, 1976: 226). Nevertheless, conjunctions do specify relationships between meaning already textualized and meaning that is yet to be textualized. The relationship of succession, in time, primarily, is the predominant type of relationship expressed by conjunction. This sequential relationship in time is also very important to the cohesion of narratives, when events are often related in a chronological order of occurrence. In the twenty-one complete narratives included in Table 1, the following usage of conjunctive cohesion can be summarized: (1) all used the conjunction Y: some of these uses of y imply temporal succession while others simply imply the succession of items in a series, without reference to time; 4 of the 6 non-complete narratives used y in them; Other uses of the conjunction y demonstrate a variety of the functions of the w o r d a n d the semantic relationships that it may entail: (a) attributive: era una gigante y estaba morada (#1) (b) additive: (items in a series): y el gigante grande le trayo pan,

100

Elizabeth A. Leone and René Cisneros y el gigante mediano le trayó palomas rojas, y el gigante pequeño le trayó . . . (#19.) (c) cause and effect: y una vez comió pintura y se enfermó. (#17) (d) succession in time : sus amigos vinieron y dijeron (#2)

Other combinations of adverbials with y include the following: y y y y

luego luego el siguiente dia de ahí ahora

y y y y y

entonces después luego entonces nunca una vez

All of the above combinations indicate some kind of temporal relationship that y is engaged in, even though the proximity of y to the adverbial units following does not in itself indicate that y belongs to the same constituency as the adverbials. Certainly the scope of y is perhaps wider than the scope of the adverbials. Another characteristic of the conjunction y that was not analyzed in this study story by story, but that nevertheless was a predominant function encountered in the children's narratives, was its use as a filler. That is, when the children were narrating they often marked junctures, when pausing between clauses, with one or more utterances of y. (2) one used the conjunction porque: only one child used the causal conjunction porque, and this child used it three times, though one occurrence was part of a self-monitoring or self-repair: — no quería comer nada porque estaba bien mal — no quería al comer pantia porque le están mal Both uses are very similar — one serves as a repetition of the other.

(#12) (3) one used the conjunction también: in the process of searching for a lexical item for the gift that the smallest giant gave to the sick purple giant, one child simply repeated the gift for the previous, mediumsized giant: — los viejos traen pan, y uno trae floras, y el . . . un trae floras también. (#16) (4) uses of conjunctive elements tended to repeat within a single child's narrative: when one conjunctive element was used by a child, it tended to be used more than once by that same child — just as porque was used three times by the same person. One child used y de ahí

Children's Storytelling Strategies

101

three times (#6); several children used luego, sometimes with y, several times throughout the narrative (#1, #3, #14, #23, #26); two children used después several times (#20, #27); a few used entonces several times (#19, #24, #25). As mentioned above, often the children repeated the same conjunctions in their narratives, though the different instances of the conjunctions were not necessarily serving similar functions. However, the use of repetition as a cohesive device is well known, whether it involves repetition of conjunctions or of other linguistic and paralinguistic elements.

Pronominalization, Referencing, and Anaphora Halliday and Hasan (1976) state that substitution is more of a grammatical type of cohesive relation than a semantic one, and for this reason we treat pronominalization not as substitution but as a referential cohesive relation, implying primarily semantic ties of nouns with their substitutes. That is, the relation occurring between nouns and their referential substitutes is such that the substitutes (personal pronouns, demonstratives, and comparatives) presuppose a relationship with something and are actually meaningless without reference to a textual or contextual identity (1976: 3 Iff.) Most of the types of reference that occur involve nomináis, and these are the types analyzed in the children's narratives. Although the class of reference involving person reference is often divided structurally by roles such as nouns, adjectives, determiners, and pronouns, these will all be treated in this study as personal pronouns/reference. The personal pronouns were the predominant type of referential cohesion in the children's narratives. The following summarizes the use of various kinds of pronouns and reference: (1) All used some type of personal reference: all children except one telling complete stories used some type of personal pronoun — subject, object, indirect object, possessive adjective, or reflexive. Of the 6 not telling complete narratives, 2 did not use any personal reference at all. Examples of different uses of personal reference are: reflexive: Y luego se enfermó. indirect object: y el chiquillo le trajo un pito . . . direct object: Se lo comió y se puso enferma. subject: Y ella se sentía muy mal. possessive adjective: Y sus amigos vinieron y . . .

102

Elizabeth A. Leone and René Cisneros

(2) Other children's stories included demonstratives and a combination of anaphoric references; many of the children included other kinds of reference that provided cohesion in their narratives: demonstrative: ¡Esto es pura pintura! the one: (see Halliday and Hasan, 1976: 92ff. on nominal substitutes) y el mas grande le trajo unas flores, y el mediano le trajo unas flores, y el chiquito le trajo . . . one, the other: el grandote trajo un poco de pan. El otro trajo cana. El otro trajo pito. The last types of examples demonstrating use of the one, one, the other are all more like substitution as a grammatical relation, and always presuppose previous mention in the text of the thing substituted. In other words, this type of cohesion is different from the personal reference type mentioned earlier in that it always ties a particular element in a text to something already mentioned in the same text. This is not necessarily the case with personal pronouns, for instance. Anaphoric cohesion is perhaps the simplest kind: one element in a text refers back to another element in the same text. (Halliday and Hasan, 1976: 14) Many instances, as cited above, are evident in the children's narratives, of precisely this type of cohesion.

Lexical Cohesion and Repetition When a lexical element is repeated in a text a very basic type of cohesion is evident. Although the repetition-referencing of nouns is considered the most common type of lexical cohesion, other types of repetition exist which add to textual cohesion. Repetition in general is thought to include not just exact reiteration but also the use of derivatives, synonyms and other closely associated semantic alternates. In the stories analyzed in this study, repetition of lexicon with grammatical roles other than nouns was found: repetition of conjunctions of time and of a large variety of verbs in their different forms. Table 2 lists the various lexical substitutes for the nouns only.

Semantic-field

Alternates

As the children narrated their stories lexicon which appeared in the original story often underwent changes. In some cases, the children retold the story without making any changes in the lexical units. In such cases we could state that the children found the lexical items in their own lexicon and retold the

Children's Storytelling

Strategies

103

story without making or needing to make any changes. In one or two cases, the children even go beyond the original story and provide a more elaborated version. However, in the majority of story retellings the use of lexical items reveals systematic alterations.

Table 2: Lexical substitution in 27 stories. La giganta:

una giganta, el gigante, una gigante, una giganta morada, he, él, ella, la giganta, un muchacho, un señor gigante, la señora gigante, una mujer, el giranta, el giranja, un grandota, un man, un viejo, un grande, un gigante, el elefante grandote, una giganta muy grande, una giganta chistosa,

Los amigos: unos amigos, sus amigos, los amigos, la amigos, estos amigos, los amigos de ella, tres amigos, los viejos, unos gigantes, they, los tres gigantes, los muchachitos, su amigos, tres amiguitos, ellos; el grande:

el primer gigante, el grande, la grande giganta, el gigante, el grandote, uno, el garganta grande, un, el grande gigante, el gigante grande, el gigante más grande, el más grande;

el mediano: otro mediano, el mediano, la mediana, el otro, el miedá, la otra, el gigante mediano; el pequeño:

el chiquito, el pequeño, el chamaquitos, el otra, el chiquillo, el gigante pequeño, el chico, el gigante chiquito;

helado de fresa: todo lo que se parecía fresa, nieve de fresa, fresas, hielo, lo, helado de fresa, una fresa, tomates, estas cosas, esta, la, esa cosa roja, una red, esta sopa, la sopa, el conto de helado, el gilo de helado, rojos, comida, helado; pintura roja: pintura roja, pintura, pintura rosa, pintura que se parecía al de fresas, roja pintura, pinta, roja pinta, paint, pentura, la color . . . el besa, el paint red . . . lojo, tinta roja, la tinta, pura pintura, rojo pintura, esa pintura, soltura, centura, el centura, rojo pinte, red ... r o j o . . . pinte; pan fresco:

pan, pan fresco, fresco pan, una martillo, una pan, un pan, un poco de pan, un pan fresca;

palmeras rosadas: planta, palmas rosadas, las flores, plantas, esta, mata rojas . . . matas rojas, una planta, palomas rojas, unas plantas, una plantas, una planta de palmas, palmeras rojos, unas flores, macetas, caña, una planta con flores, la planta, flores, floras, árbol; trompeta de plata: trompeta, trompeta de plata, una horn, un pito, un trompe, un trometro . . . tometro, el trombón, trátano, pito, trupa, un músico, un trumpet, un trompeta, un trompa, una pito;

One strategy which the children utilize in making lexical substitutions is that of substituting a lexical item with a superordinate. A superordinate is a name for a more general class of referents, such as vehicle for car or cut for pare. In retelling the story, the children employ the following superordinates:

104

Elizabeth A. Leone and René Cisneros

pito plantas árbol

for trompeta for palmeras for palmeras

elefante tinta tomates

fox gigante for pintura for fresa

Some of the other collocations or substitutes are lexical items which stand to each other in some recognizable lexical-semantic relation, or word meaning: part to whole, part to part, the same ordered series, etc. Some examples of these other substitutes are: mujer maceta fresa flores la color caña

for giganta for palmeras for helado de fresa for palmeras for pintura for palmeras

elefante tinta tomates cono sopa rojos

for for for for for for

gigante pintura helado de fresas helado helado de fresas helado de fresas

The wide variety of lexical collocations which appear in the narrative could point to a variety of influencing factors. One possible influence on the variety of substitutes could be that the lexical items which appear in the story are not well established in the lexicon of the child. Thus, when the child hears a lexical item and also sees the picture of the item in the storybook, the child substitutes the lexical item with another which bears a lexico-semantic relationship to the one in the story. So, if the story is about palmeras, then the child might retell the story with plantas. It would be interesting to compare the type of vocabulary which appear when the story is told without pictures to the type of vocabulary which appear when the story is told with pictures. Would the one without pictures produce less substitutions than the one with pictures? It is obvious that the lexical collocations which are related semantically reveal that the child has a lexical item which in his/her grammar expresses the object or concept. This is especially true of the superordinate and other lexico-semantic relationships. Another type of substitution is the one which we informally call the 'made-up' word. When the children retell the story they may utter a word which is quasi-homophonous and which has undergone phonetic changes from the original in the story. All the phonetic changes, it must be noted, are articulations which are phonologjcally correct for the Spanish language. For example, the following substitutions with phonetic alternates were found in the children's stories: trompa pienta pinte miada

for trompeta for pintura for pintura for mediano

giranta pantia pinta

for giganta for pintura for pintura

Children's Storytelling Strategies

105

A group of substitutes that are also 'made-up' words for the lexical items that they substitute yet are nevertheless actual words in the Spanish lexicon are the following: palomas

for palmeras

garganta

for giganta.

These words resemble phonetically the words that they substitute but they bear no semantic relation to the words they substitute. The last type of lexical substitution is the one where the child has heard the story told with one social or regional variant, and upon retelling the story accompanied by the same pictures seen when listening to the story he/she substitutes the regional or social variant from his/her own lexicon. One lexical item which produced a number of examples of this type of substitution was sus amigos. The friends are the giant's friends and in the story are depicted as smaller in size than the giant — in the pictures. The original story also refers to the size of the friends as the big one, the medium-sized-one, and the little one. Examples of the substitutions for amigos, el grande, el mediano, and el pequeño can be found in Table 2. Many of these seem to be social or regional variants. Another variant substitution which is most probably regional is nieve substituted for helado. What is interesting here is that the child obviously was able to decode helado and then substitute it for nieve, since the picture in the book does not graphically represent ice-cream. That is, without first decoding the word helado the child most probably would not have been able to then substitute nieve. If this is so, then the example would demonstrate that this child actually encodes and decodes two regional varieties for the term ice-cream in Spanish.

Communicative Strategies In general the children kept the language codes separate functionally. That is, they clearly used Spanish to narrate the story. English was used usually with a clear function related to the communicative competence of the social interaction appropriate to the speech event of narrating or retelling the story they had heard. That is, the administrator of the test specifically asked the students to retell the story in Spanish. Also, the earlier, preceding sections of the test — the phonetic, lexical and listening comprehension — had also been conducted in Spanish, including instructions in each of the preceding sections. Thus, the children were well informed and primed to carry out the task in Spanish. Hence one can say that the children understood the statements and instructions given which were of a metalinguistic nature. Interestingly, therefore, were the various communicative strategies that the children used which involved either switching languages or comments on the use of

106

Elizabeth A. Leone and René Cisneros

languages. In addition, suprasegmentals indicated a concern and preoccupation with being understood by the recipient of the narrative, the administrator-teacher listening to the narrative.

Metalinguistic

Comments

One strategy that supports the notion that the children comprehended the metalinguistic instructions given by the administrator is illustrated by the following examples: (4.1) "I don't know how to say it in Spanish". (#18) (4.2) "When it's too fast, I can't understand it". (#10) Statements such as the above would appear in the context where a child is ploughing along in Spanish and then comes to a phrase, concept, or lexical item which he could not express in Spanish. These metalinguistic comments made by the children were often made before the test even began, or between the various sections of the test, though they were also made in the middle of the narrative itself. The metalinguistic comments were communicative in the sense that they were concerned with conveying to the listener the language abilities of the storyteller-test-taker, so that, much as a preface to a book serves as a disclaimer, so too these comments served to pave the way for what was often a less than perfect fluency of the storyteller. Without these metalinguistic comments, the listener would not know if the children were themselves aware of their abilities, and the fact that their abilities in Spanish were not what their abilities in English were and/or not what the children themselves wished they were. These metalinguistic comments, far from indicating a lack of ability actually indicated a sensitivity of the children both for the adequacy of their own language and for the demands and expectations of the test and test administrator.

Restart-Switches Another strategy used by the children which combines a language switch with a restart of an utterance is illustrated by the following examples: (4.3) " . . . one day, he . . . un dia . . . " (4.4) " . . . there w a s . . . habia una vez . . . " The above switch is more than just a restatement. It is not a simple redundant use, nor a simple restatement for emphasis; much less is it an example of the student's inability to express the adverbial in Spanish, since the Spanish

Children's Storytelling Strategies

107

phrase follows the English switch. The switch from English to Spanish is definitely a restart. It occurs at the beginning of a topic, at the beginning of the narrative, and it repeats the message in the two languages (Ochs-Keenan, 1977). The restart with a change in language is part of the self-monitoring the children manifest in carrying out the task in Spanish. The student has two sets of linguistic resources for expressing the concept and for marking the story retelling. She starts using one set of linguistic resources, monitors herself, and swiftly shifts to the other set of linguistic resources, which are the ones which are appropriate for the carrying out of the task according to the instructions given at the beginning of the activity, specifying that the story be retold in Spanish. For these nine- and ten-year-olds, the different functions and different sets of linguistic resources and appropriateness for the use of these resources are all distinct in their understanding the metalinguistic instructions. The child might make a false start in one set of inappropriate linguistic resources, but thanks to her skillful communicative abilities and self-monitoring for appropriateness of style and other aspects of the communication, she smoothly shifts to the appropriate set of resources for the specific story retelling activity (Shimanoff and Brunak, 1977).

Verification-Switches Another type of language-switch communicative strategy is the verification or 'do-you-read-me' language shift. Here the child is smoothly sailing along in Spanish as the task indicated, and then at a certain spot, s/he will restate a point in English. That is, the restatement is made after the point has already been made in Spanish. This restatement switch is not so much a restatement for emphasis, but one which is part of the clarification, elaboration, and 'are you with me' type of clarification, as illustrated below: (4.5) " . . . la pintura . . . p a i n t . . . " This may also be a strategy of further clarification and elaboration; that is, the child may be 'covering both bases' as s/he sees them. She uses pintura, then switches to paint just in case the communication was not decoded, or if there were a problem in the decoding of pintura, or if there should be an error with pintura. By using the verification-switch, the child uses both sets of linguistic resources, to verify and confirm the referent, and the encodingdecoding process in general. The suprasegmentals of the verification switch are marked, as the lexical item usually comes after a slight pause and is stressed and is an even medial — not rising nor falling — intonation.

108

Elizabeth A. Leone and René Cisneros

Stylistic-Switches Another type of language switch which serves as a communication strategy is the stylistic-switch, illustrated below: (4.6) " . . . se felio sick . . . se estaba malo . . . " The above informal stylistic feature consists of a loanword plus a codeswitch, then a restatement in the formal standard variety of Spanish. In other words, se felio sick comes from one stylistic register while se estaba malo is a linguistic resource from a more general or formal variety. Together the two stylistic varieties comprise a type of repetition that serves as a communication strategy. The above example of a stylistic switch occurred after a clarification request (CR). initiated by the test administrator (Corsaro, 1977). That is, the child said se felio sick, then the test administrator said ¿quel, then the child said sick, then recomenced with se estaba malo. Although the second formulation has a problem with morpho-syntax, nevertheless it does represent a linguistic resource from a more general or standard variety.

Suprasegmentals Finally, a communicative strategy that perhaps encompasses many of the other types of strategies, since it includes features that are not segmented so easily, is the use of suprasegmentals, specifically pauses and intonation. The strategy marked by both intonation and pauses and that was most common in the children's narratives was the request for confirmation, illustrated below: (4.7) "¿el muchacho. ? . . " This example was accompanied by a particular lexical item plus a rising intonation, indicating that the speaker wanted some immediate response from the test administrator that the particular item was correct or appropriate. Although this particular strategy may also be common in English, and thus not necessarily indicate a language-specific strategy, nevertheless, its use was in all instances appropriate and concerned with adequate communicativeness. Whether, more particularly, the concern was with choice of style, morphosyntactic form, or some other level of language use, the use of the request for confirmation — which could potentially accompany almost any utterance — was functional for the children. Another use of pauses as a communicative strategy was simply the slowing down of speech — sometimes accompanied by a more careful enunciation of specific lexical items and a more marked juncture between items. These

Children's Storytelling Strategies

109

pauses often were clearly for the assistance of recall of a particular story fact or story referential item; other times the pauses were for a morphosyntactic form. At times, the pausing was only occasional in a particular narrative, though, especially in the narratives elicited with a question-answer format, the pauses were more hesitations that occurred frequently and even changed to silences when the lexical item was not recalled or the morphosyntactic form did not surface.

Summary The strategies used by the children — narrative strategies, coherence strategies, and communicative strategies — all reveal a command of the sociolinguistic system that, although limited, indicates a very active and viable communicative resource for these children. Far from being adult in their completeness or competency, these children's strategies are still developing, and still show signs of morpho-syntactic uncertainties. Nevertheless, these strategies, including verb usage, use of conjunctions, and language switches all demonstrate a native and a bilingual competency - not necessarily a fluency - that can and will continue to develop, given the recognition and appreciation that any skill requires for its development.

Implications for Language Planning When planning for Spanish language policy in the United States it is imperative that the many variants, some of which are reflected in this study, are considered. The first step in applying descriptive studies to language planning areas is to inform persons in positions as language planners about the results of these studies. For example, if the concept of communicative competence of bilinguals could be explained and illustrated at the appropriate stage of planning in a municipality's revision of personnel guidelines, then every time a question about language competence came up, in whatever context, at least there would be some written reference to the question. The example of a revision of city personnel procedures is only a hypothetical situation and the issue of language competence could just as logically fit into any number of situations where institutional procedures are being changed or modified. One of the most controversial and misunderstood features of bilingual communication is that of code-switching. Garcia points out that evaluations of bilingual children's language proficiency are made without regard to how the two languages work together as one system or resource (Garcia, 1980). Evidence that code-switching is still largely misunderstood is the acceptance

110

Elizabeth A. Leone and René Cisneros

of the notion of 'interference'. This notion is held despite the fact that bilingualism is more the norm worldwide than is monolingualism (Fishman, 1976). Therefore, when in the process of informing persons in health, legal, educational, and other institutions of the nature of Spanish language usage in the U.S., the concept of interference — the idea that knowing a second language will automatically interfere with knowing another language - must be put in its place. This concept of interference has been a misconception not only among laymen but also among educators and linguists. To a great extent, this misconception stems from the fact that studies of bilingualism do not reflect the native bilingual's theory of speaking but rather reflect the analyst's theory of speaking, rendering many of these studies inaccurate and invalid (Gumperz and Hernàndez-Chàvez, 1970). That is, these studies of interference are based upon linguistic data that do not take into account any of the cognitive, social and other contextual features of the bilingual's language use in ongoing communication. In order to effectively dispel this widely accepted yet erroneous theory of interference, researchers of bilingualism and persons concerned with Spanish language planning in the U.S. must communicate the underlying problems of this research to their colleagues engaged in such research. In addition, these same problems must be explained in more general terms, using if necessary examples from style-switching studies in English, to laymen in positions to affect Spanish language policy in their respective institutions and agencies. Other ways of using studies such as this one describing Spanish language variation include the incorporation of contextual examples of Spanish language variants into all types of Spanish language courses, at both the secondary and post-secondary levels. In order to do this successfully, the teacher must have a clear idea of some basic notions involved in language planning and language variation. Classes of Spanish for Spanish-speakers; Spanish for special groups such as social workers, nurses, interpreters, and teachers; and Spanish as a second language for college and high school students all could easily and enjoyably have students involved in going out to Spanish-speaking communities nearby to be first-hand collectors of Spanish variants in use in the U.S. (Leone and Cisneros, 1981). Once a large number of persons become informed as to the fact of Spanish language variation in the U.S., including language planners themselves, then the subsequent step of using this information about language variation, bilingualism, and language communities can begin to have an effect on language planning. A final suggestion for ways in which information from descriptive studies of Spanish language variation in the U.S. can be applied to language planning domains involves the investigation into language policy as it now exists in the various states, counties, cities, and federal areas of both public and private use. That is, references to language, or the absence of references to language

Children's

Storytelling

Strategies

111

in documents and other written communications such as memoranda, référendums, and legal precedents need to be researched. Once the existing language policy - whether it be an intentional language policy or simply policy by virtue of the fact that no references are available within a specific jurisdiction or institution — has been determined, then modifications based on demographic changes, resettlement patterns, and other needs in a given community can be made. These modifications and changes of existing language policies would involve the specification of variants that are used within the community as well as the goals and objectives of the community for use of Spanish in public domains. The planning and changing of policy would necessarily involve persons informed in the areas of bilingualism, language variation, and language communities — persons who would be sensitive to the needs of the communities being affected by the Spanish language planning.

References Cisneros, Rene and Leone, Elizabeth A. (1983). Mexican American Language Communities in the Twin Cities: An Example of Contact and Recontact. In L. Elias-Olivares (ed.) Spanish in the US Setting: Beyond the Southwest. National Clearinghouse for Bilingual Education, Rosslyn, VA. Corsaro, William. (1977). The Clarification Request as a Feature of Adult Interactive Styles with Young Children. Language and Society 6: 183—207. De Avila, Edward and Duncan, Sharon. (1977). Language Assessment Scales (LAS), Level I. Corte Madera, CA.: Linguametrics Group. Dubois, Betty Lou and Valdés, Guadalupe. (1980). Mexican-American Child Bilingualism: Double Deficit? The Bilingual Review 7(1): 1—7. Ervin-Tripp, Susan and Mitchell-Kernan, Claudia. (1977). Introduction. In Child Discourse, Ervin-Tripp and Mitchell-Kernan, (eds.), New York: Academic Press. Fishman, Joshua A. (1976). Bilingual Education: An International Sociological Perspective. Rowley, MA.: Newbury House. Garcia, Maryellen. (1980). Linguistic Proficiency: How Bilingual Discourse Can Show that a Child Has It. Ethnoperspectives in Bilingual Education Research, Vol. II: Theory in Bilingual Education. Ypsilanti, MI: Eastern Michigan University Bilingual-Bicultural Program. Gumpeiz, John and Hernández-Chávez, Eduardo. (1970). Cognitive Aspects of Bilingual Communication. In El Lenguaje de Los Chícanos. Washington, D.C.: Center for Applied Linguistics. Halliday, M.A.K. and Hasan, Ruqaiya. (1976). Cohesion in English. London: Longman. Hymes, Dell. (1964). Introduction: Toward Ethnographies of Communica-

112

Elizabeth A. Leone and René Cisneros

tion. In The Ethnography of Communication, Gumperz and Hymes, (eds.) American Anthropologist, 66(6), Pt. II, 1—34. — (1972). Models of the Interaction of Language and Social Life. In Directions in Sociolinguistics, Gumperz and Hymes, (eds.) New York: Holt, Rinehart and Winston. Keenan, Elinor Ochs. (1977). Why Look at Unplanned and Planned Discourse. Discourse Across Time and Space. Keenan and Bennet, (eds.) SCOPIL #5. Los Angeles: University of Southern California. Kernan, Keith. (1977). Semantic and Expressive Elaboration in Children's Narratives. In Child Discourse, Ervin-Tripp and Mitchell-Kernan, (eds.). New York: Academic Press. Labov, William. (1972). The Transformation of Experience in Narrative Syntax. In Language in the Inner City. Philadelphia: University of Pennsylvania Press. Labov, William and Waletzky, Joshua. (1967). Narrative Analysis. In Essays on the Verbal and Visual Arts, June Helm, (ed.). Seattle: University of Washington Press. Leone, Elizabeth and Cisneros, René. (1981). Modelos de Comunicación: Usando el Español en los Estados Unidos. Minneapolis: Luciérnaga Publications. Ochs-Keenan, Elinor. (1977). Making It Last: Repetition in Children's Discourse. In Child Discourse, Ervin-Tripp and Mitchell-Kernan, (eds). New York: Academic Press. Ochs, Elinor and Schieffelin, Bambi. (1979). Developmental Pragmatics. New York: Academic Press. Schegloff, Emanuel. (1973). Recycled turn beginnings: A precise repair mechanism in conversations' turn-taking organization. Public Lecture at Summer Institute of Linguistics, University of Michigan. —, Jefferson, Gail and Sacks, Harvey. (1976). The preference for self-correction in the organization of repair in conversation. Unpublished MS. Selinker, Larry. (1972). Interlanguage. International Review of Applied Linguistics, 10(3): 219-231. Shimanoff, Susan and Brunak, Joanna. (1977). Repairs in Planned and Unplanned Discourse. Discourse Across Time and Space, eds. Keenan and Bennett. SCOPIL #5, Los Angeles: University of Southern California. Stockwell, Robert, Bowen, Donald and Martin, John. (1965). The Grammatical Structures of English and Spanish. Chicago: University of Chicago Press. Tarone, Elaine. (1980). Communication strategies, foreign talk and repair in interlanguage. Paper presented at TESOL 1980, San Francisco. Walcutt. Judith. (1977). The Typology of Narrative Boundedness. In Discourse Across Time and Space, SCOPIL #5, Los Angeles: University of Southern California.

In the Match between Spanish Dialects, Who is the Referee?1 by Maryellen Garcia and Georgeanne Weller

People who determine the variety of Spanish for teaching at the university, high school, and elementary school levels look to a general Latin American standard or Castilian standard. Varieties of Spanish spoken closer to home are tacitly discounted by teachers in the United States despite the fact that U.S. speakers are more likely to use Spanish with Mexicans or Mexican-Americans rather than with Europeans or Latin Americans. As the nearest Spanish-speaking country is Mexico, and speakers of Mexican dialects of Spanish are found across the Southwest and the Midwest, a popular spoken Mexican Spanish may be more useful for communication in the Spanish-speaking neighborhoods of these locales rather than a formal variety. But little is known about how the spoken language of Mexicans differs from the norma culta, or educated norm, prescribed by texts and more appropriately suited for speaking in formal circumstances and in formal writing. Moreover, the dialect of Spanish spoken by Mexican-Americans has developed in a country where English has been the language of formal domains (e.g., school, government, national media, etc.), and is therefore likely to be different from other dialects of Mexican Spanish. The questions raised by this are: (1) what are the differences between the Mexican dialects; and, (2) which features of Southwest Spanish are likely to be radically different from other Mexican dialects, possibly leading to a negative evaluation from Mexican speakers? The present study was undertaken to survey the conscious awareness of groups of Mexican Spanish speakers as to what was acceptable, correct usage with regard to a number of Spanish constructions, taken from actual interviews with Mexicans and Mexican-Americans. The constructions which will be presented for discussion may indicate areas where linguistic norms may differ for the groups surveyed. While studies based on naturalistic, spoken language of each variety serve to point out the actual difference between Mexican and Mexican-American Spanish (see Teschner, Bills, and Craddock (1975) for a comprehensive, annotated bibliography of studies on Mexican-American Spanish), acceptability

114

Maryellen Garcia and Georgeanne Weller

studies have served to point out those linguistic areas where dialects may prove to differ in the spoken language as well as to suggest conscious awareness of the linguistic differences, i.e., norms, by members of a particular dialect group or speech community. Awareness of norms of correct usage have been explored by previous acceptability studies. Garcia (1979) found that a was preferred to para by Mexicans and Mexican-Americans in expressions indicating a locative goal, but that para was more acceptable with a greater variety of verbs for Mexican-Americans. Garcia (1977) suggested that para, the full nonreduced form, was a conscious linguistic norm for Mexican-American Spanish speakers that is shared with other speakers of New World Spanish, but that certain constructions with locative para, such as irse para atras, may be an acceptable norm in U.S. Spanish only. Constraints on the use of the subjunctive have been compared between Mexicans and Mexican-Americans (Garcia and Terrell, 1978) and between Puerto Ricans and Mexican-Americans (Lantolf, 1980). The present study will attempt to discover other areas in which norms of correctness differ between Mexicans and Mexican-Americans.

Methodology It was difficult to determine an exercise which could be administered fairly quickly, could evoke judgements about the spoken language, and also allow respondents to identify what, specifically, about the construction they felt to be unacceptable. Considered and rejected were exercises which would elicit attitudes about the speaker, as in matched guise studies (cf. Anisfield, Bogo and Lambert, 1962) rather than the sentences themselves. Also rejected was the idea of placing utterances in extended discourse contexts giving the social status of the speakers and the circumstances of the utterance. Our goal was not to determine whether or not a particular utterance was suited to formal or informal circumstances, but to get at an underlying awareness of what was acceptable Spanish and unacceptable Spanish from different groups. With the goals of eliciting both immediate reactions to the spoken language and identification of any unusual specific features of the utterance, we presented twenty-five sentences for judgement in two modes: aural, with tape-recorded, spoken sentences as the stimulus, and written; with the same sentences on a typed response sheet. The sentences were taped by a native Spanish speaker with standard pronunciation. 2 Except for certain sentences in which rural pronunciations were specified, sentences were intended to be evaluated on nonphonological criteria. Responses to the aural stimulus was on a sheet where checks could be placed on a five stage continuum from culto 'educated' to mayormente inaceptable 'extremely unacceptable'. Between these two extremes there were correcto 'correct', popular aceptable 'popular

In the Match Between Spanish Dialects, Who is the Referee?

115

acceptable', and popular inaceptable 'popular and unacceptable'. For the evaluation of the written sentences, which were distributed after the aural component of the task had been completed, subjects were asked to underline and change anything that they thought was incorrect or odd to something closer to the way they would express the same thing. One unexpected outcome of the task was that respondents did not understand some of the sentences as written, and changed those words and expressions which did not make sense. Additionally, the inappropriateness of finding a colloquial expression written down likely motivated a change to a more formal stylistic variant. These problems and their influence on the interpretation of the results are discussed further in the concluding section of this paper. Sentences. The majority of the sentences were based on spoken Spanish data collected in the El Paso-Juárez area (García, 1981). (See Table 1 for original sentence and modifications.) As these data were elicited in interactive conversation, and therefore might not be understood in the absence of their discourse context, they were modified where necessary to make sense on their own. Further, the sentence might be modified to eliminate any extraneous material, including nonstandard features which were not of immediate interest in the study. The sentences thus selected and modified for inclusion in the exercise were intended to represent the range from very educated speech to very colloquial, uneducated speech. Some were of uncertain grammatical status, but had been noted in the interviews conducted by the first author, and were included to see how their grammatically would be assessed by the groups of subjects. The sentences took 10—12 minutes to present on the cassette audiotape and subjects took approximately another fifteen minutes to evaluate the written versions. Subjects. The Mexican subjects for the study were surveyed in Mexico City, and represent both the professional and working classes, determined by level of education and occupation. There were 39 professionals and 37 in the working class. The subjects were a convenience sample of friends and colleagues of the second author, augmented by people recruited at their jobs for the administration of the exercise. All subjects were over 18 years of age and had resided in the Mexico City metropolitan area for at least 10 years. The exercise was administered in a number of settings — in offices, in kitchens, construction sites, and whatever other circumstances were convenient at the time. The Mexican-American subjects were students at a state college between the ages of 18 and 35, and were surveyed in their Spanish classes at the beginning of the semester. 3 They had all learned Spanish as a home language

116

Mary eilen García and Georgeanne

Weller

and were taking the classes to improve their literacy skills. They were also fluent in English, having been raised in the Los Angeles area. Most had fullor part-time j o b s during the day; over half were elementary school teachers or teacher aides. There were 38 in the Mexican-American group.

Table 1: Sources o f sentences. Sentence and source (underneath)

Speaker

1. En ese lugar suelen comer a las cuatro de la tarde. (Fabricated) 2. En México cada profesionista hace su servicio social. "En Méjico toda profesionista hace servicio social."

MX(mc)

3. Hasta la fecha me gusta mucho Chihuahua. "Hasta la fecha me gusta mucho Chihuahua."

MX(mc)

4. Durante ese año, Juan estaba radicando en Chihuahua. "Estaba radicando en Chihuahua, sí."

MX(wc)

5. Anduve buscándolo pa' arriba y pa' abajo todo el día. (Fabricated) 6. Y ay bajando del puente luego luego los agarraron presos. "Y ay bajando del puente luego luego los agarraron presos."

MX(wc)

7. Todos son nacidos en Tepic. "Todos son nacidos ay en Juárez, pero. . . "

MX(wc)

8. Creo que ella fue nacida allá en Yucatán. "Es nacida en México."

MA(mc)

9. La semana pasada fuimos paraGuadalajara, y nos divirtimos mucho allí. (Fabricated: used in García, 1981: 164) 10. Uno de mis hijos vive para El Paso, aunque no en la mera ciudad. "Y uno de los hijos mayores vive en Sunland Park . . . el otro vive para Ascárate."

MX(wc)

11. Después de quedarme unos tres días en Los Angeles, me regreso para atrás a mi casa. "Entonces me regreso pa' atrás para Los Angeles y me quedo unos dos, tres días ay, y regreso para atrás."

MA(wc)

12. Si se me había olvidado la cartera, fuera para atrás a la casa por ella. "Pos, yo fuera para atrás a mi casa a agarrar mi cartera."

MA(wc)

13. Ese mismo año se alistó y se fue pa'l ejército. "Se alistó y se fue pa'l Air Forcé."

MA(wc)

Key: MX = Mexican, MA = Mexican-American, mc = middle class, wc = working class.

In the Match Between Spanish Dialects, Who is the Referee?

117

Results While a comparative analysis of all sentences in all modalities is unfeasible here, we feel that a comparison of Mexican and Mexican-American evaluations of the written responses will be the most revealing of the perceptions of correctness for each group. The written versions of the sentences were intended to be visual prompts to aid the respondent in pinpointing those items that were incorrect about the sentence, but the fact that they were written called into play another set of considerations: those of the respondent's literacy skills in Spanish and his/her awareness of stylistic variance between the spoken and written language. This had also been noted in a previous study (Garcia, 1977), where respondents were asked to change items and passages they felt to be 'incorrect' on a partial typescript of an interview. Still, it had yielded qualitative information as to the reasons that certain linguistic constructions used in the interviewee's spoken Spanish might be unacceptable to speakers of other dialects. This was not possible with acceptability studies with auditory stimulus alone, done with Chicanos on a/para/pa alternation (Garcia, 1979), the subjunctive (Garcia and Terrell, 1978), and with Puerto Ricans on the subjunctive (Lantolf, 1980). The sentences to be discussed must therefore be considered in terms of the extent to which they reflect colloquial usage and may be inappropriate in a written style, and then the degree to which the respondent felt qualified to change it to a more formal written style. Of the 25 sentences, fifteen of the more stereotypically 'bad' and 'good' sentences were selected for analysis, grouped into the following categories: superstandard, colloquial, and locative para. Each category will be explained in its respective section. The results discussed in the remaining sections are based on respondents' reactions to the written questionnaire. Two procedures were used in the analysis. For the Mexican data, responses from both the middle and working classes were considered by the second author and two of her colleagues in Mexico City, all with graduate preparation in language and linguistics. From the written questionnaires they reached a consensus as to a sentence which would typify the correct version of the sentence and a popular but acceptable version for the Mexican respondents. These are presented as the (b) and (c) variants, respectively, of the sentences presented for discussion. For the Mexican-American data, we focused on those items in each sentence which were likely to be the cause of the evaluation. The first author tabulated the frequency that each item was changed or eliminated by that group. Those results are contrasted with the (b) and (c) versions of the sentences from the Mexican sample.

118

Maryellen García and Georgeanne Weller

Superstandard sentences. The sentences in this section were intended as superstandard on the basis of the lexical items or phrases underlined. It was our belief that norms for correctness would be revealed as much by judgements of sentences which were culto, i.e., pure, elegant, correct (Cuyás, 1966: 157) as they would by sentences which were vulgar, i.e., vernacular (speech) (Cuyás, 1966: 570). The items and phrases underlined in the sentences below had been used in an interview with the first author by a collegeeducated woman from the Juárez middle class. At the time of the interview, these items had seemed to the interviewer not to be common in the Spanish of Chicanos and immigrant Mexicans in the United States. The first sentence to be discussed contains a form of the verb, soler 'to be in the habit of, accustomed to'(Cuyás, 1966: 508). This dictionary indicates that it is a defective verb, in that it is used only in the present and imperfect tenses and is always followed by an infinitive (1966: 508). Bello and Cuervo (1973: 214) agree, adding that the preterite tense and verbal derivations are hardly used, and that the other forms are not used at all. The sentence from the questionnaire is example (la); the corrected version ( l b ) is more correct for the Mexican group sampled than (lc), which is more popular for the same group. (1) a. En ese lugar suelen comer a las cuatro de la tarde. 'In that place they customarily eat at four in the afternoon'. b. En ese lugar acostumbran comer a las cuatro de la tarde. 'In that place they are used to eating at four in the afternoon'. c. En ese lugar siempre comen a las cuatro de la tarde. 'In that place they always eat at four in the afternoon'. From this initial example, it is obvious that respondents change the sentence not strictly on the basis of semantic equivalence but modify meaning as well if it conveys the same general idea. It may be argued that acostumbran comer 'they are accustomed to eating', and siempre comen 'they always eat' are different ways of saying essentially the same thing for the Mexicans sampled, one slightly more elegant than the other. The Mexican-Americans sampled, on the other hand, had a problem understanding the sentence, and asked the examiner what it meant. Most of this group (58 percent) left the sentence as is on the response sheet, presumably at a loss as to what to change. The rest of the group changed or eliminated the auxiliary. The second sentence contained the word profesionista, which in context meant 'professional person': (2) a. En México cada profesionista hace su servicio social. 'In Mexico, every professional person does his social service'.

In the Match Between Spanish Dialects, Who is the Referee?

119

b. En México todos los profesionistas tienen que hacer servicio social. 'In Mexico all professionals must perform social service'. c. Todos los profesionistas mexicanos hacen su servicio social. 'All Mexican professionals do their social service'. This stimulus sentence included was assumed superstandard, i.e., culto, because of the level of education of the original interviewee and because the speech in her interview was largely free of nonstandard dialectal variants, e.g., para atrás, íbanos, -ites, etc. Neither Moliner (1966) nor Cuyás listed profesionista, but Cuyás did have profesional with that meaning (1966: 443). From examples (2b) and (2c) we can see that the word as written was accepted by Mexicans, with only stylistic variance supplied by their word order changes. Over three-fourths (89 percent) of our Mexican-American sample, on the other hand, did not change this item, the rest changed it to profesional. In sentence 3, hasta la fecha 'to this day' was expected to be judged as superstandard because it was not familiar to us in everyday conversation: (3) a. Hasta la fecha me gusta mucho Chihuahua. 'To this day, I like Chihuahua very much'. b. Siempre me ha gustado la Ciudad de Chihuahua. 'I've always liked the city of Chihuahua'. c. Me sigue gustando mucho Chihuahua. 'I've continued to like Chihuahua very much'. As we can see from the type of change typically made by the Mexican group sampled, the meaning of the target phrase was retained, but the form changed by writing in another adverbial modifier, or by changing the tense of the verb. Possibly hasta la fecha sounded stilted, and may be inferred not to be part of the usual way of speaking of the Mexico City respondents. Over three-fourths (87 percent) of the Mexican-American sample didn't change it, however. The remaining few made changes such as hoy en día and siempre, indicating that for them as well, hasta la fecha is not part of their usual speech. Colloquial expressions. Another set of sentences contained slang lexicon, folksy expressions, or rural pronunciation (indicated by apostrophes in the written questionnaire). (4) a. Durante ese año, Juan estaba radicando en Chihuahua. 'During that year, Juan was living in Chihuahua'. b. Durante ese año Juan vivió en Chihuahua. 'During that year Juan lived in Chihuahua'.

120

Mary eilen García and Georgeanne Weller c. Durante ese año Juan radicaba en Chihuahua. 'During that year Juan lived in Chihuahua'.

We can see from the changed variant (4b) that vivir 'to live' is considered more formal than radicar, which has as a first meaning 'to take root', and 'to settle, establish oneself as the second (Cuyás, 1966: 455). This verb was not listed at all in several sources reporting colloquial usage (Santamaría, 1942, 1974; Galván and Teschner, 1977; Vásquez and Vásquez, 1975; Brambila Pelayo, 1957; Gerrard and Heras, 1964). The Mexican-American sample for the most part seemed unfamiliar with the meaning of the verb. Several respondents asked the examiner about its meaning during the exercise. Almost three-fourths (74 percent) of the Mexican-Americans sampled did not change the sentence despite the fact that they did not seem familiar with it. Those who did recommend a change replaced the verb with some form of vivir 'to live'. In the next sentence, the progressive form of the verb was conjugated with andar rather than the more formal estar 'to be' prescribed by Spanish grammar texts (e.g., MLA, 1960: 126). Further, para is shortened to pa' in the colloquial expression pa' arriba y pa' abajo: (5) a. Anduve buscándolo pa' arriba y p a ' abajo todo el día. 'I was looking for him all over the place (literally, up and down) all day'. b. Lo busqué por todos lados durante todo el día. 'I looked for him all over (literally 'on all sides') all day'. c. Lo anduve buscando todo el día. 'I was looking for him all day'. Cuyás has as its primary meaning for andar: (for a person) 'to walk, to go'. We are advised that 'to be' is a fourth meaning, especially as an auxiliary verb with a gerund (Cuyás, 1966: 36). De Mello adds that andar as an auxiliary adds the idea of 'wandering about', or not going straight to a specific goal (1974: 209). The other underlined item in the sentence, pa' arriba y pa' abajo had been included in another questionnaire eliciting judgements in fieldwork in El Paso-Juárez (García, 1981: 73). The expression was found not to be consciously stigmatized by Spanish speakers in that locale. In a prior acceptability study both Mexicans and Mexican-Americans corrected pa to para on a portion of typescript of an interview with a Mexican-American woman. Awareness of the full form was suggested as a conscious linguistic norm which Mexican-Americans shared with other New World Spanish speakers (Garcia, 1977: 207).

In the Match Between Spanish Dialects, Who is the Referee?

121

It is evident from the changes made by the Mexican group sampled that andar was felt to be a colloquial element, and that eliminating it in (5b) made the sentence more 'correct'. Importantly, none of the Mexican-Americans changed the auxiliary. Further, the Mexican group altered pa' arriba y pa' abajo by either eliminating or replacing it. Of the Mexican-American group, only 45 percent indicated its elimination or replacement, while 26 percent indicated changing pa' to para. The rest made no changes to the sentences. The next sentence included phonological reduction, ay for ahi, lexical reduplication, luego luego, and semantic extension, i.e., agarrar 'to grasp, seize' (Cuyas, 1966: 17) to mean 'to capture'. (6) a. Y ay bajando el puente luego luego los agarraron presos. 'And there, getting off of the bridge, right away they took them prisoners'. b. A1 bajar del puente inmediatamente los tomaron presos. 'Upon descending from the bridge, they immediately took them prisoners'. c. Bajando del puente luego los tomaron presos. 'Descending from the bridge, they then took them prisoners'. Of ahi, Santamaria writes that it is used popularly in Tabasco and in Puerto Rico, often as an expletive (1974: 43). Furthermore, in the expression de ahi, used as a temporal adverbial, said to be equivalent to 'despues', the stress is usually on the lower vowel in the vernacular (Santamaria, 1974: 43). We have observed that in popular Mexican dialects, there appears to be a single locative adverb pronounced as the glide ay. The elimination of this adverb in the responses of the Mexican groups sampled indicates that it is recognized as nonstandard. The Mexican-American group sampled also reccommended changing or eliminating ay by 63 percent, indicating perhaps a shared norm with the Mexican groups. Also in this sentence was the expression, luego luego. In peninsular Spanish, luego by itself in colloquial usage is reported to mean 'later, in a moment, or afterward' (Gerrard and Heras, 1964: 70). In one tHctionary on Chicano Spanish, three forms are presented: luego, luego luego, and luego lueguito, all glossed as 'right away, immediately' (Galvan and Teschner, 1977: 51). The reduplicated form was used along with en seguida to provide the gloss for luego, lueguito in Santamaria's book on Mexicanisms (1974: 668). The response to this colloquialism, clearly widespread in Mexican usage, was to change the adverb to a more standard semantic equivalent, inmediatamente in the (b) version of the sentence, and to use the single word in the more

122

Mary eilen García and Georgeanne Weller

popular (c) version. Only 31 percent of the Mexican-American group suggested that the alteration of the adverb was necessary. Agarrar, the last underlined item, is extended to a greater variety of semantic contexts in Southwest Spanish (Sánchez, 1972: 49; Lance, 1975: 46; Garcia, 1977: 203) and is therefore more syntactically versatile as well. Santamaría reports that it is used 'promiscuously' (promiscuamente) for asir, tomar, coger, adoptar, recoger, pillar, adquirir, cobrar, contraer, etc., in Latin America (1942: 52). In his book on Mexicanisms Santamaría reports that it is accepted by the Spanish Academy for asir, coger, and tomar con la mano, verbs which indicate a physical component in the core meaning. He reflects his personal distaste for its popular usage by negatively sanctioning its use in a wider number of semantic contexts: ' . . . sigue siendo entre nosotros sumamente vulgar y no usado por gente de mediana preparación cultural' (1974: 35). In the responses from the Mexican group, we see that the majority of both the middle class and working class replaced the verb with tomar-, in contrast, only 18 percent of the Mexican-American group recommended a change of verb. The next two examples were included to focus on the ser + naca- construction. (7) a. Todos son nacidos en Tepic. 'All are born in Tepic'. b. Todos son nativos de Tepic. 'They are all natives of Tepic'. c. Todos nacieron en Tepic. 'There were all born in Tepic'. (8) a. Creo que ella fue nacida allá en Yucatán. 'I think that she was born in Yucatan'. b. Creo que ella nació en Yucatán. 'I think that she was bom in Yucatan'. c. Se me hace que ella nació en Yucatán. 'It seems to me that she was born in Yucatan'. This construction had been evaluated in an earlier study (Garcia, 1977) in an example taken from actual speech, fui nacida allá which in context meant, 'I was born there'. It was found to be extremely unacceptable (71 percent) to Latin Americans and somewhat less so for Mexicans (50 percent). For both of these groups the most frequent change suggested was to the simple preterite, nací. Only 17 percent of the Mexican-Americans surveyed in that study found that expression unacceptable. It was noted then that

In the Match Between Spanish Dialects, Who is the Referee?

123

while ser was used as an auxiliary in passive constructions, it was so used only with intransitive verbs (Kenniston, 1937:40) and only when the semantic direct object was to be made the focus of attention in the sentence by using it as the grammatical subject (Moliner, 1966, II: 55). The conclusion drawn was that fui nacida in that example quite likely reflected English language influence on selectional and functional restrictions on the Spanish language norms of Mexican-Americans (Garcia, 1977: 206). The Mexican respondents in the present study changed the passive to the preterite forms, seen in versions (b) and (c) of (7) and (8). Of the MexicanAmericans in our sample, 53 percent suggested a change of the passive in (7), and 26 percent suggested a change in (8). This would indicate that the Mexicans sampled have norms which don't allow the use of nacer in the passive construction, while Mexican-American norms for that construction are more tolerant. Pa(ra) + Location. The last sentences which we shall consider are those containing para to introduce a locative complement. We will consider each singly: (9) a. La semana pasada fuimos para Guadalajara, y nos divirtimos mucho allí. 'Last week we went toward Guadalajara and we enjoyed ourselves very much there'. b. La semana pasada fuimos a Guadalajara y nos divertimos mucho. 'Last week we went to Guadalajara and we enjoyed ourselves very much'. c. La semana pasada estuvimos en Guadalajara y nos divertimos mucho allí. 'Last week we were in Guadalajara and we enjoyed ourselves very much there'. Sentence (9) was intended to imply arrival at Guadalajara, making a 'to' a more logical choice than para 'toward'. Previous acceptability studies (Garcia, 1977, 1979) have shown para to be the most acceptable to both Mexicans and Mexican-Americans with the verb ir, and less so with other verbs of motion. But because para prescriptively focuses on the direction of the motion rather than the goal (Gili Gaya, 1964: 254; Seco, 1967: 257;López, 1970: 169), we would expect that the preterite form of the verb, plus the additional context that indicates that the goal was reached, especially alii, would make para semantically unacceptable in this case. In fact, the Mexican groups both eliminated para and allí in the more standard version, (9b), and changed the focus from motion to static location with en in the (9c) version. Of the Mexican-Americans sampled, 45 percent recommended a change of

124

Maryellen García and Georgeanne Weller

para to a, while 50 percent recommended no change at all. The remainder prescribed the elimination of allí, possibly to shift the focus away from arrival at the goal. The next sentence tested the acceptability of what has been called 'static locator' para, used to introduce locatives after non-motion verbs where prescriptive grammars would prescribe en (Garcia, 1979: 107). (10)a. Uno de mis hijos vive para El Paso, aunque no en la mera ciudad. 'One of my sons lives in the direction of El Paso, but not in the very city'. b. Uno de mis hijos vive cerca de El Paso, pero no en la ciudad. 'One of my children lives near El Paso, but not in the city'. c. Uno de mis hijos vive en las afueras de El Paso. 'One of my children lives on the outskirts of El Paso'. The use of static locator para had been observed in the colloquial Spanish of the El Paso border and so was included in the earlier acceptability study which showed it to. be not widely accepted as a norm (1979: 110). In the present study, the majority of the Mexicans also found it unacceptable. The (a) version of this sentence retains the intended meaning of the original, i.e., that the location was in the vicinity of El Paso but not the city proper. We note that both para and mera are eliminated from this version. During the exercise, the Mexican groups commented negatively on the use of both of these items in (10). The more colloquial paraphrase in version (10c) re-states the location as the outskirts of the city by replacing para with en and eliminating mera. The majority (61 percent) of the Mexican-Americans sampled also prescribed changing or eliminating para, but only 16 percent recommended the same for mera, indicating a shared norm with the Mexican group for the former, but no shared stigma for the latter. In (11) the redundant and highly stigmatized para atrás was expected to elicit extreme negative comment from the Mexican group. Their responses are indicated in (1 lb) and (1 lc): (11)a. Después de quedarme unos tres días en Los Angeles, me regreso para atrás a mi casa. 'After staying some three days in Los Angeles, I return back to my house'. b. Después de quedarme unos tres días en Los Angeles, tengo pensado regresar a mi casa. 'After staying in Los Angeles for three days, I'm thinking of returning home'.

In the Match Between Spanish Dialects, Who is the Referee?

125

c. Voy por tres días a Los Angeles y regreso de nuevo a casa. 'I'm going to Los Angeles for three days and then returning home again'. In addition to the use of regresar alone in these paraphrases, the Mexican groups sampled commented frequently that regresar para atrás should not be used. It was also negatively judged by almost three-fourths (74 percent) of the Mexican-American group, who suggested changing or eliminating para atrás. Interestingly, in a sentence where para atrás was used with a form of ir, the Mexican groups again rejected the adverbial phrase, while MexicanAmericans were more receptive. (The sentence also contains nonstandard tense sequence included intentionally, but which produced insufficient data for discussion at this time.) (12) a. Si se me había olvidado la cartera, fuera para atrás a la casa por ella. 'If I had forgotten my wallet, I would have gone back to the house'. b. Si se me hubiera olvidado la cartera me regresaría a la casa a recogerla. 'If I had forgotten my wallet, I would return to the house to retrieve it'. c. Si se me olvida la cartera regreso a casa por ella. 'If I forget my wallet I return to the house for it'. As we can see from the (b) and (c) versions, the preferred verb for the Mexican groups is regresar rather than the English based calque ir para atrás. The Mexican-American group was much less negative, but the majority, 58 percent, recommended changing or eliminating para atrás. The final example has para used as a complementizer for an abstract noun, ejército 'military service'. It has been proposed (Garcia, 1981: 195—6) that this use of para is modeled on its locative usage in introducing specific locative referents after verbs of motion. This interpretation is reinforced when we note the syntactic overlap of ir + para to introduce locative complements, and its semantic extension here to introduce a figurative move in terms of personal employment, which perhaps implies a change of physical location as well. (13)a. Ese mismo año se alistó y se fue pa'l ejército. T h a t same year (he) readied (himself) and went with (into) the service'. b. En ese año ingresó al ejército. 'In that year (he) joined the service'.

126

Mary eilen García and Georgeanne Weiler c. Ese mismo año se preparó y se fue con el ejército. 'That same year (he) readied (himself) and went with into the service'.

As we can see from (13b) and (13c) from the Mexican groups, para was not accepted in this semantic context. The Mexican-American group either changed pa' to para (39 percent) or changed pa' to a (38 percent), the remainder leaving the sentence as is. Further, alistar has two entries in the Cuyás dictionary, one meaning 'to enlist, enroll', and the other, 'to get or make ready' (1966: 26). Both alistar and enlistar are reported by Rael (1975: 22) as familiar vocabulary to New Mexicans, with the meaning 'to enlist' reinforced from English. The Mexican respondents paraphrased the verb in both the (b) and (c) versions, while only 18 percent of the Mexican-Americans sampled recommended a change of that item.

Concluding Discussion The constructions just discussed were evaluated on the basis of written responses to examples taken out of context from the spoken language. As such, the evaluation of the construction itself must be differentiated from changes that are made to shift the register from the vernacular to a written style, and those which add a few words of context to make the sentence read better in isolation. However, it seems reasonable to suggest that those constructions which are not altered in a written format are not in the conscious awareness of the respondents as unacceptable written variants. Where the Mexican sample and the Mexican-American sample appear to have different evaluations of the written language, it may be reasonable that those items are of difficult status in their respective speech communities. In Table 2 we summarize our interpretations of the responses made by both groups to the words, phrases, and constructions presented in the written versions of the exercise. The evaluations should be interpreted as each group's assessment of the item's acceptability in a written mode. In a previous article, Garcia (1977: 207) suggests that "Chícanos have their own norms of grammaticality for their dialect of Spanish, and thereby comprise a speech community of Spanish speakers". It is the question of speech communities which emerges as the one to be treated more definitively before conclusions can properly be made. The social dimension of the notion 'speech community' is inseparable from the linguistic dimension of the construct, whereby the speakers so identified must have at least one linguistic variety in common, and share norms for the negative or positive evaluation of either the conduct of speech (Hymes, 1972: 54), or have similar evaluations of a

In the Match Between

Spanish Dialects,

Who is the Referee?

127

Table 2: Comparisons of evaluations of constructions from written data. Sentence 2 4 1 13 6 10 3 6 5 5 7,8 9 10 11 12 5,13

Construction NOUNS profesionista VERBS radicar soler alistar ('enlist') MODIFIERS (Adj./adv.) ay (for ahí) mera hasta la fecha luego luego pa ' arriba y pa ' abajo VERB TENSES andar + progressive ser + nacer (passive) VERB OF MOTION +paraÍL Ir para vivir para regresar para atrás ir para atrás pa' (hipara)

Mexican sample

Mexican-American sample

OK

U

OK S N

U U OK

N N S N N

N OK U OK U '

N N

OK OK

N N N N N

OK N N OK N

Key: S = superstandard, OK = acceptable, N = not acceptable, U = unfamiliar or of uncertain status. a. For a more detailed discussion of these constructions based on interview data from 59 speakers, see Garcia (1981).

set of linguistic features of the variety (Labov, 1972: 158; Bailey, 1973: 65). In the Mexican-American situation, there are clearly some features which are not shared across dialects of Mexican Spanish, due to the group's separation from the Mexican monolingual Spanish mainstream, and its fluency in English. Where two dialects have the same features in common, and share the same social interpretation of their use, then it is reasonable to treat reactions to those features as being on a single continuum of relative stigmatization, as is the case of para for pa' and ahi for ay. Where there are some features that are not held in common across two groups of speakers, their evaluations cannot be compared, and the argument for their treatment as two different speech communities is strengthened. The recommendations for teaching which can be made on the basis of these results are that constructions or lexicon stigmatized by the Mexican group be avoided in written Spanish. Where Spanish is used for purposes of spoken communication in the United States, the speech community is not likely to stigmatize speakers who use nonstandard constructions. However,

128

Mary eilen García and Georgeanne Weller

for purposes of teaching reading and writing, which Spanish teachers usually include as goals of instruction, a more formal style is needed. In answer to the question posed by the title of this paper, we must conclude that there is no single referee, because there appear to be two arenas, but for teaching writing to students in the United States, it may be better to be in the more conservative corner.

Notes 1. We would like to thank Benji Wald, Elizabeth Leone, and René Cisneros for their comments on an earlier draft of this paper. We also thank Víctor Rodriguez for his editorial help. 2. We wish to thank Consuelo Sigiienza for taping the sentences. 3. We wish to thank Ms. Sara Del Oro Krauthamer, Dr. Alfredo Morales, and especially Dr. Hildebrando Villarreal of California State University at Los Angeles for lending us their students for the exercise. We also wish to thank friends and colleagues at CIIS in Mexico City, and all of the people in Mexico City and Los Angeles who took the time to participate in the exercise.

References Anisfield, M., Bogo, N., and Lambert, W.E. (1962). Evaluational Reactions to Accented English Speech. Journal of Abnormal and Social Psychology 65: 223—231. Bailey, C.J. (1973). Variation and Linguistic Theory. Washington, DC: Center for Applied Linguistics. Bello, A. and Cuervo, R.J. (1973). Gramática de la Lengua Castellana. Buenos Aires: Editorial Sopeña Argentina, S. A. Brambila Pelayo, A. (1957). El Lenguaje Popular en Jalisco; Exportación Lexicográfica. Guadalajara, Jalisco, México: Editorial Brambila. Cuyás, A. (ed.). (1966). Appleton's New Cuyás Spanish Dictionary: EnglishSpanish and Spanish-English. 5th Edition. New York: Appleton-CenturyCrofts. De Mello, G. (1974). Español Contemporàneo. New York: Harper and Row. Galván, R., and Teschner, R. (eds.). (1977). El diccionario del español chicano. The Dictionary of Chicano Spanish (Revised Edition). Silver Spring, MD: Institute of Modern Languages, Inc. García, M. (1977). Chicano Spanish/Latin American Spanish: Some differences in linguistic norms. The Bilingual Review 4(3), 2 0 0 - 2 0 7 . - (1979). Paira) usage in United States Spanish. Hispania 62(1), 106-114. — (1981). Pa(ra) in Locative Phrases in the Spanish of El Paso-Juárez. Ph.D. Dissertation. Washington, DC: Georgetown University.

In the Match Between Spanish Dialects, Who is the Referee?

129

— and Terrell, T. (1978). Is the use of mood in Spanish subject to variable constraints? In Studies in Romance Linguistics, Hagiwara, M.P., (ed.), 214-226. Rowley, MA: Newbury House. Gerrard, A. and Heras, J.H. (1964). Beyond the Dictionary in Spanish: A Handbook of Colloquial Usage. New York: Funk and Wagnalls. Gili Gaya, S. (1964). Curso Superior de Sintaxis Española, 2nd ed., Barcelona: Bibliograf, S.A. Hymes, D. (1972). Models of the Interaction of Language and Social Life. In Directions in Sociolinguistics, Gumperz, J. and Hymes, D. (eds.), 35-71. New York: Holt, Rinehart and Winston. Kenniston, H. (1937). The Syntax of Castillian Prose. Chicago: University of Chicago Press. Labov, W. (1972). Sociolinguistic Patterns. Philadelphia, PA: University of Pennsylvania Press. Lance, D. (1975). Dialectal and Nonstandard Forms in Texas Spanish. In El Lenguaje de los Chícanos, Hernández-Chávez, E., Cohen, A., and Beltramo, A. (eds.), 37-51. Washington, DC: Center for Applied Linguistics. Lantolf, J. (1980). The Variable Constraints on Mood in Puerto Rican-American Spanish. In Contemporary Studies in Romance Linguistics, Suñer, M. (ed.), 193—217. Washington, DC: Georgetown University Press. López, M.L. (1970). Problemas y Métodos en el Análisis de Preposiciones. Madrid: Editorial Gredos. Modern Language Association. (1960). Modern Spanish. New York: MLA Publications. Moliner, M. (1966). Diccionario del Uso del Español, II. Madrid: Editorial Gredos, S.A. Rael, J. (1975). Associative Interference in New Mexican Spanish. In El Lenguaje de Los Chícanos, op. cit. Sánchez, R. (1972). Nuestra Circunstancia Lingüistica. El Grito 6(1), 45—74. Santamaría, F.J. (1942). Diccionario General de Americanismos. México: Editorial Pedro Robledo. — (1974). Diccionario de Mexicanismos, 2a ed. México: Ed. Porrúa, S.A. Seco, M. (1967). Diccionario de Dudas y Dificultades de La Lengua Española. Madrid: Gráficas Une, S.A. Teschner, R., Bills, G., and Craddock, J. (1975). Spanish and English of United States Hispanos: A Critical, Annotated, Linguistic Bibliography. Washington, DC: Center for Applied Linguistics. Vásquez, L.K., and Vasquez, M.E. (1975). Regional Dictionary of Chicano Slang. Austin, TX: Jenkins Publishing Company.

Section 2 Using Spanish in Public Life

Spanish Language Planning in the United States by Joan Rubin

It is no secret that in the last twenty years we have witnessed and participated in an increased public awareness of the communication needs of Spanish speakers in the United States. This awareness and the subsequent publication in this arena has been part of two decades of public demands for the rights of a wide variety of underserved groups (for example, blacks, women and other minorities). Several papers at the El Espanol en los Estados Unidos Conference in October, 1981, pointed to aspects of linguistic inadequacy occurring in specific domains. For example, in the health domain, Martinez, Leone and Sternbach de Medina (this volume) discussed the role of language in limiting access to health care, and Arjona (this volume) detailed some of the complexities involved in implementing the federal law which governs the certification of court interpreters. Further, most of us are aware of at least some of the many activities of the federal and state governments, as well as those of MALDEF (the Mexican-American Legal Defense and Education Fund) and PRLDF (the Puerto Rican Legal Defense and Education Fund), which identify some communication inadequacies and propose some sort of solution to them. What I intend to do in this paper is to review how various communication issues vis-à-vis Spanish-only speakers have been dealt with, by considering the following questions: (1) What language/communication inadequacies have been identified and by whom? (2) Who are the planners who have the authority and power to make and influence language-related decisions? (3) What plans or goals have been set out to attend to the communication inadequacies identified? and (4) What attempts have been made at planning — that is, in what situations has there been or is there a real effort being made toward implementation of or feedback to a plan? In order to detail what linguistic inadequacies are identified by whom, where responsibility is placed for attending to these problems, and what policies are applied to whom, my review will consider these problems in the following domains: health/medical; legal ; employment ; communication ; citizenship ; social welfare ; and education. 1 Through this analysis we will observe attempts to break through and redirect

134

Joan Rubin

the focus of administrative agencies in serving limited-English speakers, as well as consider the role of private agencies in meeting the needs of limitedEnglish speakers.

The Health/Medical Domain In order to understand what might be identified as a communication inadequacy in the domain of health and medicine, it is helpful to consider what the main functions of this domain are. I think we could agree that they are to provide health care to its clients. For Spanish-only speakers, the question is whether patients have access to such services and whether the medical staff understands the patient's symptomatology well enough to make an accurate diagnosis of the illness and to negotiate an appropriate treatment with the patient. In the U.S., the identification of language/communication issues in the health domain has largely been left to the private sector. For Example, MALDEF (1979) in a brochure entitled Chicanas and Mental Health, noted Hispanic underutilization of health and mental health services, finding that one of the factors contributing to this state of affairs is that 'Few mental health facilities provide bilingual services or employ a bilingual/bicultural staff. They note that 'A therapeutic relationship cannot be established when the client and the therapist cannot communicate. This can occur when the emotional experiences of the patient cannot be fully expressed in the patient's primary language'. Martinez, Leone and Sternbach de Medina (this volume) have also focused on identification of the problems that discourage Hispanics from seeking treatment and affect the quality of care that they receive. Their preliminary investigation suggests that language may indeed be a crucial barrier to services. Another study was undertaken by the California Department of Health Services in response to complaints from the Hispanic community regarding the lack of bilingual services in the emergency center of the San Francisco General Hospital. Some of the principal findings of this study noted by Aguirre (1980) are: (1)

Investigators witnessed hospital employees in the emergency triage area trying to communicate in English and in hand gestures with monolingual Spanishspeaking persons.

(2)

In psychiatric emergency, a 24-hour unit with three shifts, none of the nineteen people on the day shift, none of the thirteen people on the evening shift, and only one of the fourteen people on the night shift could speak Spanish.

The medical domain is an area in which we are just beginning to recognize some of the important communication gaps. There is a strong need for further studies to detail how access to and effectiveness of medical care relate to

Spanish

Language

Planning

in the United States

135

communication barriers. It is interesting next to ask who the planners in this domain might be. That is, who are the individuals who have the authority and power to make and influence language-related decisions? In public hospitals, the decision rests with the hospital authorities or with the state or federal department of health which can insist that plans be made to correct this inadequacy. It is obvious that the public can also influence such decisions by complaining to the proper authorities. To date, I know of no legal suit brought in this area to influence the plans made by hospitals in providing health care to Spanish-only patients. We can next ask what plans or goals have been set out to attend to these communication inadequacies. There are several that have been made: (1)

Some hospitals now provide signs in Spanish indicating where particular offices are found.

(2)

Several textbooks have been published to train medical personnel in Spanish (note for example Medical Spanish. A conversational Approach (DilorenzoKearon and Kearon, 1981), Communicating in Spanish for Medical Personnel (Tabery, Webb and Mueller, 1975), or Basic Spanish for Health Personnel (Seyman, 1973).

(3) The M.D. Anderson Hospital in Houston, Texas, through its Patient Education Office provides Spanish language translations of the treatment explanations provided English-speaking patients. (4) Tarrant County in Texas now has Spanish speakers monitoring the Tel-Med system. Tel-Med is a free health and medical information service available by telephone, sponsored by the Fort Worth Academy of Medicine, the Tarrant County Medical Society, and its auxiliary members. Spanish speakers are available every Tuesday from 9:00 a.m. until noon. The service also has 40 tapes narrated in Spanish on a number of medical subjects such as family planning, alcohol problems, mental health, among others (from the Arlington Texas News, 11 February 1981).

Finally, we can ask whether there has been much planning in the medical arena — that is, what attempts have been made to implement the goals or promote the use of the products suggested, and what information has been sought to evaluate the effectiveness of the implementation? Apparently there has been little or no planning in this area.

The Legal Domain Most would agree that the main function of the law is 'to provide justice for the members of the community'. To the extent that the communication process prevents the carriage of justice because the accused cannot follow the trial or understand the charges, we can note a communication gap. Within the legal domain, perhaps the major communication inadequacy for Spanish speakers (and others who do not speak English) has occurred in

136

Joan Rubin

the role of the court interpreter. Until very recently, despite the critical importance of the interpreter in court cases, there were no measures (1) to require certification of the capabilities and training of an interpreter for the post nor to ascertain whether or not their understanding of the law and the interpretation process was adequate to the important assignment they were to carry out; or (2) to record the actual testimony of the witness so that the translation of the interpreter's version could be verified. On October 18, 1978, the U.S. Congress, acting as a language planner, enacted Public Law 95—539 'to provide more effectively for the use of interpreters in courts of the U.S. and for other purposes'. The goals/plan of this law is that the director of the administrative office of the U.S. courts shall establish a program to facilitate the use of interpreters in the courts. Here we see that the director of the administrative office becomes a language planner — charged with the authority to influence language-related decisions. The law further spells out what the director's main goal shall be, namely: to prescribe, determine, and certify the qualifications of persons who may serve as certified interpreters in U.S. courts in bilingual proceedings and proceedings involving the hearing-impaired. In so doing the director is to consider the education, training, and experience of those persons. Not only does the law appoint the director of the administrative office of the U.S.court as its chief implementor and give this officer a specific goal, it also gives some suggestions about how this plan is to be implemented: The director (1) shall maintain a current master list of all interpreters certified by the director, and (2) shall report annually on the frequency of requests for and the use and effectiveness of interpreters. Here we see that some evaluation suggestions have been built into the plan by the Congress. The law further spells out when the plan is to be implemented: The presiding judicial officer . . . shall utilize the services of the most available certified interpreter . . . if the presiding judicial officer determines . . . that. . . such party . . . or a witness: 1) speaks only or primarily a language other than the English language: or 2) suffers from a hearing impairment. . . . ' (204).

It also specifies under what circumstances: if the problem is great enough so as to inhibit such party's comprehension of the proceedings or communication with counsel or the presiding judicial officer, or so as to inhibit such witness' comprehension of questions and the presentation of such testimony.

Hence the law is quite specific as to its goals and more than a little specific about who is to carry out the plan and how they are to do so. It is less than specific about what is to be done with the evaluation which it prescribes. That is, if there are more requests than interpreter services, they do not specify what action the director of the administrative office of the U.S. courts is to take. Aijona (1981) details the complex procedure followed in

Spanish Language Planning in the United States

137

developing an instrument for certifying Spanish/English interpreters to comply with the stipulations outlined in PL 95—539. In this law, we see clearly (1) how the linguistic inadequacy was defined; (2) who the planners are; (3) what the plan is: and (4) what implementation and evaluation procedures are envisioned. Clearly, more work needs to be done here on the problem of implementation and on the evaluation of the effectiveness of such procedures in meeting the perceived inadequacy. Finally, we should note that no efforts have as yet been made to tape record the witness' testimony or, in state and local courts, to certify interpreters. I do not have complete information for state and local courts, but I understand that California does require certification of interpreters, and the Circuit Court of Cook County does train and test interpreters (Rodriguez and Ruiz, 1981). To my knowledge, there are no standards for determining the language competence of witnesses, defendants or plaintiffs in a court of law. Sometimes private individuals perceive a need and prepare a product to try to meet it. Such an example is A Spanish Manual for Law Enforcement Agencies (Luna and Meneses, 1973). These men tried to provide a product which met an inadequacy that they perceived — namely, the need by law enforcement officers to be able to communicate more effectively with Spanishspeaking inhabitants. The only indication that we have that this product has been recognized as being suitable to meet that need is a memo from a Park Ranger in the U.S. Department of the Interior National Park Service to the superintendent of another park, suggesting that the volume may prove of help to rangers in the second park. Clearly no planners (persons with authority to make or influence language-related decisions) were involved, nor is there a plan for the evaluation of this volume. It is possible that such a need exists and is so perceived by others, and that the volume is an effective means of meeting this need, but in this case there has been no planning involved.

The Work Domain In the domain of work, there are a number of relationships which can be identified between problems of getting work done effectively or efficiently and problems in the communication process. In the US, the issues identified in the work area relate to (1) access to employment information by a consumer or client in relation to their language knowledge, and (2) certification as a bilingual for certain positions. What language inadequacies have been identified in the work domain and how have these been dealt with? In its role as 'influencer of language related decisions', the Puerto Rican Legal Defense Fund (PRLDF) has brought suit in Barcia vs. Sitkin in the

138

Joan Rubin

State of New York. The linguistic inadequacy identified by the Fund is a Hispanic claimant's inability to understand English, which threatened to cost her hundreds of dollars in unemployment insurance benefits. The plan which this suit proposed was that the New York Appeal Board accord claimants elemental due process guarantees, such as the right of notice of the evidence considered against them. The Fund also seeks to require the New York Appeal Board to take into account the fact that in local unemployment insurance offices, claimants are often denied their rights to Spanishlanguage assistance, rights which were guaranteed by a 1976 court order in Pabon vs. Levine. The perceived linguistic inadequacy is the Hispanic's inability to understand information in English about unemployment benefits. The planners here are the PRLDF and the courts. The plan requires that information be made available in Spanish to those who do not understand the English materials. The case is of interest because it begins to illustrate the problems connected with language planning when done by the courts. The court can issue a decision or an order in a particular case. It may also provide for some guarantee of implementation at least in a specific case. However, unless the court case is connected with a regular administrative procedure which is carried out by the executive branch of the federal or state government, the chances of its implementation are remote. Even more remote is the possibility of feedback for improving implementation. The PRLDF notes this difficulty in the case of Pabon vs. Levine, where although the U.S. Department of Labor has regulations similar to those involved in the Law case, nonetheless, the court did not require an interpretive memorandum. Hence, in the case of employment benefits, it is necessary to prove a disparity of benefits. In educational agencies, unequal benefits are presumed. With the advent of state bilingual education acts, a need has been identified to specify what the language skills of bilingual teachers must be. The perceived linguistic inadequacy is that monolingual English teachers cannot provide adequate educational services to LEP/NEP (limited-English-proficient or non-English-proficient) children. The planners in this instance vary from state to state. In the case of the State of California, authority to set goals and insure compliance is vested in the State Commission for Teacher Preparation and Licensing. The plan is that all agencies — in this instance, (1) colleges and universities granting multiple and single subject credentials and granting bilingual/cross cultural credentials: (2) 28 designated assessor agencies granting a certificate of competence for Bilingual Education; and, (3) school district administrators who apply for an emergency credential for their teachers — must assess the language competence of the prospective candidate. Further, the Commission states that the teacher must be assessed in all four language skills (speaking, reading, writing and listening) and must demonstrate competence and proficiency roughly equivalent to the Foreign Service Level 3,

Spanish Language Planning in the United States

139

which is the professional proficiency level.2 Implementation is vested in the agencies named above and it is widely acknowledged that there is a great deal of variation in this implementation. However, a procedure for evaluation of these agencies does exist. In fact, the Commission is constantly monitoring them, and if necessary, it can either close an assessor agency or can withdraw credential approval. In an effort to tighten up this assessment process, the state has appointed a committee to provide clearer standards for the bilingual certificate of competence, though the report of this committee has not yet been issued.3 Florence Barkin (this volume) described the model which she and her committee are developing for a language assessment instrument which will be the official language proficiency examination for the State of Arizona, required of all teachers before receiving bilingual endorsement. In her paper, Barkin touches on some important problems of assessing language varieties and domains of usage, something which I understand has also been of concern to the California committee on standards,4 and on the need to involve the educational staff in the preparation of the test. Some other states (Michigan and Texas, for example) also have similar procedures to certify the language competence of those involved in bilingual education. It is of interest that in the area of work where language is seen as a scarce resource, the procedures for planning are fairly well elaborated though certainly there is room for improvement. There is another area where there should be some perception of linguistic inadequacy in the work domain, but in which there has not been much to date. That is, international business corporations have not seen a need for their American employees to learn Spanish (or any other foreign language for that matter) to any great extent. According to three studies (Olympus Research, 1976: Inman, 1978: and Berryman et al. 1979), many U.S. companies require their overseas employees to learn English rather than have their executives learn Spanish. It is very hard to identify cases where lack of Spanish knowledge has been perceived as negatively affecting marketing, contract confirmation or the provision of service. However, increasingly there are individuals in specific companies who are pointing out that language knowledge can make the difference in the sealing of a contract or the provision of a service.

The Communication Domain The issue here is one of access to information and/or service. In California, the Hispanic Coalition has argued for the past ten years that Hispanic customers are not treated equally because they are not given services equal to

140

Joan Rubin

that of English speakers. The linguistic inadequacy is the inability of the California Telephone Company to understand and provide services to its Spanish (and in some cases) Chinese customers. There are a variety of planners in this instance: (1) the Hispanic Coalition which has been pressing the California PUC (Public Utilities Commission) to see that services are provided for Spanish-speaking customers; (2) the California PUC which has compelled the Pacific Telephone Company to provide emergency services; and, (3) the Pacific Telephone Company which has decided on the nature of the services. The plan is a very complicated one. In 1976, the California PUC ordered Pacific Telephone to provide emergency telephone services in Spanish and Chinese to those areas where the population of Spanish-speakers or Chinesespeakers was more than 5 percent. In the same year, Pacific Telephone implemented this order by contracting with a phone answering service in Los Angeles. The service was called ESLAB (Emergency Spanish Language Assistance Bureau). The Hispanic Coalition has questioned the effectiveness of this service. Although there is no documentation of the number of Spanish language calls which are referred to it, in 1980 Pacific Telephone reported to the Commission that 90 percent of the calls that go to ESLAB are handled satisfactorily. The Coalition has questioned the definition of satisfactory for users in towns other than Los Angeles, since the problem of transmitting information is compounded by lack of direct interaction between the operator in the other town and the person reporting the emergency. They argue that Spanish language operators should be part of the Pacific Telephone personnel, available in all cities where there are sufficient Spanish telephone users. In 1979, the California PUC reopened the bilingual services case and held hearings to determine if the Pacific Telephone had complied with the order for bilingual emergency services and to investigate whether Pacific Telephone was providing the same services to non-English speakers as it was to English speakers. To answer these questions, Pacific Telephone contracted the services of Herman Gallegos, a member of the Pacific Telephone Board, to study the needs of the Spanish-speaking community. A MALDEF staff member reported that Mr. Gallegos interviewed only Pacific Telephone customers. The PUC, recognizing that his study was biased, is now calling for an independent survey of the telephone needs of the Spanish-speaking community. What we can observe in this instance is a complex interaction among those who have the authority and power to make or influence languagerelated decisions. We can also note the problems which arise due to the lack of clear-cut specification of the implementation and evaluation process. Consequently, the only recourse of the pressure groups is to take every opportunity to reopen the issue.5 In the area of media, the linguistic inadequacy identified is the need for Spanish language broadcast time. Until the deregulation of communication

Spanish Language Planning in the United States

141

which began with the Carter administration, there was some basis for arguing that lack of provision for broadcasting to monolingual Spanish speakers constituted a communication gap. The Communication Act of 1934 (now being rewritten by the present Congress) argued that airwaves are public property and that stations are licensed to use these airwaves in the public interest. Therefore, the Federal Communication Commission (FCC) ruled that stations were obliged to serve the needs and interests of the communities in which they were licensed. Since the FCC has had the authority to grant and suspend licenses (that is, it is the planner in this case), the Commission has construed its responsibility to serve the interests and needs of specific communities as a basis for, at times, requiring broadcasting in languages other than English (for us, this is the plan or goal). Further, the FCC has interpreted the public interest to mean minority ownership of stations, since ownership influences station policies, and those in turn influence the way in which a population is served. We can mention a case in point which has recently been ruled on in Phoenix. In this case, the American International Development (AID), which is owned and managed by a Chicana, proposed to bring Phoenix a twenty-four hour Spanish-language FM radio station. The applicant was, however, competing with two other (Anglo) applicants for Phoenix's last remaining FM channel. The case has gone through three review boards: (1) an FCC administrative law judge who ruled in favor of the Chicana; (2) an FCC review board which overturned that decision on technicalities; and (3) the FCC itself, which sustained the first decision in favor of the Chicana applicant. The case has been appealed to the U.S. Court of Appeals but in the opinion of the attorney for the Chicana,6 the decision of the Commission will most likely be sustained. There is very little planning in the media domain. The Commission does not monitor the number of non-English speakers in a community on a continuing basis to ascertain need, as it relates to the number of stations or other non-English media outlets. Rather, the issue is decided on a case-bycase basis. To date, at least, the Commission has had a social policy — the 1934 law - which it could invoke to decide specific cases. However, if the law is revoked and deregulation continues, issues of access to the media by Spanish-only inhabitants will probably be heard with less sympathy.

The Citizenship Participation and Representation Domain If we look next at the domain of citizenship participation and representation, we find that there are two areas where language inadequacies (or related issues) have been identified: (1) the need for Spanish language forms in

142

Joan

Rubin

taking census counts, and (2) the need for Spanish language services connected with voting. Looking first at the census question, we can see that the linguistic inadequacy is of two kinds — first the need for Spanish language forms and second, the question of what kind of Spanish should be used in such forms. The 1980 census provided Spanish language questionnaire forms. In the May 4, 1978, Los Angeles Times, Martinez (for MALDEF) indicated how pressures were brought and rationale provided for this decision. She noted that an advisory committee was set up by the Census Bureau to identify the reasons for the 1970 census undercount of many language minority groups. The committee found that one of the reasons was the 'inability of census takers to communicate with respondents speaking Spanish'. They also found that in 1970 'Spanish-speaking families who could not read the forms often did not respond and, to compound the problem, non-Spanish speaking interviewers were frequently sent out to collect the missing information'. In other words, at the insistance of various civil rights groups, acting as language planners, the Census Bureau did investigate a problem which it turns out included a language inadequacy in their own administration of the census and as we know did attempt to correct the problem by providing for Spanish language forms. There are two examples of how the question of what kind of Spanish to use in a census form has been met. In 1975, the Bureau of the Census pilottested a Spanish language form in Texas which was to be used in the 1980 United States census. In a letter to the Census Bureau, Troike (1981), identified this census translation as inadequate, and his observation motivated the Bureau to give more serious consideration to the translation process. We will describe that process below. However, in 1976 the National Center for Educational Statistics contracted with the Center for Applied Linguistics (CAL) to provide a questionnaire in Spanish for the survey of income and education to be administered in collaboration with the Bureau of Census count of the children of poverty. The goal was to provide a form which would be understandable to speakers of the several regional varieties of Spanish in the U.S. Troike (1981) describes the process of implementation in the preparation of this form. CAL assembled a group of four translators, one Puerto Rican, one Cuban, one Mexican-American from Texas, and another from California. one member of the team was first assigned to translate the English version of the questionnaire into Spanish, and the resulting version was then given to the others to back-translate into English. The team was then brought together and on the basis of discussions, comparisons of back-translations with the original, and consultation with various sources and resource personnel, they developed - or, rather, negotiated - a consensus translation.

Spanish

Language

Planning in the United States

143

Troike describes (1981) some of the language issues which emerged. For example, employer was translated as Para quién trabaja1 rather than as patrón, since the English could refer to a company as well as a person. Armed forces was translated as fuerzas armadas, although several members preferred ejército, because the latter was too restricted; el militar was not accepted by any.

Troike gives other interesting compromises, expressing regret that there were neither time nor funds available to field-test this form, since no real evaluation of its effectiveness did in fact take place. The Census Bureau took a different tack in meeting the problem of a need for intelligible Spanish, given the several varieties of Spanish used in the United States. For the Commonwealth of Puerto Rico, they prepared a form in Puerto Rican Spanish.7 For the rest of the U.S., the Decennial Census Division worked with an advisory committee in preparing a form which would accommodate the several varieties of Spanish. They used two strategies: (1) They provided two alternate forms side-by-side. For example, for 'toilet' they said inodoro (excusado), and for 'truck', troca (camión). (2) They gave a Spanish form along with the English equivalent: bombas eléctricas (heat pumps). The plan was created by the Decennial Census Division working with an advisory committee. The Bureau did attempt to get feedback (evaluation) on the validity of the translation by pre-testing the questionnaire in Austin, Texas; New York City; and Oakland, California, where speakers of the three major Spanish language varieties are concentrated. 8 Still, even with a Spanish translation, the question of collecting data from Spanish-only inhabitants had to be dealt with. MALDEF expressed concern in the May, 1978 Los Angeles Times article about whether the implementation would be adequate, since at the time only English language forms were to be mailed to homes, even in areas where the population was predominantly Chicano. They noted that there would at least be a small notice in Spanish telling recipients that they could request a version in their own language. The Census Bureau did, in fact, have a fairly elaborate implementation plan to reach non-English speakers. They worked with community groups, which contacted people in the census and apprised them of the importance of the community; they also had spot notices on television in Spanish, telling of the forthcoming administration of the census. They also had similar notices in other media and informed the public through brochures of the importance of the 1980 census.9 Their effectiveness can be measured by the enormous increase in the count of non-English speaking inhabitants. Another language issue in the citizenship area is the perceived need for Spanish language services for non-English-speaking citizens who participate in the voting process. This became a hotly contested issue, since many voting administrators have challenged the provision for bilingual voter services in the

144

Joan Rubin

Voting Rights Act of 1975. Nonetheless, it is helpful to understand some of the issues involved and to try to see how much we know about the planning process. Until 1975, the right to bilingual election procedures derived only from section 4(e) of the Voting Rights Act of 1965, in which it was noted that it was a manifest injustice to bar Puerto Ricans from voting because of illiteracy in English (Teitelbaum and Hiller, 1977). Section 4(e) was targeted to prohibit the application of New York's literacy test to Puerto Ricans residing in that state. But, as Teitelbaum and Hiller (1977) note: 'eradicating the use of literacy tests was not enough, since registration and voting was conducted in English, a language incomprehensible to the intended beneficiaries of section 4(e)'. Consequently, beginning in 1973 a series of lawsuits were filed charging that English-only elections constituted a condition on the right of Spanish monolingual Puerto Ricans to vote. In the important case, Puerto Rican Organization for Political Action vs. Kusper, the court determined that 'the right to vote' meant the right to an effective vote, and that a Spanish monolingual could not cast an effective vote when confronted with election materials she or he could not understand. All of these cases, and several others, provided the legal underpinning for the bilingual election provisions contained in section 203 of Title II and section 301 of Title III of the 1975 Voting Rights Amendments. The Attorney General is charged with the implementation of the Act. The law specifies that a jurisdiction is deemed in violation of the Act if the Attorney General determines that any voting or registration materials provided to voters are exclusively English when five percent of the voting age population within any state or political subdivision are members of a single language minority and the rate of illiteracy of the particular language minority exceeds the national average of voting-age citizens. We can note that while the court suits focused on the inability of speakers to vote because of a lack of knowledge of English, this law rested on figures indicating numbers of people belonging to the minority group rather than upon their knowledge of English. Only a little is known about the actual implementation of this law. The Federal Election Commission contracted with the University of New Mexico to survey creative ideas about how bilingual elections could be carried out (Department of Linguistics, 1978). Their survey considered Spanish, Chinese, and Native American language minorities in the U.S. According to David López, one of the principal investigators, the survey found that there was considerable hostility toward the legislation, but it was able to identify some exemplary procedures, most of which related to the process of voting, itself. However, Lopez considered this an inappropriate point of entry, since the first and most important barrier to voting was at the point of registration; without Spanish-speaking officials at the registration sites, Spanish-speaking voters could not begin the voting process. The law is not very specific in

Spanish Language Planning in the United States

145

identifying the specific services which the state, country and local officials must provide. It is clear that there is no regular mechanism for evaluating the quality and quantity of these services.

The Social Welfare Domain In the domain of social welfare, we found a clear instance of the access to quality and quantity of services being dependent on meaningful communication. The issue has been most extensively debated in New York City by PRLDF and others. Without going into the complexities of the planners and the plan, I think that there are a couple of interesting aspects of this domain to be considered. The PRLDF argued that because there were no Spanish employees or forms, persons speaking only Spanish could not get equal access to welfare benefits. They argued that Spanish-only applicants received fewer or no benefits and were unequally subjected to hardships because they were required to provide their own translators. This resulted in their having to return more frequently to the Center, in their having to keep young children out of school so that they could serve as interpreters (hence exposing them to family traumas which they perhaps should have been spared), or in violating their own privacy, when an acquaintance served as interpreter. These complaints were subsequently substantiated by the Mendoza Report, commissioned by the U.S. Department of Health, Education and Welfare and issued in September, 1978. Of particular interest is the fact that the PRLDF was unable to base its claims on language inadequacies, but rather had to rely on Title VI of the 1964 Civil Rights Act, which prohibits descrimination on the basis of national origin. This case of social welfare is also of interest because it illustrates the complex interaction between the need for recourse to the many sources of authority for language-related problems. In this case, in the first instance, in 1973, the complaint was made as an administrative one to the OCR (Office of Civil Rights). When nothing was done there, the complaint was filed in Federal Court, which required that the defendant prove that the complaint was incorrect. The OCR then did an in-depth study which resulted in the Mendoza report. In 1979, the OCR issued a letter of findings based on this report which triggered the New York City Department of Social Services into submitting a plan in which the city agreed to hire provisionally 272 bilingual workers. These figures were then disputed by the PRLDF. The OCR replied that it would monitor the department and then make a decision. In August, 1981, the OCR found that the city was still in violation of Title VI. The PRLDF is still not happy with the implementation of the law because: (1) there is no guarantee that Spanish-only speakers will be assigned to

146

Joan Rubin

a bilingual worker (they argue against the current policy which assigns applicants on a random basis without consideration of their language needs); (2) they want the city to provide periodic reports on the actual number of bilingual workers; and (3) they want the bilingual workers to be given civil service status so as to guarantee that the service will be available. The PRLDF is considering what the best next move is since the Reagan administration ethos tends toward deregulation. The OCR suggested that the Fund should be satisfied with its gains rather than push for too high a profile. The PRLDF still is considering going back to the courts. I have not here detailed all of the complexities of this case, but it illustrates some of the problems of not only seeing that a law applies in a particular instance but also, once having done so, seeing that the ruling/regulation is implemented in an effective manner and on a regular basis. Further, although the Civil Rights Act applies across the nation, there are considerable difficulties in seeing to it that the New York decision is brought to bear in other communities where there are non-English-speaking welfare applicants. In 1979, MALDEF petitioned the Department of Health, Education and Welfare to have this law apply across the nation. The PRLDF sent a similar petition in 1979 to the Department of Labor. To date, there has been no response, and the likelihood of a response is slim in an era of deregulation and less government coercion in all areas. 10

The Educational Domain Finally, we come to the domain of education, which I have deliberately left for last because it is so complex. In the other domains we have considered, the correlation between language knowledge and access t o or the quality of services provided is relatively straightforward. Further, although the process of ensuring implementation is complex in the other domains, it is overwhelmingly and extraordinarily diverse in the educational domain. This is in large part because of the complex relationship between federal and state regulations in the domain of education. Hence, one would have to investigate, and I hope that someone will do so on a statewide basis, who the planners are, what the plan is and how planning is carried out. Another continuing area of concern in planning is the problem communities have in agreeing on what the function(s) of education is/are: is it to provide for children's cognitive growth, is it to socialize them, is it to prepare them for work, is it to make good citizens out of them, is it to help in their self-image, or something else? Unlike the medical domain, where the function is relatively clear (although there are many hidden agendas among the medical personnel which prevent the satisfactory delivery of services) agreement on the function^) of education is not common. Indeed there has been continual debate

Spanish Language Planning in the United States

147

around this issue. Most recently, the debate about the functions of education has focused upon questions of the validity and justice of competency testing. A large part of the problem is that people have differing opinions about what education is supposed to be achieving. However, we do have several planners and plans which have agreed that LEP/NEP children are not being adequately served and that something other than the traditional English program must be provided for them. The legal cases in which this has been mandated are well known. Lau vs. Nichols is the landmark case on the federal level. Aspira of New York vs. New York City Board of Education is an important landmark case in the states. There are and continue to be many other cases which specify that limited Englishspeaking children must get special care. How is this decision to be implemented? On the federal level, the OCR issued the Lau remedies which spelled out how the Lau vs. Nichols court decision is to be implemented. The remedies specified that only bilingual education was an acceptable program unless a school district could prove that another treatment was acceptable. These remedies were finally challenged by the State of Alaska on a technicality — that normal federal procedure, requiring a hearing before a regulation becomes law, had not been followed. The OCR attempted to provide new regulations which were more specific about the nature of the instruction, the certification of teachers, the process of entry and exit into these programs, and the nature of testing. Public hearings were conducted in six cities throughout the U.S. and over 7000 depositions were taken. Secretary Bell rejected these regulations in January 1981, striking down a federal means of ensuring bilingual education for LEP/NEP children. Nonetheless, the federal government still continues to provide monies toward the implementation of bilingual programs by supporting such programs as: materials development centers; technical assistance centers; teacher training programs; evaluation, development and assessment centers; research on bilingual education; and the National Clearinghouse for Bilingual Education. Many of the states have also enacted bilingual education laws which provide for limited English speakers. That of California tries to spell out much of the implementation process in detail - how many children, what kind of treatment, how many teachers, what kind of competence teachers should have, and what sort of testing and placement. California Law AB507 took effect in September, 1981; however, all of the regulations associated with it are not yet in place and administrators are still looking to the state for guidance in this area. There are two other language-related issues in education that should be mentioned, both of them concerned with what could be called 'corpus planning issues'. The first relates to the testing of the children's knowledge of Spanish, and the second with the development of materials for Spanish-

148

Joan

Rubin

speaking children. Troike (1981) described how the Materials Development Center in Miami tried to serve children using different varieties of Spanish. He notes that they developed separate parallel curricula for Mexican-American, Cuban, and Puerto Rican Spanish in response to local demands. Further, he feels that most of the differences among these varieties for written purposes appear to lie in the names of objects (nouns, for example) and cites a few examples: ' . . . the Southwestern tortilla, a flat, thin, circular, unleavened corn or wheat bread, is an omelette in the East, while an orange is a china in the East but a naranja in the West'. Keller (1982) has presented extensive information on the corpus planning issues in Spanish in the U.S. He notes that the problem has been to decide which variety of Spanish to use in the classroom. He finds that 'The answer has been often made in the form of one or two extremes. There are those who exalt the ethnic form of their locality and denigrate what the American Association of Teachers of Spanish and Portuguese has called "world standard Spanish" '. Conversely, there are those who exalt 'world standard Spanish' and denigrate the ethnic or folk form. As a more specific example, Keller cites his experience in 1974 as linguist for an NIE (National Institutes of Education) project, which was to evaluate approximately 100 curriculum titles in Spanish bilingual education. The project found eight types of Spanish in actual existence in 1974 in Spanish bilingual education programs. (1)

Programs which use 'world standard Spanish'. Some of the language is free of regionalisms. Some of the language may not be understood by United States Spanish speakers who use a regional or ethnic designation instead of a standard one (e.g., these programs use autobús, but not camión, nor guagua).

(2)

Programs which use language specific to particular regions or social groups of the Hispanic world outside of the United States, such as Spain, Bolivia or Chile. For example, these programs may use micro (Chile) or autocar (Spain), but not autobús, camion or guagua.

(3)

Programs which use language characteristic of all the regions and ethnic varieties of United States Spanish. That is, the programs use guagua and camión but not autobús.

(4)

Programs which use language characteristic of the eastern United States and the Caribbean: for example, they use guagua, but not autobús nor camión.

(5)

Programs which use language characteristics of the western United States and Mexico. For example, they use camion but not guagua nor autobús.

(6)

Programs which use non-standard, non-Spanish (as in bad translations).

(7)

Programs which use both the regional or ethnic varieties of language as well as the 'world standard Spanish' variety. For example, they use camion and guagua in addition to autobús.

(8)

Programs which use controlled 'world standard Spanish', using only common language forms, for which there are no alternative regionalisms or ethnic varieties (hence eliminating camion, guagua and autobús from instructional materials) (Keller, 1982).

Spanish Language Planning in the United States

149

Keller argues that there needs to be appropriate compartmentalization. Type (1) he feels is best used in more advanced Spanish language courses, particularly in the content areas such as math and the sciences. Types (4) and (5) should be used in the relevant regions in transitional bidialectal education. Type (8) he feels successfully deals with miscues that would otherwise crop up in teaching literacy. The complexities of language planning in education are fascinating and would require several volumes to discuss all of the issues, planners, plans, and processes of implementation and feedback.

Summary We have surveyed some of the major kinds of actors who have appeared and the activities they have undertaken in the United States in the last 20 years in an attempt to provide services to the Spanish language community. We have focused largely on federal governmental planning activitites, but have provided examples of state or non-governmental attention to language inadequacies. All of the examples are ones where the 'instrumental' communicative function of language is clear. We can note that in some cases the issues could be treated in terms of language, whereas in others, it was necessary to use legal precedent to provide services for Spanish-only language clients. As a summary of cases from Teitelbaum and Hiller (1977) notes: Accommodating non-English speakers in areas other than education requires a showing that without bilingual services, the protection of substantial rights is dimiinished. Thus, non-English-speaking plaintiffs must premise their language discrimination claim on the preservation of established rights, and not naked demands for bilingual ism (96).

They note, further, that the greatest gains for non-English speakers has been where the rights at issue have been deemed to be of particular importance by the courts: for example, both voting rights and inequality in the criminal justice area are scrupulously probed, whereas the courts have given less favorable treatment to unequal delivery of social services. The range of interaction of planners in the identification of problems and the making of a plan in any one domain is great. We have noted the extensive participation of the PRLDF and MALDEF, as well as others in influencing language-related decisions. At the federal level, there is a very wide range of planners who have the authority and power to make language-related decisions: decisions can be made through judicial decrees, they can be made through executive regulatory decisions, or they can be made through congressional law. Although the PRLDF and MALDEF have thus far concentrated

150

Joan Rubin

on litigation as a means of affecting policy, cases decided in a particular court are not automatically applied to other states without further litigation or executive regulation. Further, decisions made through litigation are difficult to implement, since the courts do not have the staff to insure implementation and often appoint someone without much power to oversee the court order. When a decision is taken in the courts and then implemented by an administrative branch of the executive, the possibilities for effective implementation are greater. For example, once the courts ruled in Lau vs. Nichols that limited English speakers needed more than 'English only' treatment, the Office of Civil Rights had the basis to issue regulations implementing bilingual education in all of the school districts. As we have noted, there are other planners on other levels; often these planners can be found in the private sector. The plans occur in a wide variety of forms and products. In the federal case, as we have indicated above, they come most often in the form of court decisions, laws or administrative regulations. But they can also consist of textbooks which contain specific kinds of language, or dictionaries which prescribe correctness or designate regionalisms, and so forth. Finally, I would like to note that the area of planning is the one where the actors are most diffuse. To the extent that the plan specifies who is to implement the plan in what way, who is to monitor the implementation process according to what sort of guidelines, and what they are to do with this monitoring, we can observe that the planning stage can be more effective. However, in most of the instances we have examined, the plan is not very specific so that we find that there is a very disjointed, continual attempt to define the basis for an effective and appropriate implementation procedure. For example, the PRLDF in two separate cases in the educational domain' brought suit to ensure adequate implementation. One suit related to the dispersal rather than the concentration of Hispanic children in Delaware (Evans vs. Buchanan). Another related to the need to hire more Hispanic teachers (Morgan vs. McDonough). Nonetheless, we can note that when regulations are too specific and the implementors do not agree with them, often because they do not suit their specific situation, then implementors may choose to ignore or violate them as long as they can. So, enlisting the participation of the implementors in the process of making a plan has been recognized as essential by some students of social planning (Webber, 1978). We can conclude that over the past twenty years the United States has been deeply involved in defining language inadequacies which limited English speakers may have in many public domains. Several of the issues have been carried out on behalf of Spanish-only speakers. Finally, I would like to note that scholars have a very important role to play in this process. First and foremost, they can be very influential in defining language-related difficulties

Spanish

Language

Planning

in the United States

151

experienced by Spanish-only speakers. Second, they have been and will continue to be important in providing information for making a plan and in the actual planning process, b y helping to define adequate criteria for implementation as well as criteria for the evaluation of such plans. It is an extremely exciting and important area for the application o f linguistic and sociolinguistic skills t o monitor the ways in which attention to and solutions for language inadequacies arise and are resolved in the public domain.

Notes 1. A version of this paper was the Keynote Address at the El Español en Los Estados Unidos conference, Chicago, October 3, 1981. I am indebted to Richard Baldauf for inspiration on the organization of the paper. 2. Level 3 means that the person can participate fully in conversations with native speakers of the language on a variety of topics, including professional ones, with relative fluency and ease. The person should have mastered most of the major grammatical features of the language and enough vocabulary to cover a large number of topics. 3. This information was obtained through the courtesy of Maria Ortiz, California Department of Education. 4. Personal communication, Concepción Valadez. 5. According to information provided by attorney Ann Hill of MALDEF, one other state has looked into the language needs of its Spanish-speaking population: Connecticut. The Southern New England Telephone Company (SNETCO) has studied the needs of the Hispanic community and held meetings to ascertain whether their needs were met. This was done because in filing for a rate increase, SNETCO was required to demonstrate that it was satisfying its customers. SNETCO has decided that the way to meet the needs of Hispanics is to provide an Emergency Spanish Assistance Lab modeled after that of Los Angeles. 6. Personal communication, attorney Margo Polivy. 7. Personal communication, Angel Landron, of the Census Bureau. 8. Personal communication, Carmina Young, Bureau of the Census. 9. Ibid. 10. I am indebted for this information to Robert Becker of the PRLDF and to Adalberto Aguirre (1980) for details on the Mendoza report.

References Aguirre, Adalberto. ( 1 9 8 0 ) . The political e c o n o m y context of language in social service delivery for Hispanics. Discussion paper, Conference on Ethnicity and Public Policy, May 3 0 - J u n e 1, 1980. Berryman, Sue E., Paul F. Langer, John Pincus, and Richard H. Solomon. ( 1 9 7 9 ) . Foreign Language and International Studies Specialists: The Marketplace and National Policy. Santa Monica, CA: R A N D . Prepared for the National Endowment for the Humanities.

152

Joan Rubin

Carrillo, C., G. Gibson and C.A. Estrada. (1979). Chicanas and Mental Health. San Francisco, CA: MALDEF. Di Lorenzo-Kearon, M.A. and T.P. Kearon. (1981). Medical Spanish: A Conversational Approach. New York: Harcourt Brace and World. Department of Linguistics (DOL). (1978). Provision of Bilingual Registration and Election Services. Final Report for the Federal Election Commission. December, 1978. Albuquerque: University of New Mexico. Inman, Margaret E.P. (1978). An investigation of the foreign language needs of U.S. corporations doing business abroad. Unpublished dissertation. University of Texas at Austin. Keller, Gary. (1982). What can language planners learn from the Hispanic experience with corpus planning in the United States. In J. Fishman and J. Cobarrubias, (eds.), Progress in Language Planning: International Perspectives. Berlin: Mouton. Luna, Larry and Adalberto Meneses. (1973). A Spanish Manual for Law Enforcement Agencies. Las Vegas, NV: Latino Languages Enterprises, Inc. Mexican-American Legal Defense and Education Fund (MALDEF). (1981). Docket. San Francisco, CA: Litigation Department. Martinez, Vilma S. (1978). Uno, dos, tres . . . will the next census again skip millions? Los Angeles Times. May 4. (Distributed by MALDEF). Mendoza Report. (1978). Access ofNon or Limited English Speaking Persons of Hispanic Origin to the New York City Department of Social Services. US Department of Health, Education and Welfare: Office of Civil Rights. Olympus Research. (1976). Language for the World of Work. Salt Lake City, UT. Public Law 95-539. (1978). Court Interpreters Act. October 28. Puerto Rican Legal Defense and Education Fund, Inc. (PRLDF). (1981). Docket of cases. June, 1981. Mimeo. Rodriguez, M. and C. Ruiz. (1981). Interpreters Manual. Circuit Court of Cook County. Office of the Chief Judge. Mimeo. Seyman, M.R. (1973). Basic Spanish for Health Personnel. Garden Grove, CA: Trainex Press. Tabery, J.J., M.R. Webb and B.V. Mueller. (1975). Communicating in Spanish for Medical Personnel. Boston: Little Brown. Teitelbaum, Herbert and R.J. Hiller. (1977). The Legal Perspective. Bilingual Education: Current Perspectives. Vol. III. Arlington, VA: Center for Applied Linguistics. Troike, Rudolph. (1981). Language Problems and Language Planning of Spanish in the United States. Language and Development. Webber, Melvin. (1978). A difference paradigm for planning. In R Burchell and G. Sternlieb, (eds.), Planning Theory in the 1980s: A Search for Future Directions. New York: Rutgers University Press.

Language as a Barrier to Health Care by Deborah Martinez, Elizabeth A. Leone, and Jennifer Sternbach de Medina

Introduction In order to study bilingual health care in the largest medical center in the upper Midwest, many obstacles still must be overcome; these are many of the same obstacles that still need to be overcome by Spanish-speaking persons seeking health care in the Midwest in general. The problems consist of a lack of accessibility of information needed by health care workers and a lack of accessibility of services for the Spanish-speaking persons seeking health care. In other words, the obstacles of both health workers and Spanish-speaking patients could be summarized in more abstract terms as involving a lack of power because of a lack of access to sources of power (Melville, 1980a). In this paper, the authors have documented the process of empowerment that a group of health care professionals in the 'twin cities' have begun in order to ensure that the second-class health care available to Spanish-speaking persons be changed to a first-class health care delivery system.

Background The lack of power and numbers of minority persons in most professions has long been documented and is still a reality in the United States. Alvarado challenges the assumption made by many educators that the ethnic balance of nursing professionals is being achieved by examining the statistics of proportional representation in the nursing field (Alvarado, 1980). In the large medical complex of hospitals and medical school facilities at the University of Minnesota in the Twin Cities, there is still no course that exists in medical terminology at the time this article is being written. Persons entering the

154

Deborah Martinez, Elizabeth A. Leone and Jennifer

Sternbach

School of Public Health at the University of Minnesota, for example, are still being told that they must choose between international and domestic health care programs, and that Hispanic health care is not a specialization open or available to students (Sternbach de Medina, MS). In genetic counseling training at the University of Wisconsin at Madison, students are told that translation services are often provided by friends or relatives accompanying the patients (Martinez, MS). Just as important as the need for health care workers to know the Spanish language is the need for these health care professionals to be trained in the appropriate medical terminology. Many times the accompanying friend or relative of the Spanishspeaking patient may be well-educated but nevertheless unable to translate the very technical and usually very intimate questions and answers between doctor and patient. Consequences of the lack of trained Hispanic health care workers - who are informed about the language and culture of Hispanics and well-versed on technical medical procedures and terminology — are also well-documented. In a court action in 1978, Vêlez describes the courtroom environment and consequent breakdown in communication as influential in the decision against the ten Mexican women appealing their case of nonconsenting sterilization in Los Angeles (Velez-I, 1980). The chief argument in favor of the ten plaintiffs - "the ideology of cultural differences" - was used against them and referred to as the "atypicality of their cultural traits". the women were so culturally different that the doctors could not have known that the sterilizations would have affected them in so adverse a manner. This belief removes the legal and moral responsibility for their (the doctors') actions (Velez-I, 1980: 246).

Eight of the ten women suffered irreparable damage in their husband-wife dyadic relationships as a result of the nonconsenting sterilizations. Many cases such as these exist which document the consequences of language and cultural barriers to first-class health care. In the Twin Cities, for example, a young woman not given translation services was not informed as to the extreme pain she was to undergo during the cerebral ventriculogram. This jeopardized both the patient's care and the hospital's image, in addition to creating problems for the attending physician who was unable to continue the procedure under first-class medical conditions (Velasco, 1981). In addition to the documentation of the lack of Hispanic health care workers and of the consequences of the lack of these Hispanic workers, there is a growing documentation of the cultural differences and barriers that language and culture create when no trained Hispanic health care workers are available. In an 18-month study of the Model Cities Family Planning Clinic in Alcala, Texas, Urdaneta documents the use of abortion by MexicanAmerican women and states that "contrary to the widespread impressions and cultural arguments, Mexican-American women do use this option (abor-

Language as a Barrier to Health Care

155

tion) when available" (Urdaneta, 1980). She goes on to examine the factors of the health care delivery system and their role in either facilitating or inhibiting Mexican-American fertility regulation. In Urdaneta's interviews not only was Spanish used, but the interviewers used both the lay Spanish terminology and the medical terminology in Spanish, in order to explain to patients the medical procedures that were to take place. Andrade, in a review of Mexican American family planning practices, succinctly pinpoints the problem of accessibility: It is the disjuncture of the Anglo, English-speaking, middle-class health care delivery system in terms of implementation as it intersects with the Mexican-American system of utilization that causes the problem - not the microsystem of the Mexican American population and their beliefs (Andrade, 1980).

Naranjo examined the attitudes of 117 college undergraduates, and the 28 Mexican-American females in the group demonstrated some of the most positive attitudes on birth control in described cases of genetic risk. (Naranjo, 1976). And so, the typical misconception that Mexican-Americans and Latinos in general do not advocate abortion was found to be inaccurate. Other studies of Mexican-Americans and their use of the health care system document that knowledge of the English language is the most important cognitive facilitator of acculturation and therefore familiarity with the Anglo culture and the utilization of the social service system (Melville, 1980b). In a study on the health and illness perceptions of the Chicana in Santa Clara County, California, it was found that place of birth (Mexico, Texas and California) were factors in determining sexual preference of physicians and ethnic preference of physicians: the first choice would be a female Spanish-speaking physician (Manzanedo, Walters and Lorig, 1980). The same study indicated a high utilization and a strong trust of pharmacists and the importance of considering husbands/partners when making serious health decisions. Regarding the relationship of the Chicana to the health care system, Manzanedo et al. mention three important facts: (1) it has long been recognized that the Chicana has had a major role in the health and economics of her family, at least within the cultural context brought from Mexico; (2) for many years the health service institutions have not made the effort to get Chicana participation because of the stereotype that certain programs were contrary to cultural values; (3) "the role models for minority females in the health care system are no longer only the cleaning women. At present, community workers, physicians, nurses, dental hygienists" and others are of ethnic minorities.

156

Deborah Martinez, Elizabeth A. Leone and Jennifer Sternbach

In concluding, Manzanedo, Walters and Lorig mention five important factors in successful participation of Chicanas in health care programs and practices: (1) Programs should meet health care needs expressed by the community. (2) Programs should involve community members and participation of community members should be sought. (3) Programs should be a part of the participant's life and not separate from it. (4) Programs should, whenever possible, utilize community members as providers. (5) The standards to evaluate success of a program should be set by the community. Two studies on Chicanas and health care both emphasize the very important role of the family in making certain kinds of decisions about health care. In a paper on childbirth using an ethnographic approach, Kay (1980) documents the different practices of three generations of Mexican-Americans and one group of Mexicans in southern Arizona. In brief she states that today some Chicanas "have come full circle" and have adopted their grandmothers' practices of childbirth, thus "returning to the family the functions that the dominant society had delegated to official medicine", and "for the Mexican woman who does not have resources for scientific medical care, there is a long tradition of women's mutual help and lore that is being revitalized". Wagner and Schaffer (1980) examine the social networks and survival strategies of female family heads in San Jose, California and conclude that despite the urban setting and changing lifestyles in the urban areas, the preferred resource was the kinship network. Velasquez and Velasquez's (1980) framework of bicultural assessment is one first attempt at identifying the cultural background of Latinos for purposes of better social service delivery. They have put together a type of ethnicity measure that examines the 'language comfort level', the food, the values, the settlement and household patterns, generational patterns and other variables that may indicate to health care professionals the type of language and cultural needs of the person being served. If such a bicultural assessment is possible, at least in part, then perhaps training could be more tailored to the specific language and cultural needs of Hispanic community members in a given area or region of the country. All of the above literature points to the need to develop a core of trained Hispanic health care professionals, and the resources needed in the institutions of higher learning to provide this training where it is not already available. Because of the lack of training in the Midwest available to persons wanting to specialize in bilingual Hispanic health care, a common meeting ground for these persons has become the courses in language and culture

Language as a Barrier to Health Care

157

that are offered outside the medical and public health professions. For the most part, persons in the fields of history, anthropology, political science, Chicano studies, Spanish and education have been dealing with the issue of empowerment of Hispanics for several decades in the institutions of higher education in the US. In fact, it was in the context of a Chicano studies course dealing with language issues, that the impetus for gathering together the information on Hispanic health care in the Midwest and for continuing work on a bilingual manual for nurses, social workers and physicians took place.

The Twin Cities Hispanic Community Our investigation to document the services available to patients who are not fluent in English convinced us that, generally, the community is not aware that as many as 35,000 to 38,000 Latinos live in the Twin Cities area. This may or may not include some of the 600 to 800 recent Cuban refugees in the state of Minnesota (U.S. Immigration Department, St. Paul, MN, 1982). In addition, demographic data collected in an important study, Ramsey County Mental Health Needs Assessment (Garza, 1980), suggests that many Hispanics would welcome bilingual personnel in clinics where they seek medical attention (see Table 1).

Table 1 : Results of Ramsey County needs assessment. 14.12% - -— Considered Spanish the language in which they are most comfortable 51.77% - -- - Use and prefer a combination of Spanish and English 34.12 % - • - Prefer English 33.33 % - — Indicated a need for translation services on a daily basis when encountering monolingual English speakers Source: Gaiza (1980).

While we recognize the problems associated with the 1980 census, we know also that it does not accurately reflect the current growth of the Latino community. Among other things, it does not include the recent immigrants from Cuba and other Latin American countries. We can safely state that the population of Spanish speakers in the state is growing. Yet, there has been little corresponding growth in awareness on the part of the majority Anglo community (Cisneros and Leone, 1983).

158

Deborah Martinez, Elizabeth A. Leone and Jennifer Sternbach

A study by the Minnesota Historical Society in 1977 indicates that the Spanish language seems to have increased in use in the Twin Cities (Saucedo, 1977). A 1977 survey by the St. Paul Public Schools (1977) indicates that 84 percent of parents or guardians of Mexican-American children are either Spanish-dominant or bilingual and use Spanish often at home. Both the number of persons using Spanish in the Twin Cities and the increase in the number of Spanish-speaking persons migrating to the Twin Cities is welldocumented. From this information we infer an urgent need for translation services in stressful or technical situations such as those encountered in a hospital setting.

The Hispanic Health Care Coalition The Hispanic Health Care Coalition (HHCC) is composed of participants from a variety of organizations in the Twin Cities area working together to provide better medical and social services to Latinos. What follows is a brief documentation of the formation of the HHCC, its goals, and its project to improve translation service to the area's major medical centers. The promise of this coalition can be attributed to a combination of factors, including the crossing of the geographical barrier of the Mississippi River and the participation of persons who are themselves Latino and Spanish-speaking. The HHCC recognizes that Hispanics live not only in the neighborhood of St. Paul's west side, long seen as the Latino barrio, but also in various other parts of both metropolitan areas. The coalition represents, therefore, not only the Twin Cities but also the various agencies that serve the Hispanic community. Included are the following groups: Catholic Charities Chicanos, Latinos Unidos en Servicio Hispanos en Minnesota Minnesota Cuban Refugee Committee Ramsey County Community Human Services Dept., Minority Services Office Spanish Speaking Seniors Program, Our Lady of Guadalupe Church West Side Health Center

The HHCC was formed to address the following questions: (1) Where do Hispanics go for medical treatment? (2) What services are available to them? (3) What is the influence of language, culture, and traditional attitudes toward health care on the accessibility of these services and on the effectiveness of care received?

Language as a Barrier to Health Care

159

(4) Does the Hispanic consider available services to be adequate? (5) Is there a need for an assessment of Hispanic health care services? (6) Who are the groups, agencies, and individuals interested in the question of medical services for the Hispanic population? As indicated, the Hispanic Health Care Coalition is composed of many of these groups and individuals, and together the persons in the HHCC are in the process of answering some of the above questions. Some of the problems that recur in areas such as the twin cities where translation services are not available to Spanish-speaking patients have been documented by persons working with the HHCC. For example, the following types of cases are common: (1) the patient waits for treatment (sometimes causing needless complications) until a translator can be found; (2) the patient and physician experience a misunderstanding about symptoms or diagnosis, causing at times incorrect and dangerous results; (3) the translator arrives too late to be of any help; (4) a person who speaks Spanish arrives but is not trained in the technical terminology needed for effective translation; (5) a person who speaks Spanish arrives but is not fluent in Spanish nor familiar with the Spanish-speaking cultures, causing ineffective translation; (6) the patient is provided with a translator but because the person is not bicultural (or familiar with the culture) the patient does not ask the usual and necessary questions to ensure full consent and participation in the health care process; (7) the patient is given a translator who is trained only in the technical terminology in Spanish but not the lay terminology, thus rendering translation ineffective; (8) the patient is asked if she needs translation services, and this autoidentification system by the monolingual Spanish-speaking person is less accurate, excluding some persons from the needed service (i.e., other identification systems may be more effective). The above list of cases or hypothetical situations that commonly occur when translation services are inadequate or non-existent can be elaborated, but they serve to make the point well: the lack of professional translation services can result in second-class health care for the limited English speaking person. With the above questions to be addressed and problems of the lack of services documented, the HHCC sought to establish alternative health care services for Latinos in the Twin Cities area. Although the HHCC represents

160

Deborah Martínez, Elizabeth A. Leone and Jennifer

Sternbach

public and private agencies and individuals, there were certain common interests and goals. All members of the HHCC wanted to define the basic framework that would guide the actions of the group. These common goals are: (1) That the groups in the HHCC would support each other, in the search for funding sources for health and other services for Hispanics; (2) That the HHCC would identify the needs not yet met in Hispanic health care and then propose better ways to meet these needs and advocate for funding to do so; (3) That the HHCC would advocate that the health care services for Hispanics operate more effectively; (4) That the HHCC would provide a network for communication among Hispanics, in the health care area and in other related services. These goals, formed in response to the documentation of health care inadequacies by individuals and groups in the HHCC of the Twin Cities, are the starting points of the coalition. One project in progress that seeks to carry out the coalition's goals is a series of negotiations with the Spanish-speaking task force of a large hospitalmedical center in the area. These negotiations focus on translation services. As a result of a meeting with representatives of the hospital, the HHCC formulated eight points which are essential for a translation service in the hospital-medical center and its clinics: (1) Translating in the health care field is a professional activity that requires training and certification leading to a license in translating. (2) All official documents (applications, prescriptions, brochures, instructions, etc.) that are used by the patient need to be translated into Spanish. (3) It is necessary to identify as soon as possible the process by which the patients are identified as needing translation services. (4) Patients' medical records and administrative records must be clearly marked in order to facilitate the conveying of translations (translated information) to both physician and patient. (5) It is the responsibility of the medical center to publicize and disseminate information to the community to inform them of the existence of the new translation services. (6) Both the medical center and the HHCC will investigate the procedures of other hospitals regarding translation services. (7) The medical center and the HHCC will collaborate in seeking funds for the improvement of services for Hispanics. (8) The medical center will investigate the possibility of sharing translation services with other medical centers and hospitals in the area.

Language as a Barrier to Health Care

161

The coalition has asserted that it is the moral obligation of the medical center to assist in the establishment of translation services because communication to the patient and from the patient is an essential part of health care services. Unfortunately, in a time of cutbacks in budgets, it is difficult to expect that a hospital will respond to issues such as those addressed by the HHCC, since it is probable that there are no funds available for the provision of these services. It must be noted, however, that a medical treatment or health service once well understood does not often need to be repeated. That is, the use of adequate and effective translation is one way to ensure the effectiveness of the treatment and the unnecessary cost of repeated treatments not assisted by translation. One hopeful note in awaiting implementation of translation services in the Twin Cities area medical centers is that currently there is an overabundance of hospital beds. If translation services are offered, these services could play a part in attracting a clientele not currently being served, thus filling a demand of Hispanic patients for better care and of hospitals for an improved, costeffective use of their facilities. This is just one of the arguments that the HHCC hopes will be useful in encouraging the institutionalization of translation services in the twin cities health care centers, in order to replace the haphazard, ineffective, and potentially unsafe practice of providing translation by untrained friends and relatives of the patients, or by non-fluent, non-bicultural staff also untrained in medical and lay terminology in Spanish. In conclusion, the future of Hispanic health care services in the Midwest does look promising. One excellent translation program, for example, has recently been formalized at the University of Iowa, Iowa City Hospitals. Maria Martinez et al. at the University of Iowa have established a variety of training manuals for personnel in health care and related professions, and these translation documents are the result of several years of work with Hispanics in Iowa by health care and other social service personnel (Martinez, 1977; 1982). At a recent conference (May, 1982, Our Lady of Guadalupe Church, St. Paul) on Hispanic health care there was much discussion of many of the same issues that have been raised by the HHCC. So, little by little, the participation of Hispanic community persons and Hispanic health care professionals in the advancement and promotion of translation services is taking place in the Twin Cities area. In time, just as the HHCC has begun to establish channels of communication among Twin Cities area medical centers, regional and national channels of communication will be set up and resources shared so that public health professionals and other Hispanic social service workers can obtain the necessary training, certification, and internship experience to better participate in and improve the delivery of these services to the Spanish-speaking communities in the United States.

162

Deborah Martínez, Elizabeth A. Leone and Jennifer Sternbach

References Alvarado, Anita. (1980). Hie Status of Hispanic Women in Nursing. In Margarita Melville, (ed.). Andrade, Sally. (1980). Family Planning Practices of Mexican Americans. In Margarita Melville, (ed.). Cisneros, René and Leone, Elizabeth A. (1983). Mexican American Communities in the Twin Cities: An Example of Contact and Recontact. In Spanish in the US Setting: Beyond the Southwest, Lucía Elias-Olivares, (ed.). Rosslyn, VA: National Clearinghouse for Bilingual Education. Garza, Roy. (1980). Ramsey County Mental Health Needs Assessment Report - Clinical Implications for Policy and Program. St. Paul, MN: Ramsey County Community Human Services Department. Kay, Margarita A. (1980). Mexican, Mexican American and Chicana Childbirth. In Margarita Melville, (ed.). Manzanedo, Hector Garcia; Walters, Esperanza García and Lorig, Kate R. (1980). Health and Illness Perceptions of the Chicana. In Margarita Melville, (ed.). Martinez, Deborah. (MS). Personal Communication and Journal Notes. Martínez, Maria. (1977, 1982). Communicating with the Spanish-Speaking Patient. Trans, and rev. Ozzie T. Díaz-Duque and Temo J. López. Iowa City, Iowa: University of Iowa Hospitals and Clinics, Department of Social Services. Melville, Margarita. (1980a). Introduction. In Margarita Melville, (ed.). — (1980b). Selective Acculturation of Female Mexican Migrants. In Margarita Melville, (ed.). — (ed.). (1980c). Twice a Minority: Mexican American Women. St. Louis, Mo.: C.V. Mosby. Naranjo, M.S. (1976). Cross-cultural Comparison of Social Psychological Factors Contributing to the Effectiveness of Genetic Counseling Therapy. Unpublished Master's Thesis, the University of Texas at Austin. St. Paul Public Schools. (1977). Title VII Continuation Proposal. St. Paul, MN. Saucedo, Ramón. (1977). Mexican Americans in Minnesota. St. Paul, MN: Minnesota Historical Society. Sternbach de Medina, Jennifer. (MS). Personal Communications and Journal Notes. Urdaneta, Maria Luisa. (1980). Chicana Use of Abortion: The Case of Alcala. In Margarita Melville, (ed.). Velasco, Maria. (1981). Personal Communication and Journal Notes. Minneapolis, MN.

Language as a Barrier to Health Care

163

Velasquez, Joan and Velasquez, C.P. (1980). Applications of a Bicultural Assessment Framework for Social Work Practice with Latinos. The Family Coordinator, (Summer). Vélez-I, Carlos G. (1980). The Nonconsenting Sterilization of Mexican Women in Los Angeles: Issues of Psychocultural Rupture and Legal Redress in Paternalistic Behavioral Environments. In Margarita Melville, (ed.). Wagner, Roland M. and Schaffer, Diane M. (1980). Social Networks and Survival Strategies: An Exploratory Study of Mexican American, Black, and Anglo Female Family Heads in San Jose, California. In Margarita Melville, (ed.).

Testing Bilingual Language Proficiency: An Applied Approach by Florence Barkin-Riegelhaupt

In spite of the fact that sociolinguistics has progressed considerably in its methods of inquiry and analysis, a basic dichotomy exists between its theoretical contributions and their application to practical situations. Accepted premises based on sociolinguistic concepts such as language varieties (regional and social), role relationships, situation, and repertoire range have had little if any impact on actual formulation and implementation of language policy and language planning. Certain areas beg for an application of findings in sociolinguistics. One such area is that of testing language proficiency for specific purposes. This paper will address the feasibility of applying a model based on sociolinguistic and ethnographic research to a practical end, i.e., testing language proficiency of children and adults who must be able to function in two languages. This model has been used at the elementary, secondary and university levels for the purpose of placement, language diagnosis and language development, and at the professional level for the purpose of diagnosing language proficiency for specific vocational purposes such as teaching, translating and interpreting, secretarial positions, etc. This paper offers a model for the development of any or many tests. Each examination allows for modifications as the need arises. It recognizes the importance of carefully documenting and analyzing the language of the particular group in question, i.e., the people with whom the examinee must interact, and the actual language needs of the individuals being tested. Examples for the purpose of illustrating the implementation of the model will be taken from the process and product developed by the Arizona Spanishlanguage task team, which has designed an instrument for the purpose of testing the Spanish-language proficiency of present and future teachers in bilingual programs. In order to develop a language proficiency examination in an area such as bilingual teaching certain essential questions must be addressed (Barkin, 1981). Part of the problem in developing examinations for state certification

166

Florence

Barkin-Riegelhaupt

of bilingual teachers in the language(s) of instruction is the fact that the state departments of education, the departments of foreign languages and the colleges of education are uncertain about the actual responsibilities of the bilingual teacher, how the bilingual programs are organized and administered, what the community attitudes are toward bilinguals and bilingual education, the relationship between the majority and minority cultures in the schools, and their mutual responsibility in addressing these issues. There is no doubt that testing language proficiency for almost any purpose is a delicate question, one which must be treated with utmost care. No one likes to undertake the responsibility of determining what actually constitutes 'a successful teacher, secreatary, etc., on the basis of language skills alone'. Anyone who has worked with bilingual teachers is aware of the fact that language abilities alone do not insure success as a bilingual teacher. However, it is expected that the ability of the teacher to effectively communicate with students is predicated upon an ability to express ideas, communicate concepts and oversee their implementation. Examinations designed to measure language proficiency for specific purposes cannot ignore the actual functions expected of the person to be tested after passing such a test. Too many examinations have been based on what has been called 'underlying linguistic competence' or 'global language ability' rather than the actual reality of the task to be enacted. In clear cut cases such as these where we know or can discover the functions of a person in a particular position, our language diagnosis need not be based on abstract concepts of language proficiency when what is needed is a reality-based language proficiency instrument. Ultimately performance in the language is of key importance. The first step in developing a model for testing language proficiency for specific purposes is to clearly outline the tasks the individual in a particular position has to perform. The process of uncovering these functions is one that requires keen observational techniques based on ethnographic models. For example, in the area of developing a test for bilingual teachers, it is necessary to document the reality of the teacher in the classroom. This documentation should include the whole range of necessary skills from use of simple commands in the target language, to full instruction in that language. In addition to the investigation of such details, it is also necessary to move beyond the present reality to future projections. It must be remembered that a bilingual teacher, once certified, will continue in that position despite changes in programs. The task of documenting the present reality cannot be accomplished by only one individual. The model I propose involves appointment of representatives from various sectors of the population to be addressed. For example, in the development of an instrument t o test bilingual teachers, it is essential to know about the language needs of a bilingual teacher in the target language

Testing Bilingual Language Proficiency:

An Applied

Approach

167

from the point of view of parents, children, and administrators and test designers as well as the observations of teachers in the classroom. Testing language proficiency of teachers is a political issue as well as a social reality. The more representative the group involved in consultation during the process of examination design, the more likely that the results will reflect both the realities of the bilingual teacher as well as other individuals' perceptions of these realities. Examples for this paper will be selected from the deliberation of the Spanish-language task team in the State of Arizona. Two committees were appointed. The first committee was made up of representatives from the state department of education, colleges of education, departments of foreign languages from the three universities in the state and one representative from the community colleges. This first committee was called the 'design committee'. These individuals are all active in either language instruction or bilingual content areas. The second committee, comprised of representatives solicited by members of the first committee, is made up of parents, children, administrators and legislators active in developing educational policy and/or with children in the public school bilingual programs. The review committees work closely with the representative to the design committee whose task it is to ultimately come up with an instrument that takes into consideration all the above issues. While the philosophy of the individuals involved in both committees encompasses a basic belief in quality education for all and in the importance of providing the best education in the target language, the approaches vary. For example, in the present committee attitudes toward local, so-called non-standard varieties of Spanish vary from complete acceptance of any variety of the language which communicates the message to 'correctness' and recognition of a 'universal Spanish language norm'. A committee with such varied views allows for an open forum in which each of the issues is treated in-depth and a compromise is reached after extensive debate. Further issues are clarified along the way. In the case of the state examination issues about standard/non-standard varieties of Spanish were debated in detail and decisions were made according to the specific tasks at hand. For instance, if the teacher were to write a formal letter or translate an announcement from English to Spanish, spelling errors, non-standard syntax and lexicon, etc., will not be permitted, without some penalty, whereas in actual oral communication with the students certain non-standard, widely used and understood forms are more acceptable. These issues are clearly reflected in the elaborate system of grading, which assigns the same number of points to non-standard as to standard varieties in certain areas while not in others. Once the selection of committee representatives of the various sectors of the community to be addressed is complete, the committee members must be informed of the task at hand. However, it must not be assumed that because

168

Florence Barkin-Riegelhaupt

these people have been solicited for a specific task that they understand its intricacies and implications. The eight member task force made up of members who agreed to work on a Spanish-language proficiency instrument spent the first two three-hour sessions deciding on whether they wanted to be involved in such a task. Philosophies about language proficiency testing, standard/non-standard varieties, state teacher certification, actual bilingual programs in the state, the importance of literacy in the target language, etc., were exchanged. In addition to actual philosophies about these issues, problems were discussed. At the end of the second meeting a vote was taken in favor of cooperatively designing an instrument and the actual work was about to begin. Nevertheless, each of the above issues continued to surface throughout the process and had to be addressed at every point during the design phase. Once there was a stated public commitment to the task at hand, the identification of the tasks became easier. For the state task force it was decided that the test we designed would be reality-based and reflect the concept of a proficiency test as proposed by Clark (1975: 10): Any measurement procedure aimed at determining the examinee's ability to perceive or transmit information in the test language for some pragmatically useful purpose within a real-life setting.

It was also decided that the test should reflect current research findings in bilingual education, first and second language acquisition, psycho- and sociolinguistics, micro- and macro-ethnography and language testing. The committee recognized the importance of testing in some way or another receptive and productive skills (listening, speaking, reading and writing) even though many bilingual teachers were not required to do more than communicate orally in the language of the students only when absolutely essential. The committee felt that it was their task to accurately assess the future needs since teachers would be certified 'for life'. Each of the members of the committee was assigned to observe in the classroom and/or prepare material from their present files, on language use and needs in specific areas. For example, when a committee member was involved in training teachers, he/she would be asked to survey a teacher and future teacher perceptions of their duties and language needs as well as their actual language use. Since all representatives of the task team were involved at one level or another in the training of teachers in language and/or content areas, this task became a natural outgrowth of their experience and observations. Following the compiling of this information, a list was drawn up of all of the points identified and these were discussed in-depth at every step. Naturally, the quantity of tasks was far too great for all to be included. Nevertheless, the committee made the ultimate determination of which areas

Testing Bilingual Language Proficiency: An Applied Approach

169

should be included in the examination on the basis of their importance as expressed by the people consulted, their own observations, and consultations with language testing experts and final deliberations with other committee members. One of the concerns in a task such as this is the fact that while the examination must reflect the reality of the teacher which is complex and extensive, it must also be relatively short and effectively measure language proficiency and not pedagogy. This point was a continual issue during the committee's deliberations. Our role was not to see whether a person knew how to teach but rather was able to effectively communicate in the target language. Naturally, the idea of effectively communicating involves an aspect of expertise in the area of specialization, which in this case was elementary education. However, the committee agreed to provide contexts for the activities which included the lesson to be taught, or task to be completed along with its importance to a bilingual teacher. Questions about sequencing of lessons which crossed the areas of memory, language proficiency and knowledge of content were discussed in detail and efforts were made to make the final instrument as context — complete and language - and not pedagogy-oriented as possible. This issue is clearly reflected in the development of grading criteria, where the interaction of content and language is reflected but lack of knowledge of appropriate pedagogical methods is played down. In a sociolinguistic/ethnographic approach to testing language proficiency, in addition to actual content, the importance of context and the functions of language related to specific tasks must be recognized. The committee, therefore, not only identified teacher needs, language use and attitudes but also documented the functions of language in each of the areas. Questions about how bilingual teachers have to use the language and for what continue to surface. It was noted that skills in praising, requesting, commanding, reprimanding, arguing, eliciting information, questioning, etc., were essential to effective teaching (Guntermann 1979, 1982). Therefore, communication content was not the only concern of the committee. Issues such as cultural and/or professional appropriateness as well as grammatical correctness, fluency, content and style were all considered. A system of grading of written and oral work based on such issues was instituted and each area was assigned a particular weight according to its relative importance in the entire picture of the language proficiency of the bilingual teacher. I would like to point out that the issues mentioned here have been consistently identified as the essential ones in testing the Spanish language proficiency of future teachers. The following list represents a summarized version of the sections of the examination which reflect committee members' reports from their observations. Each of the sections reflects emphasis on one particular language skill, i.e., listening, reading, speaking and writing. Never-

170

Florence Barkin-Riegelhaupt

theless, the committee thought it inappropriate and unrealistic to attempt to completely separate these skills since a teacher is required to be able to integrate them in many of the activities undertaken. The list reflects both functions of language and competence in the language in specific areas. Each section of the examination is described and some of the issues which came up during discussion are identified. Many of these issues are addressed in the grading system which is included following each section. The exam is based on a total of 100 points. It is essential to point out that the areas identified here are not the only important ones nor do they represent a prioritized list of skills needed. Rather, they are some of the important areas identified by students, teachers, community members, parents and legislators for success as a bilingual teacher.

Outline of Examination Section 1. Oral Comprehension of Students

(10 points)

Justification. One area that has been considered of utmost importance for the success of the bilingual teacher is understanding children's discourse among themselves and with the teacher. Many teachers claim that while they can present material in the target language they are often at a loss when the children communicate their ideas. Sometimes this is due to cultural differences and a non-mutual context, while at others it is due to a teacher's inexperience with the speech of children and/or the regional dialect. Description of section. In this section the examinee must demonstrate comprehension of children's speech by answering questions following a videotaped presentation of children's classroom interactions. There are ten short scenes involving children's interacting in natural settings. Following each one the examinee must write a short statement identifying the major points communicated by the child. Each correct response is worth one point. Grading scale. Grading for this section is relatively simple. The evaluator must determine from the written response of the examinee whether or not the children's utterances were understood. Examinees are graded solely according to their demonstration of comprehension of the children's speech and not their own production either orally or in writing. Section 2: Oral Reading (5 points) Justification. One of the most important activities of an elementary school teacher especially in the early grades is the reading of stories to the class. Bilingual teachers often state that they avoid such an activity because certain words are hard to pronounce and the children often do not understand the

Testing Bilingual Language Proficiency:

stories. From our sion comes from patterns, rhythm read by teachers, being read.

An Applied

Approach

171

observations, we have noted that often such miscomprehenthe teacher's inability to read with the correct intonation and pronunciation. Sometimes, by listening to selections it is obvious that teachers do not comprehend the passages

Description of the section. The examinee will read aloud and record on tape, with expression, a short literary selection, chosen from a book commonly read aloud to children, using correct pronunciation, intonation, word groupings. Grading scale. The grading scale for this section reflects the committee's concern for the importance of oral reading as a precursor to and a precipitator of development of reading skills in children. The examinee must be able to read in such a way that the story is communicated in an enjoyable and entertaining as well as comprehensible way. The following system for delegating points is shown in Table 1. Table 1: Grading of oral reading. 0 2 4

The examinee has a lot of difficulty with the pronunciation of the words, his/her intonation and rhythm do not correspond to the selection being read. The examinee reads with the appropriate expression for the selection read but mispronounces some of the words; or the examinee pronounces all of the words correctly but his/her expression does not correspond with the selection read. The examinee reads with the appropriate and correct expression, intonation and rhythm and pronounces all of the words correctly.

Section 3: Oral Presentation of an Instructional Activity (15 points) Justification. The committee identified as essential to the actual functioning of the teacher the ability to teach a lesson. Description of the section. In this section the examinee is asked to present a lesson, as if teaching a group of elementary school children. The lesson to be taught is based on instructions provided in a teacher's guide commonly used in bilingual classrooms. Grading scale. This section was deemed the most complex and important by the committee. Therefore, it is worth more points than any other individual section (15 points). The grading scale (see Table 2) is based on both grammatical and such linguistic criteria! as organization of discourse (to include correct sequencing when necessary for the comprehension of the lesson, etc.)

172

Florence Barkin-Riegelhaupt Table 2: Grading of oral presentation.

0

The examinee does not say anything or what he/she says does not relate to the activity. 5 The examinee provides some sentences but does not explain the entire activity or does not explain it in the necessary and appropriate order. 10 The examinee explains the activity in complete sentences, and in the correct and appropriate order but commits grammatical errors and/or repeats the words from the teacher's guide (does effectively vary lexical items). 15 The examinee explains the activity in his/her own words, using complete sentences without errors in grammar.

Section 4: Question Formulation (10 points) Justification. The committee identified teacher questioning strategies specifically pertaining to content areas and readings done by the students as an important teacher activity. Description of section. In this section the examinee is requested to formulate questions relating to a reading selection as if these questions were being posed to a group of elementary school children. Grading scale. The scale for this section involves grammatical and content components, for each question formulated (see Table 3).

Table 3 : Grading of question formulation. 0 The examinee cannot formulate the question. 1 The examinee tries to formulate the question but cannot use complete sentences and/or the question does not relate to the content of the story; he/she commits many grammatical errors. 2 The examinee formulates the appropriate question based on the content of the story without grammatical errors.

Section 5. Technical Vocabulary (10 points) Justification. The committee did extensive research in the particular areas of vocabulary most needed by the bilingual elementary school teacher. One of the most common deficits noted by teachers is that of the appropriate vocabulary in Spanish for things they have learned in English.

Testing Bilingual Language Proficiency: An Applied Approach

173

Description of section. The examinee will translate English vocabulary items into Spanish. This section includes some of the areas identified as being particularly important to the daily functioning of the bilingual teacher. Sections include translation of mathematical terms, educational terms such as eraser, sentence, board, etc., dates and times, common classroom phrases (commands, etc.), and terms relating to family members. Grading scale. Since this section involves the correct translation of an item from English to Spanish, it is considered right or wrong. Each item is worth a half point.

Section 6: Oral Communication with Parents (12 points) Justification. One of the important tasks that must be carried out by the bilingual teacher is communicating any problems or success to parents. The bilingual teacher in many cases is the only link between the parent and the school since he/she is often the only individual who can communicate with parents in their language. Description of the section. In this section the examinee must demonstrate that he/she can communicate orally with parents using a professional and culturally appropriate style to the context and situation, while demonstrating his/her ability to use correct grammar, vocabulary, pronunciation, and fluency. This section involves responses to a contrived phone conversation where the parent's voice is provided. Grading scale. The grading scale for this section takes into consideration social and contextual appropriateness, grammar, vocabulary, and naturalness and completeness or fluency (see Table 4).

Table 4: Grading of oral communication with parents.

Appropriateness Grammar Vocabulary Naturalness and Completeness (Fluency)

Level

Level

Level

Level

0 0 0 0

1 1 1 1

2

3 3 3 3

2 2 2

174

Florence

Barkin-Riegelhaupt

Definitions: Appropriateness (Social and Contextual) 0 The responses were generally not professional and/or appropriate to the context. 1 Half or fewer of the responses were appropriate or professional 2 Only one or two of the responses was not completely appropriate or professional 3 All of the responses were appropriate and professional. Grammar 0 Several serious errors in grammar were committed. 1 A few errors were committed. 2 No serious errors were committed; while the usage was not that of an educated native speaker, it was understandable. 3 The use of the linguistic forms and structures was native/near native and educated. Vocabulary 0 Vocabulary was inadequate for getting the message across. 1 Choice of words was definitely non-native and sometimes inadequate or incorrect. 2 Words or expressions were near-native; or native but regional, colloquial, or anglicized. 3 Vocabulary was that of an educated native speaker of Spanish. Naturalness and Completeness (Fluency) 0 Communication was virtually impossible; or so brief that it was incomprehensible. 1 Expression was slow and uneven but sufficient to get the meaning across. 2 Speech was perceptively non-native but adequate to the task. 3 Expression was natural, native and completely clear. In addition to the oral section which is recorded on tape, there is a section in which the examinee is required to demonstrate appropriate written competencies. The following includes activities for this section. Section Notice

7: Translation

of an Official Announcement

(12 points) and/or

Justification. Even though it has been pointed out that the bilingual teacher is a teacher and not a translator and/or administrator, one of the duties which often falls into the hands of the teacher is the translation of notices to be sent home with the children. For this reason a section measuring the examinee's ability to perform such a task has been included. Description of the section. The examinee will be given an announcement to be translated from English to Spanish while being expected to use correct grammar, spelling, vocabulary, style and content. These selections have been chosen from actual announcements circulated to students in the schools.

Testing Bilingual Language Proficiency:

An Applied Approach

175

Grading scale. In this section the grading scale reflects the committee's awareness of regional and colloquial varieties and anglicisms which in some cases have become integrated into the local dialects (see Table 5).

Table 5 : Grading translation of a document. Grammar and Use of Structure. 0 Several serious errors in grammar were committed. 1 A few errors were committed. 2 No serious errors were committed; while the usage was not that of an educated native speaker, it was understandable. 3 The use of the linguistic forms and structures was native/near native and educated. Spelling and Accentuation 0 The examinee cannot express him/herself in writing. 1 The examinee commits many errors. 2 The examinee commits a few errors. 3 The examinee does not commit any errors. Use of Vocabulary 0 Vocabulary was inadequate for getting the message across. 1 Choice of words was definitely non-native and sometimes inadequate or incorrect. 2 Words and expressions were near-native; or native but regional, colloquial, or anglicized. 3 Vocabulary was that of an educated native speaker of Spanish. Style and content 0 Translation was incomprehensible. 1 The meaning was incomplete and/or style was awkward. 2 Complete meaning was expressed but style was inappropriate and/or unprofessional 3 Complete meaning was communicated in an appropriate and professional style.

Section 8: Reading a Professional Journal (14 points) Justification. One of the most common complaints of bilingual teachers is the fact that they would like to be exposed to professional written material in Spanish, but that the level of difficulty of such material makes it impossible for them to understand. This section was designed to test the ability of the examinee to read and comprehend a selection from a professional journal. Description of the section. The examinee will read a short selection from a professional journal and write a summary of it and comment on its content, using proper grammar, spelling, vocabulary and style. Grading scale. The grading scale for this section is divided into two parts. The

176

Florence

Barkin-Riegelhaupt

first part is a scale from 0—8 points on which the evaluator must determine the degree of comprehension of the selections as indicated by a 50- to 75word summary of the selection. Grammar and level of vocabulary, etc., are not graded in this part. In the paragraph, about his/her personal reactions to the content of the selection, is grading is according to both form and content (see Table 6). Table 6 : Grading reading of a professional journal. Comprehension

0 Did not understand selection.

1234567

8 Shows complete understanding of selection.

Form (Grammar, Spelling and Vocabulary) 0 Several serious errors in grammar were committed; examinee could not express himself in writing; vocabulary was inadequate for getting the message across. 1 A few errors in grammar were committed; many errors in spelling and use of accents; choice of words was definitely non-native and sometimes inadequate or incorrect. 2 No serious errors in grammar were committed; while the usage was not that of an educated native speaker, it was understandable; few errors in spelling and use of accents were committed; words and expressions were near-native; or native but regional, colloquial, or anglicized. 3 The use of the linguistic forms and structures was native/near-native and educated; no errors in spelling or use of accents were committed; vocabulary was that of an educated native speaker of Spanish. 4 Vocabulary was that of an educated speaker of Spanish. Style and Content 0 Translation was incomprehensible. 1 The meaning was incomplete and/or style was awkward. 2 Complete meaning was expressed but style was inappropriate and/or unprofessional. 3 Complete meaning was communicated in an appropriate and professional style. Section 9: Reading Student Compositions (12 points) Justification. Our observational research has shown that all teachers are involved in evaluating the written work of their students. Often they would like to point out problems in grammar and spelling, as well as clarity or expression, etc., but are unable to do so because they are not familiar enough with rules of spelling, grammar and accentuation. Description of section. The examinee will be given student compositions and will rewrite any incorrectly written word or group of words. Many of the words underlined are common mispellings due to confusion of v/b, y/11, problems in division of words such as yeste, etc., and accent problems.

Testing Bilingual Language Proficiency: An Applied Approach

177

Grading scale. In this section the examinee is told that the underlined words may not be incorrect and therefore is required to correct only the ones which are incorrect. A half point is awarded for each word which is identified as correct or incorrect and if incorrect is corrected.

Revision of Examination Each section of this examination went through approximately five stages of revision not only in terms of the content but in terms of the réévaluation of objectives and interpretation based on the grading scales. The examination was piloted using a group of teachers who already were certified as proficient in Spanish on the basis of another examination. These people's examinations provided us with some initial data both on the administrative and substantive areas of the exam. Since the pilot test given in May of 1981, the examination has been revised three times in both major and minor ways. It is constantly undergoing revisions and therefore is constantly evolving. Revisions are based on data extracted from the examinations which have undergone careful content and grammatical analysis. In addition, each examinee has completed a detailed questionnaire immediately following the exam. In the vast majority of cases, examinees' evaluations of each section of the examination in terms of its appropriateness, level of difficulty and importance indicate that the general philosophy underlying the development of the examination and its implementation are congruent with the expectations of the examinees. This, incidently, was not the case with the previous examination, which had used a more global approach toward testing language proficiency, and was not geared specifically to teachers. Comments provided by the pilot examinee questionnaire confirm that our approach to identifying important issues in testing teacher proficiency and the product, the examination itself, have effectively met their goals: Touches on real life situations typical for a bilingual teacher. Gives the test-taker real life problems that the bilingual teacher actually must encounter. You are giving them a taste of what is really out there! The situations given on the exam are situations that are encountered daily by a bilingual teacher. If a person is expected to know all these terms, he/she ought to be provided with a class before certification. Many people who grow up speaking Spanish have had no opportunities to learn the proper spelling, accentuation, technical terms, etc. I would hate to see them banned from the classroom. On the other hand, a Mexican surnamed person does not think this is an excellent attempt to improve the linguistic abilities and the quality of bilingual teachers.

178

Florence Barkin-Riegelhaupt

This model for developing language proficiency instruments based on performance and not 'competence' does not disassociate the two. It recognizes performance as essential to the tasks at hand. The process which I have outlined is ongoing. It can serve as a model for the development of other instruments. I have used it in modified form at two other sites, one is for an elementary school and the other for a secondary diagnostic/placement and evaluative examination. The development of an examination is predicated on the specific needs and wishes of the exam designers, those very people who will be involved in ultimately receiving results from these exams. Once the determination is made as to the particular needs of a school system, an exam can be designed to include only an oral section, or only a written section, etc. Such is the case with an examination presently being designed under my direction for the purpose of diagnosing and placing elementary school students in bilingual classrooms. Following the first meeting, teachers got together to determine which aspects they wanted to include in their examination. They became part of the very process of conceptualizing their own needs and approaches to testing and placement. For that purpose, the appendix to the examination includes an outline of the examination development procedure. It takes into account the time necessary for observations and presentation of findings, réévaluation, and sharing of information at all stages. Some of the meetings are with individual members of the committee rather than the whole committee, in order to carefully assess and modify the individuals particular contribution to the overall task. For example, during the initial design process of the secondary instrument, teachers from the areas of language arts, science, mathematics and social studies volunteered to have their classes recorded. Tapes were carefully edited and questions for a listening comprehension test were formulated in cooperation with the teachers involved.

Conclusion Language proficiency testing whether for the purpose of certifying individuals as competent to perform in a particular language, or for placement of students in bilingual elementary and secondary programs, must be based on the reality of the specific situation at hand. The proposed model for the designing of language proficiency instruments serves as an outline for procedures for evaluating proficiency in any situation; the examinations described here are appropriate for the needs of our particular populations. The model presented here has the following advantages: (1) It is flexible and can be modified as the situation changes. (2) It is custom-made to fit the needs of a particular situation, i.e., school system, community or perhaps even individual.

Testing Bilingual Language Proficiency: An Applied Approach

179

(3) It does not rely on an unrealistic model of language use, but rather reflects the language used in the community. (4) Once developed for a particular purpose, it can be easily administered and analyzed. (5) It involves the participation of administrators, teachers, students, and parents as well as university consultants and therefore, represents their needs and goals. (6) As individuals participate in the development of the instrument, they may become more aware of the needs of bilingual students and more willing to modify materials and approaches to the teaching and testing of these students. (7) As individuals take examinations they become more aware of their own areas of deficiencies and even in cases of a passing grade may decide to pursue further course work.

References Barkin, Florence. (1981). Evaluating Linguistic Proficiency: The Case of Teachers in Bilingual Programs. In Teaching Spanish to the Spanish Speaker: Issues, Aims and Methods, Guadalupe Valdes, Anthony Lozano and Rodolfo Garcia, (eds.). New York: Teachers College Press. Clark, John. (1975). Theoretical and Technical Consideration in Oral Proficiency Testing. In Testing Language Proficiency, Randall Jones and Bernard Sposky, (eds.). Arlington, Virginia: Center for Applied Linguistics, 10-29. Guntermann, Gail. (1979). Purposeful Communication Practice: Developing Functional Proficiency in a Foreign Language. Foreign Language Annals 12(3): 219-225. — (1982). Meaning and Function in Language Learning: Classroom Adaptation of Functional/Notional Concepts. Washington, D.C.: Center for Applied Linguistics.

Appendix Outline for Spanish Language Assessment Workshop on Designing an Oral Communication Examination for Elementary School Pupils. Week

Day

Topics

1

1

Orientation to procedures for testing language proficiency. Discussion of problems involved in trying to determine bilingual language proficiency view of instruments commonly used to place students in bilingual programs, e.g., BSM, LAS, etc. Decision by the teachers to

180

Florence

Barkin-Riegelhaupt participate in the task as to whether or not they want to undertake it and how (if response is positive). Participants are asked to begin observing students in their classes as well as themselves and others for language use, in a general way.

2

2

The time that has elapsed between the first and second meeting has been devoted to consideration of major questions and collection of initial data (see above). The language needs of the groups in question are identified. Domains and skills suspected as necessary for functioning in a Spanish language, English language and/or bilingual classroom are identified. The meeting is held as an open forum allowing for the free-flowing of ideas. Assignments are made for taping and detailing accounts of skills and areas to be tested.

3

3-4

4

5

Private interviews are conducted for the purpose of reviewing materials prepared by the teachers. Revisions are suggested and assignments made. Monday of the fourth week is devoted to discussion of second revisions. Should further changes be deemed necessary they will be done by Friday.

4

6

Sharing of information for discussion and further revision.

7

5

By Friday of the following week this typed tentative version will be available for further cuts and changes. Committee will work together at combining sections, etc. At the end of the section a form will be available for piloting.

7

6

Pilot tests to the selected students from grade levels in question. Evaluation and review of results.

8

7

Report of examination results and questionnaire about the test. Revisions based on results.

9

8

Administer the examination following revisions and retyping.

9

9

The examination is subjected to frequent revision procedures, discussions with me in consultation with individual committee members and other teachers and administrators as well as bilingual parents and children. This procedure is done intermittently throughout the first year of the examinations administration.

The Court Interpreters Certification Test Design by Etilvia Arjona

This paper focuses on the examination developed by the administrative office of the U.S. courts for the certification of federal court interpreters in the English and Spanish languages. The influx of significant non-English speaking populations into this country during the past few years has necessitated new public services. Provision of interpretation services for the hearing-impaired and for persons who do not speak the official language of the court is one such instance. The need for interpretation services, especially conferenceinterpretation services, is clearly acknowledged in other countries. In the U.S., however, the use of this type of multilingual service is still in its developing stages. This situation is particularly true in the courts, where the prevailing view has been a come-as-may attitude — service being provided, literally, by the first available bilingual speaker in the court at the time. This relaxed attitude is all too frequently accompanied by a disregard of appropriate interpretation services. Under such conditions, the accused, unfortunately, all too often finds his/her position distorted and his/her rights jeopardized by faulty interpretation. The need for a highly skilled professional in such cases is questioned in many quarters and only slowly is public awareness of the multiple issues raised by this matter growing. In recent years, however, increased concerns for the due process guaranteed by the Constitution, as well as increasing demands for minority rights, and the rights of all linguistically handicapped populations (such as the hearingimpaired), have led to the passage of Public Law 95—539 of October 28, 1978 by the 95th. Congress amending Chapter 119, Title 28 of the U.S.Code. This law provides 'more effectively for the use of interpreters in courts of the U.S. and for other purposes'. The director of the administrative office of the U.S. courts is charged with implementing a program to facilitate the use of interpreters in the federal courts as well as with prescribing, determining, and certifying the qualifications of persons who might serve as certified

182

Etilvia Ar joña

f ì t p W 1 g S -a .g ï % < S « 8 f I » fr« y z£ o.8a